Early Language Learning: Complexity and Mixed Methods 9781783098323

Citation preview

Early Language Learning

EARLY LANGUAGE LEARNING IN SCHOOL CONTEXTS Series Editor: Janet Enever, Umeå University, Sweden The early learning of languages in instructed contexts has become an increasingly common global phenomenon during the past 30 years, yet there remains much work to be done to establish the field as a distinctive area for interdisciplinary investigation. This international research series covers children learning second, foreign and additional languages in educational contexts between the ages of approximately 3 and 12 years. The series will take a global perspective and encourage the sharing of theoretical discussion and empirical evidence on transnational issues. It will provide a platform to address questions raised by teachers, teacher educators, and policy makers who are seeking understanding of theoretical issues and empirical evidence with which to underpin policy development, implementation and classroom procedures for this young age group. Themes of particular interest for the series include: teacher models and teacher development, models of early language learning, policy implementation, motivation, approaches to teaching and learning, language progress and outcomes, assessment, intercultural learning, sustainability in provision, comparative and transnational perspectives, cross-phase transfer issues, curriculum integration – additional suggestions for themes are also most welcome. Full details of all the books in this series and of all our other publications can be found on http://www.multilingual-matters.com, or by writing to Multilingual Matters, St Nicholas House, 31–34 High Street, Bristol BS1 2AW, UK.

EARLY LANGUAGE LEARNING IN SCHOOL CONTEXTS: 1

Early Language Learning Complexity and Mixed Methods

Edited by Janet Enever and Eva Lindgren

MULTILINGUAL MATTERS Bristol • Blue Ridge Summit

In memory of our good friend and colleague Magdalena Szpotowicz

DOI: 10.21832/ENEVER8316 Library of Congress Cataloging in Publication Data Names: Enever, Janet, editor. | Lindgren, Eva, editor. Title: Early Language Learning in School Contexts: Complexity and Mixed Methods/Edited by Janet Enever and Eva Lindgren. Description: St Nicholas House: Bristol; Blue Ridge Summit, PA, USA: Multilingual Matters, [2017] | Series: Early Language Learning in School Contexts: 1 | Includes bibliographical references and index. Identifiers: LCCN 2017007051| ISBN 9781783098316 (hbk : alk. paper) | ISBN 9781783098309 (pbk : alk. paper) | ISBN 9781783098347 (kindle) Subjects: LCSH: Language and languages—Study and teaching (primary)—Foreign countries. | Language acquisition—Foreign countries. | Language acquisition—Age factors. Classification: LCC P57.E8 E275 2017 | DDC 372.65—dc23 LC record available at https://lccn.loc.gov/2017007051 British Library Cataloguing in Publication Data A catalogue entry for this book is available from the British Library. ISBN-13: 978-1-78309-831-6 (hbk) ISBN-13: 978-1-78309-830-9 (pbk) Multilingual Matters UK: St Nicholas House, 31–34 High Street, Bristol BS1 2AW, UK. USA: NBN, Blue Ridge Summit, PA, USA. Website: www.multilingual-matters.com Twitter: Multi_Ling_Mat Facebook: https://www.facebook.com/multilingualmatters Blog: www.channelviewpublications.wordpress.com Copyright © 2017 Janet Enever, Eva Lindgren and the authors of individual chapters. All rights reserved. No part of this work may be reproduced in any form or by any means without permission in writing from the publisher. The policy of Multilingual Matters/Channel View Publications is to use papers that are natural, renewable and recyclable products, made from wood grown in sustainable forests. In the manufacturing process of our books, and to further support our policy, preference is given to printers that have FSC and PEFC Chain of Custody certification. The FSC and/or PEFC logos will appear on those books where full certification has been granted to the printer concerned. Typeset by Nova Techset Private Limited, Bengaluru and Chennai, India. Printed and bound in the UK by Short Run Press Ltd. Printed and bound in the US by Edwards Brothers Malloy, Inc.

Contents

Figures and Tables Contributors 1

vii xi

Introduction: Mixed Methods in Early Language Learning Research – Examining Complexity Eva Lindgren and Janet Enever

1

Part 1: Overviews of Research Findings 2

Early Language Learning in Complex Linguistic Settings: Insights from Africa Agatha J. van Ginkel

3

Considering the Complexities of Teaching Intercultural Understanding in Foreign Languages Patricia Driscoll

24

4

Literacy Development in Children with English as an Additional Language (EAL) Victoria A. Murphy

41

9

Part 2: Empirical Studies Using Mixed Methods 5

Verbal Working Memory and Foreign Language Learning in English Primary Schools: Implications for Teaching and Learning Alison Porter

6

Piecing Together the Jigsaw: Understanding Motivations of English Learners in Chinese Primary School through a Questionnaire and Elicited Metaphor Analysis 85 Jiang Changsheng, Zhang Jie, Liang Xiaohua, Yuan Yuan and Xie Qun

7

Codeswitching Your Way to Language Learning? Receptive Codeswitching with Digital Storybooks in Early Language Learning Judith Buendgens-Kosten, Ilonca Hardy and Daniela Elsner v

65

108

vi

Early L anguage Lear ning

8

Individual Differences and English L2 Learning in Two Primary Classrooms in France Heather Hilton

127

9

Content and Language Integrated Learning (CLIL): A Panacea for Young English Language Learners? Karmen Pižorn

145

Part 3: Longitudinal Perspectives Using Mixed Methods 10 The Dynamics of Motivation Development among Young Learners of English in China Yuko Goto Butler

167

11 Young Italian Learners’ Foreign Language Development: A Longitudinal Perspective Lucilla Lopriore

186

12 Employing Mixed Methods for the Construction of Thick Descriptions of Early Language Learning Eva Lindgren and Janet Enever

201

13 Developmental Aspects of Early EFL Learning Jelena Mihaljević Djigunović

222

Part 4: Evaluating Early Language Learning Programmes 14 Child EFL Interaction: Age, Instructional Setting and Development María del Pilar García Mayo and Ainara Imaz Agirre

249

15 Evaluating the Educational Outcomes of an Early Foreign Language Programme: The Design of an Impact Study for the Primary English Programme in Mexico 269 Peter Sayer, Ruth Ban and Magdalena López de Anda 16 The Development of a Curriculum-Based C-Test for Young EFL Learners Raphaela Porsch and Eva Wilden 17 Mixed Methods in Early Language Learning Research Janet Enever and Eva Lindgren Index

289 305 315

Figures and Tables

Figures Figure 2.1

Comprehension scores Dutch, English, Pokomo and Sabaot (Graham & Van Ginkel, 2014) 12 Figure 2.2 English scores of the Kom L1 based bilingual programme (Walter & Trammel, 2010) 13 Figure 2.3 Primary school English proficiency in Ethiopia (American Institute for Research, 2012) 15 Figure 4.1 PIRLS 2006: Reading scores by immigrant status at age 10 (OECD, 2012: 4) 44 Figure 4.2 The multiword phrase test (Smith & Murphy, 2015) 50 Figure 5.1 Test scores of higher and lower verbal working memory participants 75 Figure 5.2 Sample attempts at French spellings (extract 1) 76 Figure 5.3 Sample attempts at French spellings (extract 2) 76 Figure 5.4 Sample attempts at French spellings (extract 3) 77 Figure 5.5 Sample attempts at French spellings (extract 4) 77 Figure 5.6 Sample attempts at French spellings (extract 5) 78 Figure 5.7 Sample attempts at French spellings (extract 6) 78 Figure 6.1 Yr 1/Yr 3 motivational differences 95 Figure 6.2 Yr 1/Yr 3 intrinsic/extrinsic motivation (mean) 96 Figure 6.3 The major types of source domains for intrinsic/ extrinsic-motivation metaphors 98 Figure 7.1 MuViT screenshot 110 Figure 7.2 Simple examples of macro- and micro-switches 113 Figure 7.3 Number of macro-switches, plotted by duration in seconds and distance in forward pages 115 Figure 9.1 The distribution of communication, content and cognition dimensions 157 Figure 10.1 Students’ mean responses by time (grade levels) and SES (mother’s educational level) 174

vii

viii

Early L anguage Lear ning

Figure 11.1 Learners’ perceptions of the FL difficulties at age 13 Figure 11.2 Learners’ explanations for their comparison to their friends at age 13 Figure 11.3 Exposure to and use of English at home (age 13) Figure 11.4 Exposure to and use of English at home: Opportunities (age 13) Figure 11.5 Surfing the internet at home (age 13) Figure 11.6 Exposure to and use of English at home: Watching films in English Figure 11.7 Exposure to and use of English out of home: Travelling abroad Figure 11.8 Listening test results (age 13) Figure 11.9 Reading and writing test results (age 13) Figure 12.1 Illustration of Tomas’ achievements over six years Figure 12.2 Illustration of David’s achievements over six years Figure 12.3 Illustration of Mikael’s achievements over six years Figure 13.1 Development of the participants’ motivation (three-point scale) Figure 13.2 The participants’ preferences for classroom arrangements (%) Figure 13.3 Development of the participants’ self-concept (three-point scale) Figure 13.4 The participants’ outside school EFL behaviour as reported by their parents (%) Figure 13.5 Development of the participants’ accuracy and fluency in oral production (five-point scale) Figure 14.1 Age group differences in the percentages of use of conversational strategies at Time 1 Figure 14.2 Age group differences in the percentages of use of conversational strategies at Time 2 Figure 14.3 Instructional setting differences in the percentages of use of conversational strategies at Time 1 Figure 14.4 Instructional setting differences in the percentages of use of conversational strategies at T2 Figure 14.5 Percentages of conversational strategies at T1 and T2 in the CLIL group Figure 14.6 Percentages of conversational strategies at T1 and T2 in the mainstream group Figure 15.1 Progression in the Mexican programme Figure 15.2 The puppet study Figure 15.3 Where did the highest achieving students learn English? Figure 15.4 A student’s notebook describing a science experiment

192 193 194 194 195 195 196 196 196 210 213 214 230 232 233 234 235 258 258 259 259 260 260 271 278 280 282

Figures and Tables

ix

Tables Table 2.1 Table 4.1 Table 5.1 Table 5.2

Table 6.1 Table 6.2 Table 6.3 Table 6.4 Table 6.5 Table 6.6 Table 7.1 Table 7.2 Table 7.3 Table 7.4 Table 7.5 Table 8.1 Table 8.2 Table 8.3 Table 8.4 Table 9.1 Table 9.2 Table 9.3 Table 9.4 Table 9.5 Table 9.6

Pass rates at end of school exams in South Africa EAL versus non-EAL and achievement at age 5, 7 11 and 16: England 2013 Summary of constructs and tasks Summary of the influence of verbal working memory on FL Reading Aloud (FLRA), FL Reading Comprehension (FLRC), FL Receptive Vocabulary (FLRV) and FL Elicited Imitation (FLEI) at all test times Example of elicited metaphor analysis Background information of the participating schools Intrinsic/extrinsic-motivation metaphors in four schools Examples of intrinsic-motivation metaphors Examples of extrinsic-motivation metaphors The role of significant others Number of subjects per condition Number of different switch types in the logfiles (arithmetic mean, in trilingual treatment) Pearson’s correlation coefficients for measures of rCS and vocabulary gain Inter-rater reliability rCS ratings (trilingual treatment only) Individual variables measured in the Seine & Marne Primary English project Tasks used to measure emergent knowledge and skill in English Composition of the primary English groups, institutional and methodological variables Class means for measures of individual differences and English knowledge and skill, with group comparisons (column 4) Listening comprehension test results before and after the experiment The percentage of learners who improved their results on the listening comprehension test Reading comprehension and writing test results The percentage of learners who improved their results on the reading comprehension and writing test Learners’ attitudes to English language instruction prior to the experiment Learners’ attitudes to English language instruction after the experiment

16 46 71

73 90 92 96 97 97 97 114 117 117 119 120 132 133 135 137 150 151 151 151 152 153

x Early L anguage Lear ning

Table 9.7 Table 9.8 Table 9.9 Table 9.10 Table 9.11 Table 9.12 Table 9.13 Table 10.1 Table 10.2 Table 10.3 Table 11.1 Table 12.1 Table 13.1 Table 13.2 Table 13.3

Table 13.4 Table 14.1 Table 15.1 Table 16.1 Table 16.2 Table 16.3 Table 16.4

Learners’ improvement in their attitudes to English language instruction The students’ attitudes to English language learning prior to experiment Learners’ improvement in their attitudes to English language learning The students’ attitudes to 22 CLIL activities Students’ arguments why they liked or disliked the activities The students’ perceptions of the CLIL teaching approach Chi-square test for equal proportions Variables used for quantitative analyses in this study ANOVA with repeated measures (shaded if significant at the 0.05 alpha level) Main findings from the interview data Longitudinal study structure Overview of instruments used over the six-year period to collect data of production and reception achievements, attitudes and context Participants’ ages in grades 5–8 Administration of instruments Summary of significant (+) and non-significant (–) interactions of motivation, self-concept (Pearson correlation, n = 18) and language learning behaviour (t-test, n = 81). Summary of significant (+) and non-significant (–) interactions of accuracy and fluency scores with out-of-school exposure to English in grade 8 Distribution of the participants in the study Alignment of framework, competencies and research questions Descriptive data, mean (of sum score), difficulty (p), discrimination (rit) and reliability (α) Correlations of C-test scores and school grades in English Test-management strategies employed in the C-test Language learner strategies employed in the C-test

153 153 154 154 156 157 157 173 174 176 189 205 225 228

237 237 255 274 295 295 297 298

Contributors

Ruth Ban is an Associate Professor in the Department of Curriculum and Instruction at Barry University in Miami, FL. She holds a PhD from University of South Florida in Second Language Acquisition and Instructional Technology. Her research interests include professional and language identity of language teachers and teaching young learners. She participated in the development of the English language teaching programme for public primary schools in the state of Aguascalientes, Mexico. Judith Buendgens-Kosten is a postdoctoral researcher at Goethe University Frankfurt (Germany). She received an MA in English Linguistics, German Literature and Philosophy from RWTH Aachen University, a doctorate in English linguistics from RWTH Aachen University and an MA in Online and Distance Education from the Open University, UK. Her research focuses on computer-assisted language learning (CALL) and multilingualism. Patricia Driscoll is a Reader in Education and Director of Research Development at Canterbury Christ Church University, UK. She has taught in primary, secondary and tertiary education and delivered professional development courses for pre-and in-service teachers. She also teaches on doctoral programmes. Her research focuses mainly on early language learning and teacher education. Her research interests include: intercultural and cultural education, multicultural schooling, creativity in early language learning, integrated pedagogic approaches in foreign and community languages, cross-curricular teaching, inclusive education and social justice. She has conducted national, regional and local-level research studies and has published in professional and academic journals and books. Her most recent edited volume is Debates and Modern Languages Education (Driscoll, Macaro and Swarbrick, Routledge, 2014). Daniela Elsner is Full Professor of Teaching English as a foreign language (TEFL) Pedagogy at Goethe University Frankfurt Main (Germany), where she is also one of the directors of the Centre of Teacher Education and Educational Research. Her research focuses on early foreign language education, bi- and plurilingualism, multilingual CALL, and language teacher xi

xii

Early L anguage Lear ning

education. In 2014, she was awarded the national Ars Legendi Prize for Teaching Excellence in Higher Education. She is co-author of ‘Sally’ (Oldenbourg Publishers), Germany’s most popular textbook for the English language classroom in primary schools. Janet Enever is Professor of Language Teaching and Learning at Umeå University, Sweden and Visiting Professor at the University of Reading, UK. Her research interests focus on the fields of early foreign language learning, language globalisation and language policy. She holds a doctoral degree from Bristol University, UK in Primary Foreign Languages Policy and has previously worked at universities in London, Krakow and Budapest. Recent publications include: ELLiE. Early Language Learning in Europe (2011) (ed.) and ‘Primary ELT: Issues and trends’, in Routledge Handbook of English Language Teaching, (2016, G. Hall, ed.). She is coordinator for the AILA Research Network in Early Language Learning (2014–2017). María del Pilar García Mayo is Full Professor of English Language and Linguistics at the University of the Basque Country. She has two main lines of research: the acquisition of English morphosyntax from a generative perspective and EFL young and adult learner interaction. She has published numerous articles and chapters and (co-)edited several special issues and volumes. She is the main researcher of the Language and Speech Research Group (www.laslab.org), noted by the Basque Government for excellence in research in the field and the director of the MA programme Language Acquisition in Multilingual Settings Yuko Goto Butler is Associate Professor of Educational Linguistics at the Graduate School of Education at the University of Pennsylvania. She is also the director of Teaching English to Speakers of Other Languages (TESOL) programme at Penn. Her research interests are primarily focused on the improvement of second/foreign language education among young learners in the USA and Asia in response to the diverse needs of an increasingly globalising world. Her work has also focused on identifying effective ESL/EFL teaching and learning strategies and assessment methods that take into account the relevant linguistic and cultural contexts in which instruction takes place. Ilonca Hardy is Professor of Elementary Education, Faculty of Education, Goethe-University of Frankfurt, Germany. She obtained her PhD in Educational Psychology (University of Iowa, USA) in 1998; was a postdoctoral research assistant at the Max Planck Institute for Human Development, Berlin (1998–2007); and at the Habilitation in Education at the Free University Berlin in 2007. Her research focuses on early childhood education, instructional processes in science and language teaching, and children with German as a second language.

Contr ibutors

xiii

Heather Hilton moved to France, after working as a French teacher and completing a doctorate in the USA (Emory University), spending 25 years at the Université de Savoie, before moving on to a research position in Paris and is now Professor of Language Acquisition and Teaching at the Université de Lyon. Her research in foreign language teaching methodology and language acquisition is strongly grounded in sociocognitive theories of learning and in psycholinguistic theories of language processing and use. Ainara Imaz Agirre obtained her PhD in 2015 at the University of the Basque Country (UPV/EHU) and currently teaches in Primary and Infant Teaching Training Degrees at University of Mondragon (Mondragon Unibertsitatea). Her main interests include second and third language acquisition from different perspectives (generative and interactionist frameworks). Changsheng Jiang received his PhD at the Chinese University of Hong Kong and has been working as Associate Professor in the School of Foreign Languages in Zhongnan University of Economics and Law, China. His research areas include L2 motivation, L2 learner identity and World Englishes. Xiaohua Liang received her PhD at Hong Kong University and is now Professor of Applied English Linguistics at Zhongnan University of Economics and Law, working as the Associate Dean of Teaching and Learning. Fascinated by theories, research methods and practices, she is enthusiastic about both teaching and learning in applied linguistics. With her experience of 32 years of teaching and researching, she has been involved in the complete Chinese education system at all levels, and is recognised by her publications and presentations at international conferences such as AILA, AAAL, AERA, ALA, Asia TEFL and HAAL. Eva Lindgren is Associate Professor of Language Teaching and Learning at the Department of Language Studies at Umeå University, Sweden. Her research interests include multilingualism, literacy and young learners. She is the director of literacy research at Umeå University (LITUM) and takes a great interest in the development of research in close collaboration with schools, for example, in indigenous Sami communities. Another line of research focuses on the way society values languages and she uses large databases in order to investigate how knowledge of different languages may, or may not, impact on future life chances. Magdalena López is a Professor in the Sociocultural Studies Department at ITESO University, Guadalajara, and the coordinator of Sciences of Communication Undergraduate Program. She holds a PhD in Information and Communication Society. She has developed dozens of research and intervention projects related to the learning and communication processes from the

xiv Early L anguage Lear ning

sociocultural approach; her research focuses on the impacts of the use of telematics technologies in the ways of knowing, being, interacting and constructing meaning. Since 2010, she has collaborated in several projects about the implementation process of the National Program of English in Basic Education in Mexico. Lucilla Lopriore is an Associate Professor of English Linguistics, Roma Tre University. She has an MA in TEFL Reading University and a PhD in Italian as a Foreign Language, Siena for Foreigners University. She was TESOL Italy President (1996–98), TESOL International Board of Directors (2001–2004) and TESOL International Research Committee (2013–2017). She was the Italian national coordinator of the Early Language Learning in Europe (ELLiE) longitudinal research study (2006–2010). As a teacher-educator and course book writer, her fields of interest are: early language learning, English as a lingua franca (ELF), language teacher education, assessment and evaluation, content and language integrated learning (CLIL). She has published extensively in the field of teacher education, early language learning, ELF, CLIL and assessment. Jelena Mihaljević Djigunović retired from the University of Zagreb in 2014, where she held the post of Professor of Second Language Acquisition and TEFL. Her main research interests centre on second language acquisition, teaching modern languages to young learners, role of affective factors (attitudes, motivation, language anxiety and self-concept) in language learning and foreign language teacher education. She has been involved in a number of large scale national and international research projects, the latest one being the Early Language Learning in Europe (ELLiE) study. Her publications include two research books on affective learner factors and over 100 papers. She has co-edited several research volumes. Victoria Murphy is Professor of Applied Linguistics at the Department of Education, University of Oxford. She is the research group convener of the Applied Linguistics and the REAL (Research in English as an Additional Language) research groups. She is also the course director of the MSc in Applied Linguistics/Second Language Acquisition. Victoria’s area of research lies mainly within the realm of child L2/foreign language (FL) learning, vocabulary and literacy development. She has published in a wide range of applied linguistics journals and is the author of Second Language Learning in the Early School Years: Trends and Contexts published by Oxford University Press in 2014, as well as the lead editor of Early Childhood Education in English for Speakers of Other Languages published by the British Council in 2016. Karmen Pižorn is an Associate Professor of English in Education, University of Ljubljana, Faculty of Education, Slovenia. She obtained her PhD in English

Contr ibutors

xv

Language Teaching Methodology from the University of Ljubljana, Faculty of Arts in 2003. She has been a primary teacher educator and national testing team leader of English for primary and lower-secondary students for 15 years, co-leader of the National Project of Implementing Foreign Languages to Year 2 students of primary school in Slovenia (2009–2012), advisory board member to the EU Surveylang project and Slovenian coordinator of the EU DysTEFL2 project (2014–2016). She has published articles and books in the field of early language learning, language assessment, language education and language policies. Her research interests include: language assessment, early language learning, CLIL, teacher education, ICT in language learning and language policies and politics. Raphaela Porsch is a Senior Researcher in Education/Empirical School and Classroom Research, University of Muenster, Institute of Educational Science, Germany. Her PhD was about assessment of EFL writing competencies and was obtained at the Humboldt University of Berlin, Institute of Educational Science (2010). Her research interests include: teacher education, teacher qualification/teaching out-of-field, emotions and (early) foreign language teaching. Alison Porter is Lecturer in Applied Linguistics at the University of Southampton. She obtained her PhD in Modern Languages in 2014. She has been a primary FL (French) teacher for 10 years and EFL/ESL teacher for seven years. She teaches on both MA Applied Linguistics and undergraduate degree programmes. Her research interests include: early classroom language learning, foreign and second language reading, language learning assessment and the role of individual differences in language learning. Peter Sayer is an Associate Professor in the Department of teaching and learning at Ohio State University. He holds a PhD in Language and Literacy from Arizona State University. His work focuses on language education and teacher preparation and sociolinguistics. He is the author of Tensions and Ambiguities in English Language Teaching (Routledge, 2012). Since its inception 2009, he has collaborated with the Ministry of Education to support the implementation of the English programme for public primary schools in Mexico. Agatha J. van Ginkel has, for most of her career, worked in the areas of literacy and languages, and aspects of first, second and foreign language acquisition, applying this knowledge in formal and non-formal education in Africa and Asia. Her research interest includes factors that influence transfer of reading across languages and scripts and measuring of reading performance in different languages providing context-appropriate solutions to a variety of reading problems in linguistic-specific contexts. She received a PhD in Applied

xvi

Early L anguage Lear ning

Linguistics from Leeds Metropolitan University (UK), an MA in TEFL from University of Reading (UK), a BA in Teaching English from Windesheim University (NL) and a BA and MA in Special Educational Needs (SEN) from Fontys University of Applied Science and Free University of Amsterdam (the Netherlands). Eva Wilden is Professor for EFL Education and teacher educator at Vechta University/Germany. Her PhD is in Foreign Language Learning Research, Kassel University/Germany. She holds a Postgraduate Certificate in Education (history through the medium of German) from Nottingham University and has worked as a secondary school teacher in England and Germany (EFL, history and CLIL history). Her empirical research currently focuses on assessing receptive EFL skills, primary and secondary EFL education, inclusive foreign language education, culture and language learning, media and multi-literacy education and CLIL. Qun Xie is Professor of English, Zhongnan University of Economics and Law, China. She obtained her PhD in 2004 from the Department of English, Chinese University of Hong Kong. Her research interests include modern drama studies, English education and comparative cultural studies. She has published two books and over 20 essays on modern drama studies and English teaching, and has translated six English novels and books of literary criticism into Chinese. Yuan Yuan is an Associate Professor at the Zhongnan University of Economics and Law, China. Her doctoral dissertation looked at the relationship between second language users and L2 user identity by employing narrative inquiry and conversation analytic approaches. Her main research interests include discourse analysis, intercultural communication and identity. She is the author of Cultures of Learning: An Evolving Concept and an Expanding Field; Stereotypes and Talk-in-Interaction: The Strategies Used in Identity Negotiation and Identity (Re)formation in a Study-Abroad Context: The Case of a Chinese Learner of Japanese. Jie Zhang received her PhD from Macquarie University in 2011. Her PhD research is an investigation of language ideologies, language policies and language practices in the context of the 2008 Beijing Olympic Games. She is now Associate Professor and Associate Head of the Department of English Language and Literature, Zhongnan University of Economics and Law, China. Her research areas include language policy and planning, language learning and identity, and second language acquisition and intercultural communication. Her book, English Desire in the Olympic Spotlight, will be published by De Gruyter.

1

Introduction: Mixed Methods in Early Language Learning Research – Examining Complexity Eva Lindgren and Janet Enever

In this volume we examine the full complexity of language learning, and in particular the foreign, or second, language learning of young learners in instructed contexts. The aim of the volume is twofold: first and foremost we bring together and present recent research results from the early language learning classroom context. Second, the volume reviews the choice of research methodologies for early language learning research in schools with a particular focus on mixed methods, and proposes that in the multidisciplinary context of early language learning, this paradigm allows for a more comprehensive understanding of the evidence than other approaches might provide. Before we describe the volume and its underpinnings, we would like to give a brief note on terminology. The term foreign language (FL) will be used for languages other than the mother tongue(s) a person learns in an instructed context where the national language used outside the formal context is not the FL. The term second language (L2) will refer to a language learnt in, or outside, a formal context within a society where the L2 is used as the national, or most widely used, language. However, a word of caution may be necessary. In the case of, for example, Sweden or the Netherlands, English does not have the status of national language, but it is used to such a high degree in society that children learn it in much the same way as an L2 rather than an FL. Language learning has traditionally been studied from a variety of theoretical perspectives drawing on a range of research methodologies. Recently, Atkinson (2011) called for a broader, holistic approach to the understanding of the learning of other languages than the mother tongue(s) in which the full dynamics and complexity are acknowledged by the use of cognitive, sociocultural and dynamic systems theories. In terms of methods, the 1

2

Early L anguage Lear ning

cognitive approach has typically included testing and quantitative methods, while the sociocultural approaches have preferred qualitative approaches such as ethnography and case studies. In a thorough review and discussion about complex/dynamic systems in second language research LarsenFreeman (2011: 62–63) describes how a variety of methodologies can and have been used to study complex systems: quantitative, computer-based modelling, qualitative modelling with or without computers, dynamic descriptions using an ethnographic approach, formative experiments and design experiments. Mixed methods research (MMR) appeared in the late 1980s as a pragmatic response to debates about the acceptability of combining qualitative and quantitative approaches to research, particularly in the social sciences. Underpinning this debate was the contested area of research methods, where quantitative methods may be identified as connected with a positivist (or post-positivist) worldview and qualitative methods with a constructivist worldview (Creswell, 2003; see Riazi & Candlin, 2014 for a critical review and Gheitasi & Lindgren, 2015 for an overview). MMR is based on the philosophical assumption that both quantitative and qualitative methods are necessary throughout the research process (Creswell & Plano Clark, 2006) and that they allow researchers to explore a variety of research questions within the same study (Teddlie & Tashakori, 2003). By employing MMR, it is argued that validity can be increased by triangulation, complementarity or by the use of results from one study in the development of another (Greene et al., 1989). Greene et al. (1989: 259) further explain that mixed methods can ‘extend the breadth and range of the study’, as well as increase the depth and scope of the outcomes. A challenge in mixed methods research ‘concerns the level of theorising and conceptualising of the research problem so that different layers and various dimensions of the issue in question can be studied and integrated by means of an MMR [mixed methods research] framework’ (Riazi & Candlin, 2014: 161). However, MMR is a contested area and Fetters (2016: 3) draws our attention to the existence of many previous research studies, which have included both quantitative and qualitative perspectives, dating back to the earliest part of the 20th century at least. He cites particularly the disciplines of anthropology, social psychology and sociology as reflecting an approach that includes a mix of methods in a single research programme. Similarly, Maxwell (2013) reports that such traditions date back some centuries in the fields of natural, social and health sciences, reminding us that barriers as to what constituted legitimate research methodologies did not exist in Ancient Greek times where astronomy ‘used both observational description and quantitative measurement’ (Maxwell, 2013: 14). Regardless of the contested philosophical and historical underpinnings of MMR, the use of more than one theoretical and methodological approach seems useful for early language learning research. The complexity of the

Introduc t ion

3

language classroom, including how children learn foreign languages, what may influence their learning and what role the teacher and the context play, all lend themselves well to the mix of methods. For example, in order to understand more about children’s reading in a foreign language, students may be asked in a questionnaire how they feel about reading in a foreign language; in parallel, data on their reading skills is collected by means of a test and the teacher is interviewed about teaching strategies. When analysing the data, quantitative methods can be used for test and questionnaire data, which may then be discussed in the light of the teachers’ response. In a sequential study, children’s interviews can serve as input to a larger survey about, for example, motivation for learning foreign languages.

This Volume In this volume we report on a variety of studies, which represent a mix of instructed contexts, countries and methods. In the final chapter we return to the question of mixed methods and discuss its potential for early language learning research. The contexts for early language learning in the studies include English as an additional language (EAL), English as a second or foreign language (L2/FL), French as a modern foreign language (MFL), controversies on choices of medium of instruction (MOI) and classrooms exploring the potential for content and language integrated learning (CLIL). Each of these themes reflects an area of current growth in the broad field of early language learning worldwide today. The contexts from which the research studies are drawn include: China, Croatia, France, Germany, Italy, Mexico, Slovenia, Spain, Sweden, UK and the African countries of Kenya, South Africa, Cameroon and Ethiopia. Across the continents of Africa, Europe, Asia and Latin America (as reflected in this choice of research papers) growth in early language learning has been substantial since the beginning of the 21st century. Research in each of these regions is now beginning to build, but further knowledge is needed to establish understanding and expertise in the very varied contexts emerging today. The first part of the book includes three overviews of recent research findings in quite different contexts. This part is introduced by Agatha van Ginkel who discusses research evidence from a range of multilingual contexts in Africa to consider the complex question of decisions on the medium of instruction in the first few years of schooling. This is followed by a chapter focusing on the importance of including cultural, social and intercultural learning in the primary languages curriculum. Patricia Driscoll critiques recent research in the field to propose that a distinctive cultural strand needs to be placed at the heart of the primary curriculum. The first part is concluded with a wide-ranging discussion of research in the field of EAL in UK primary schools. Here, Victoria Murphy explores research on vocabulary and

4

Early L anguage Lear ning

literacy development placing particular importance on the development of rich, high quality lexical representations, which can promote more advanced and complex reading and writing skills in English. The second part of the book includes five empirical studies of early language learning, all of which adopt a mixed method approach to data collection and analysis. The introductory chapter from Alison Porter reports on a classroom study of oracy and literacy development in French in the UK with findings revealing the benefits of providing input through a number of modalities. This is followed by an innovative approach to exploring questions of motivation for primary learners of English in China where data were collected using the tools of elicited metaphor, combined with questionnaire responses. A team of researchers based in Wuhan, China worked on this study, originally initiated by Jin Lixian as a visiting researcher based in Leicester, UK. This is followed by a similarly innovative study, using a quite different methodology, to report on the effects of using multilingual virtual storybooks in primary classrooms in Germany. Here, Judith BuendgensKosten, Ilonca Hardy and Daniela Elsner have focused on revealing evidence of vocabulary development and story comprehension in their study. The fourth chapter in this part comes from Heather Hilton, providing a study of early learners of English in France. The study compares outcomes from a class of beginners starting in grade 1 with another class which started in grade 3. This close empirical study provides a fine-grained analysis of both individual case studies and a series of group studies. As a final chapter to this part, Karmen Pižorn tackles the thorny debate over CLIL in primary classrooms. Her critical analysis of research and current claims is well-illustrated by a small-scale study of a CLIL classroom compared with a parallel class where more conventional approaches to EFL teaching and learning are adopted. The third part includes four chapters that adopt a longitudinal perspective in answering questions related to early language learning. Themes included are motivation and self-concept, broad language outcomes and the development of listening, reading and writing skills, all of which lend themselves well to tracking over time as a mechanism for recording their dynamic development. This part opens with Yuko Goto Butler’s study of motivation in primary school learners of English in China. Her study triangulates a number of qualitative and quantitative measures to identify factors influencing motivation over time. The three further studies in this part all take as their starting point data initially collected in the ELLiE study (Enever, 2011). First, in Lucilla Lopriore’s chapter, the longitudinal focus is on the development of listening, reading and writing skills over time in Italy. Outcomes are reviewed in relation to both teacher assessment and learner self-assessment. Next, Eva Lindgren and Janet Enever report on a continuation of the study with a sub-sample of the original cohort based in Sweden. In their chapter they discuss final outcomes achieved after six years of English classes, providing rich, thick descriptions of learner achievements. Finally, Jelena

Introduc t ion

5

Mihaljević Djigunović discusses longitudinal evidence from 81 Croatian learners of English to review the instability of young learners’ attitudes, motivation and L2 self-concept over time. The implications of her findings are considered from a theoretical and practical perspective. The final part of the book includes three chapters offering an evaluative perspective on some of the major themes for early language learning in instructed contexts today. This part is introduced with a chapter from María del Pilar García Mayo and Ainara Imaz Agirre who discuss interactive tasks in the primary FL classroom. For many, interactive tasks are seen as essential building blocks for progress in language learning, thus our understanding of task design and effectiveness in the primary years is vital. In their study the researchers compared two learning environments, firstly a CLIL-type environment and secondly a mainstream EFL environment. Their report on interaction for the 27 dyads in their study indicates a more frequent use of conversational adjustments as one outcome from instruction in a CLIL-type environment, while age appears to be of limited significance as a variable. This use was not constant in time for any of the FL groups, however. In the second chapter in this part, the context shifts to Mexico, where Peter Sayer, Ruth Ban and Magdalena López de Anda report on the design and implementation of a large-scale impact study for the evaluation of primary English. This chapter reports on the pilot phase, implemented in one Mexican state, charting the evidence of both linguistic and non-linguistic gains made across the first six grades of schooling. The final chapter in this part, authored by Raphaela Porsch and Eva Wilden, presents a project for the development of a final test framework for use at the beginning of secondary school in Germany. The proposed C-test aims at measuring general language ability with a view to providing secondary school (year 5) EFL teachers with a tool for diagnosing their new pupils EFL learning. The authors discuss the first stages of design and development for this complex evaluative tool. In the concluding chapter to this volume we reflect on the potential contribution that MMR may offer the field of early language learning (ELL) research. Here, we discuss the variety of contributions presented, proposing that factors such as scale and multidisciplinarity may be more likely to promote the choice of an MMR design. Nonetheless, we fully acknowledge the continuing contribution that a singular research design can make to extend the knowledge base of research in ELL and call for expansion of investigations in all directions.

References Atkinson, D. (2011) Introduction: Cognitivism and second language acquisition. In D. Atkinson (ed.) Alternative Approaches to Second Language Acquisition (pp. 1–23). London: Routledge. Creswell, J. (2003) Research Design: Qualitative, Quantitative, and Mixed Methods Approaches (2nd edn). Thousand Oaks, CA: Sage.

6

Early L anguage Lear ning

Creswell, J.W. and Plano Clark, V.L. (2006) Designing and Conducting Mixed Methods Research. Thousand Oaks, CA: Sage Publications, Inc. Enever, J. (ed.) (2011) ELLiE. Early Language Learning in Europe. London: British Council. See https://www.teachingenglish.org.uk/sites/teacheng/files/B309%20ELLiE%20 Book%202011%20FINAL.pdf (accessed 16 March 2017). Fetters, M.D. (2016) ‘Haven’t we always been doing mixed methods research?’ Lessons learned from the development of the horseless carriage. Journal of Mixed Methods Research 10 (1), 3–11. Gheitasi, P. and Lindgren, E. (2015) Broadening the understanding of the language classroom: Mixed methods. In E. Lindgren and J. Enever (eds) Språkdidaktik: Researching Language Teaching and Learning (pp. 21–30). Umeå: Umeå Universitet. Greene, J.C., Caracelli, V.J. and Graham, W.F. (1989) Toward a conceptual framework for mixed method evaluation designs. Educational Evaluation and Policy Analysis 11 (3), 255–274. Larsen-Freeman, D. (2011) A complexity theory approach to second language development/acquisition. In D. Atkinson (ed.) Alternative Approaches to Second Language Acquisition (pp. 48–72). London: Routledge. Maxwell, J.A. (2016) Expanding the history and range of mixed methods research. Journal of Mixed Methods Research 10 (1), 12–27. Riazi, M.A. and Candlin, C.N. (2014) Mixed-methods research in language teaching and learning: Opportunities, issues and challenges. Language Teaching 47, 135–173. Teddlie, C. and Tashakkori, A. (2003) Major issues and controversies in the use of mixed methods in the social and behavioral sciences. In A. Tashakkori and C. Teddlie (eds) Handbook of Mixed Methods in Social and Behavioral Research (pp. 3–50). Thousand Oaks, CA: Sage.

Part 1 Overviews of Research Findings

2

Early Language Learning in Complex Linguistic Settings: Insights from Africa Agatha J. van Ginkel

Introduction Most of the people in the world speak or have to speak more than one language. Often people use different languages to communicate with different people in their neighbourhood, at the market or sometimes even within the family when in-laws are from other language communities. Companies often employ multilinguals as it will help them to communicate to a wide audience in the languages that they best understand. Multilingualism is in many ways more common and advantageous than monolingualism. Having said that, multilingualism also provides some challenges, in particular when a country has 30, 50, 80 or even more than 100 languages, as is often the case in Africa, Latin America and Asia. The challenge for these countries is to find a way for all different language communities to communicate with each other. To do so, one language is chosen that each person is expected to learn. This often happens at school. The question is also which language or languages are used at school? How do children learn a new language best? How do children become multilingual? This is a serious and challenging issue in many multilingual developing countries. This chapter seeks to gain more insight in language education in multilingual countries in developing contexts. It will do so by first examining different approaches to education in multilingual settings and reviewing them based on what is already known. Then, it will look at data from different multilingual countries to better understand what evidence comes from the field about different multilingual education models in developing countries. After that, it will briefly look at data that have emerged from situations where more than one mother tongue is spoken in a classroom. Next, it will review the most prominent context factors that influence the pace of language learning. Then, all information will be pulled together to look at what 9

10

Par t 1: Over v iews of Research Findings

picture is emerging from these findings. Finally, the chapter will briefly discuss a common economic argument that prevents implementing the best approach for becoming multilingual in developing countries.

Models of Language Education As stated in the introduction, there are many countries in Africa, Latin America and Asia that are multilingual. Different models of bi/multilingual education are used in Africa, Asia and Latin America. These will be discussed first. While it hardly seems possible to differentiate models between all the programmes that are going on, in general one can categorise the different bilingual education programmes into one of the following models: (1) Submersion models in which only the new language is used. The students tend to be minority language speakers with a low status mother tongue (Alidou et al., 2006; Baker, 2006; Skutnabb-Kangas, 1988). (2) Immersion models in which initially only the new language is used and later on both languages are developed. A well-known immersion programme is the French programme in Canada (Baker, 2006; Thomas & Collier, 1997). It is important to note that in immersion programmes, the students tend to speak a majority language. The parents deliberately choose these programmes, the programmes are well resourced and the input has been tailor made for the students (Alidou et al., 2006; Ellis, 1994). If this is not the case, then it is often a submersion model. (3) Early exit models where the first language of the child is used as medium of instruction during the first 1–4 years of education, after that the new language is used as a medium of instruction (Alidou et al., 2006; Baker, 2006). (4) Late exit models where the first language of the child is used as medium of instruction during the first 6–8 years of education, after that the new language is used as a medium of instruction (Alidou et al., 2006; Baker, 2006). (5) Additive models or dual language models where both the first language and the new language are used throughout education (Alidou et al., 2006; Baker, 2006). These five models are helpful references to understand some of the research that has emerged from different counties. The research referred to in this chapter mostly comes from Africa, as the situation in Africa is most pressing. In Africa, about 2500 language are spoken, while about only 176 languages are used as language in education in Africa (Gadelii, 2004; Ouane & Glanz, 2010).

Early L anguage Lear ning in Comple x Linguist ic Set t ings

11

Evidence for the Success of Different Models In the last few years, there has been a new body of research emerging from multilingual countries in developing contexts that are helpful in understanding what is useful for children in multilingual contexts. Although, not all the data lead to straightforward conclusions, it is helpful to review the research. The data give insights of the degrees of success of submersion, early exit and late exit models.

Mismatch between classroom practice and policy One of the first things to notice when reviewing research from different countries is that while a country might have a language in education policy stating one of the above models, the reality in the classroom might be that a completely different model is used than that stated in the policy. Some research from Kenya provides insights into this. The language policy in Kenya mandates that the mother tongue of the children is used during the first three years of primary school and that English and Kiswahili are learnt as a subject (early exit model). However, the reality in most classrooms is that English is used for more than 60% of the time (Nyaga & Anthonissen, 2012; Piper & Zuilkowski, 2015). Thus, the reality in the classroom is more of a submersion model. According to Piper and Zuilkowski (2015), the reason for using English is that English is highly valued and believed to be the way out of poverty. People also believe that children learn English by getting educated in English. People believe in the submersion model. The data from the Kenyan classrooms show that these beliefs do not hold up very well (Nyaga & Anthonissen, 2012; Piper & Zuilkowski, 2015). Even though the children hear a lot of English in the classroom and much less Kiswahili or mother tongue, the children’s mastery of English is not sufficient, even inadequate for them to understand what they are reading. The reading tests showed that children were understanding very little of what they were reading in English, while in the mother tongue their comprehension scores were much higher. This showed that the mismatch between policy and the implementation in the classroom is not helping the children at all, ‘the Kenyan education system is trying to “buy” strong English language reading outcomes through an emphasis on English-medium instruction; but what is actually being produced are children with limited fluency in either oral or written English and minimal reading skills in their own languages as well’ (Piper et al., 2016: 148). Another research study from Kenya showed that if the language of instruction is known to the children, comprehension scores for reading can be comparable with those in Western countries (Graham & Van Ginkel, 2014). The study compared the first language reading ability of grade 2 children from two European countries with that of children from two

12

Par t 1: Over v iews of Research Findings

Percent within each group

80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 0

20

Dutch

40 60 80 Score in comprehension Eng

Eng/Swa/Pok

100

Sa+Ki+En

Figure 2.1 Comprehension scores Dutch, English, Pokomo and Sabaot (Graham & Van Ginkel, 2014)

language groups in Kenya. The study included 112 children from the Netherlands, 27 children from the UK, 177 children from the Pokomo community in Kenya and 105 children from the Sabaot community in Kenya. Figure 2.1 shows the average comprehension scores for the four languages. The average comprehension scores were for English 90%, for Sabaot 82%, for Dutch 81% and for Pokomo 68%. While the Pokomo and Sabaot comprehension showed a greater spread than the two European languages, in all languages, more than 50% of the readers were able to comprehend at least 80% of what they read. There was a wider distribution among the two Kenyan languages, which could be explained by some educational and sociolinguistic reasons, but the results show that reading in a known language means that children can perform well in school. These data from Kenya show that a language policy is not always implemented in the classroom. They also show that changing the medium of instruction to a desired language rather than a language understood by children, does not mean that children learn the desired language well and do well educationally in that desired language. The data also show that if a known language is used in the classroom, learning results in developing countries can be comparable to learning results in developed countries, even though there are many other differences.

Submersion model compared with early exit models In many African countries, an early exit model is the currently stated model in the language policy. Recently, Taylor and Coetzee (2013) analysed the data for English and mathematics standardised tests from the South African Annual National Assessment over a period of five years. Their analyses show that the use of the mother tongue in lower grades leads to

Early L anguage Lear ning in Comple x Linguist ic Set t ings

13

signi ficantly better English proficiency rates in grades 4, 5 and 6 (Taylor & Coetzee, 2013). Similarly, data from India confirm that reading in the mother tongue is having a positive effect on learning to read in English (Berry, 2013). Furthermore, the data from South Africa showed that in the high performing schools, students learn English better than in low performing schools (Taylor & Coetzee, 2013). This is an important observation as it would have implications for deciding the most appropriate grade for introducing a new medium of instruction. It implies that it is very likely that children in high performing schools could probably cognitively benefit from English as a medium of instruction earlier than children in the low performing schools. Therefore, a flexible policy for deciding when a new medium of instruction can be used is necessary. The study also shows that there are other context factors that influence the rate of learning and learning outcomes. The most prominent factors will be looked at later in this chapter. Another set of data comes from Cameroon. In one of the language communities in Cameroon, a pilot project was started to see what the difference would be if children began learning through English as a medium of instruction (submersion model) from the start, or when they would have the first three years education in their home language and learnt English as a subject (early exit model). The Kom Experimental Mother Tongue Education Pilot Project (KEPP) collected data from both kinds of learners. The data from KEPP showed that the students who had their mother tongue as medium of instruction outperformed the students who had English as medium of instruction in all subjects, also in English (Walter & Chuo, 2012; Walter & Trammell, 2010). Figure 2.2 shows the score. L1 means children who had Kom as a medium of instruction; L2 refers to the children who had English as a medium of instruction.

Average score for English

60 50 40 30 20 10 0 1-L1

1-L2

2-L1

2-L2

3-L1

3-L2

Class and Language of Instrucon

Figure 2.2 English scores of the Kom L1 based bilingual programme (Walter & Trammel, 2010)

14

Par t 1: Over v iews of Research Findings

In year 4, the mother tongue students were transitioned to use English as a medium of instruction. The scores of the students dropped considerably, showing that three years education with the mother tongue as medium of instruction was not sufficient to prepare the students for the transition. Interestingly, the students who had been in an English-only class for all four and five years of education, scored lower and seemed even less prepared for the linguistic demands in class 4 and 5 than the KEPP students. Walter and Chuo (2012) attributed the higher scores of the KEPP students to the fact that their foundation in education was laid in their mother tongue. Thus, the data from Cameroon also show that learning through the home language is more beneficial for learning a new language than having the new language directly as a medium of instruction. Furthermore, the data also showed that early transition (after three years of mother tongue education) is still not sufficient for the children to have learnt the new medium of instruction well enough to be able to benefit from it well in class 4 and onwards.

Different models used in one country Ethiopia is an interesting case when it comes to language in education policy. It is one of the few countries in Africa that allows for different multilingual education models. It has a language policy in education that states that the first four years of primary school should be in the home language of the children and they learn Amharic (national language of Ethiopia) and English as a subject. From year 5 onwards, the regions in Ethiopia can choose which language to continue with. They can either continue using the home language of the child or they can use English as a medium of instruction. For different reasons, the 11 regions in Ethiopia have made different choices. Some regions have chosen to use English as a medium of instruction from grade 5 onwards, others from grade 7 onwards. A few regions have chosen to offer a dual language track from grade 7 onwards and use English as a medium of instruction for some subjects, while mother tongue is used for other subjects. Then, there are also a few regions that only use English as a medium of instruction from grade 9 onwards (Heugh et al., 2007). Thus, in Ethiopia an early exit model and a late exit model are stated in policy. Having said that, for several language communities in Ethiopia, mother tongue education is not yet applied as the languages do not have a written form and/or no materials have been developed for use in school. For those children, the medium of instruction at school is mostly Amharic, a submersion model. There are a few data sets that provide some insights on how these situations affect education outcomes. The national tests in grade 8 provide insights of how (without an intervention) the children are performing in these different models of language education. The national grade 8 tests of the different regions show that grade 8 students who had their home language as medium of instruction and English as a subject throughout primary

Early L anguage Lear ning in Comple x Linguist ic Set t ings

15

school, outperformed the students who had English as medium of instruction in all subjects (Heugh et al., 2007). Thus, indications are that using the home language as a medium of instruction throughout primary school gives children the best chance to do well in school. There is also a data set that shows how children are performing at learning English in the first few years of primary school. The data set measured the proficiency of English compared to the minimum required learning competences for the Ethiopian curriculum. Per grade, it shows the percentage of students that have reached minimum required learning competences. The scores were divided into four groups: below basic, basic, proficient and advanced. The results in Figure 2.3 show that most students in grades 1, 2 and 3 score below basic. Furthermore, in year 4 only 19% of students in grade 4 (the sum of proficient and advanced level) have reached the required knowledge and skills of the minimum required learning competences for grade 4. These minimum learning competences are set at the threshold level that would suggest that transition to using English as medium of instruction would be successful. The scores are far below the minimum threshold levels. These data suggest that using English as a medium of instruction in grade 5 (as is happening in some regions) would make it hard for students to benefit from education at all. What can be learned from language research in Ethiopia is that transition to English as a medium of instruction during the second half of primary education (year 5) is not helping the children to do well in their education, in fact it even seems detrimental to their learning results. This corroborated the findings from Cameroon. In the context of Ethiopia, the data suggest that eight years of mother tongue education, a late exit model, is necessary Grade 2

Grade 3

Grade 4

80

Percentage of student

70 60 50 40 30 20 10 0 Below basic

Basic Proficient Performance level categories

Advanced

Figure 2.3 Primary school English proficiency in Ethiopia (American Institute for Research, 2012)

16

Par t 1: Over v iews of Research Findings

to have some kind of success in learning through English as a medium of instruction after that.

Different language in education policies throughout the years The findings from Ethiopia, suggesting that a late exit models provides better learning results, are further corroborated by data from South Africa. Throughout the years, the language in education policy in South Africa has changed. Data from the end of school exams from South Africa over a few decades (from 1955–1998) provide insights on when children were performing better at school and when the results were going down. The data in Table 2.1 show the results of the final school leaving exams in class 8. The language of the exams is English. The results show some interesting developments. The pass rate is quite high in 1979 (87%), but decreased in 1992 and has continued to decline from then onwards. This seems to reflect the change in language policy. The children who had eight years of primary school instruction in their mother tongue scored well in the school leaving tests. However when in the early 1990s the policy changed to include only four years of mother tongue provision, then the scores on the school-leaving test also declined. These data seem to suggest that the early switch to English does not work in the majority of South African schools (Heugh, 2012). It could of course also be that more changes took place and the instruction was too poor for early transition. What seems to be clear is that ‘There is [a] gap between students’ English proficiency and the linguistic demands of learning through the medium of English’ (Probyn, 2006: 393).

Table 2.1 Pass rates at end of school exams in South Africa Year

African language speaking students

Pass rate (%)

1955 1976 1979 1980 1982 1992 1994 1997 1998

595 9595 14,574 29,973 70,241 342,038 392,434

43.5 83.7 73.5 53.2 48.4 44 49

Source: Adapted from Heugh (2012).

Overall total number of candidate, plus pass rate (%)

85,276 (87%) 109,897 (75%) 139,488 (69%) 448,491 (56%) 495,408 (58%) 559,233 (47.4%) 552,862 (49%)

Early L anguage Lear ning in Comple x Linguist ic Set t ings

17

Multiple Mother Tongues in a Class While the above research seems to show that it is good for children to start education in a language they understand and then have time to learn another language later, a burning issue arises. In several research reports, it is mentioned that in classrooms sometimes more than one mother tongue is spoken by the children. For example, through research in Ghana (Leherr, 2009), it came to light that nearly four out of 10 students did not understand the Ghanaian language used in that primary school. The question is then, what language would be a good medium of instruction for all these children? The KEPP data collected in Cameroon provides some interesting information. Besides the Kom speaking children, there was another group of children in both the mother tongue and the English language classes that spoke a different African language, Fulfulde. The data from the Fulfulde children were collected separately to see how they were doing in a ‘mother tongue class’ where the medium of instruction was not their mother tongue. Would they perform better or worse than if they were attending an English medium class? The data showed that the Fulfulde children did better in the Kom medium class than in the English medium class. The Fulfulde speaking students in the Kom medium schools outperformed the Fulfulde speaking students in the English medium school by a factor of 72.4% in grade 2 and by 127% in grade 3 (Walter, 2015). This seems to suggest that learning in another African language is easier for the children than to learn in a linguistically quite different language such as English. A similar report comes from Nepal, where children learned to read in their mother tongue while learning Nepali and English as a subject. After two years of education, the children can read as well in their mother tongue as in Nepali (MLE-Team, 2013). However, reading in English is not going very well, while they are also learning it as a subject (M. Wong, 18th June 2015, Personal correspondence). Still, much more research needs to be done in this area, but the initial data seem to suggest that if a language is ‘closer to home’ it is easier for the children to acquire it than when it is ‘farther from home’, as is the case with former colonial languages such as English. It seems to suggest that different trajectories of learning take place depending on the ‘closeness’ of the language. Furthermore, often no guidelines are provided for teachers in linguistically diverse classrooms and the teachers use their own interpretations and teaching methodologies for these situations (Nyaga & Anthonissen, 2012). More research is necessary to be able to provide the teachers with guidelines on how to help children to learn best in these classrooms.

Other Contextual Factors Taylor and Coetzee’s (2013) research has already indicated that other factors play a role in language acquisition in education situations. Also,

18

Par t 1: Over v iews of Research Findings

research from Ghana (Leherr, 2009) showed that additional factors also play a role. While the children in Kenya and Cameroon did well in reading in their mother tongue, in Ghana they did not perform as well as expected. Leherr (2009), in her report, indicates that the teacher seems to be an important factor in learning outcomes. Students of effective teachers score better than students of less effective teachers (Leherr, 2009). This also reflects what Taylor and Coetzee (2013) noticed with the high and low performing schools in South Africa. Even though the language factor plays a pivotal role in education, there are some other factors in different contexts that may result in a new language being learnt at a different pace. It is therefore important to consider contextual factors that may play a significant role in language education. It is beyond the scope of this chapter to discuss them in detail, so only the most important factors that have been identified by research are listed here: (1) The language level of the teachers. A minimum level is required for language teaching for good learning to take place (Enever et al., 2009; Institute of International Education, 2012; Pflepsen & Backman, 2012). (2) The use of the second/foreign language outside of the school setting. When the language is used outside the classroom (second language) it is acquired at a faster pace than when it is only used in the classroom (foreign language) (Baker, 2006; Ellis, 1994). (3) The age at which students start to learn the new language. Students that start at the age of nine or 10 have cognitive advantages over the younger students, while the younger students learn language intuitively. What is important is to make use of the advantages of the different ages (Johnstone, 2002; Mihaljevic Djigunlvic, 2014; Muñoz, 2008, 2014; Nikolov, 2009). Students that are nine years or older have the potential to learn languages faster in a school setting than students who start at the age of five or six. The ‘older’ students do better at learning a new language in an instructional setting like school, because of their cognitive maturity. Younger students have the advantage that their learning is more intuitive, but in a school setting they might learn as much in four years as the nine or 10 year old would learn in two years. (4) Clearly stated language competences and vocabulary that matches the competences. When a syllabus has clearly stated competences and vocabulary listed that match the competences, then the pace of learning improves. A lack of clear language competences and a mismatch of needed vocabulary mean that it takes longer to acquire the threshold level (Benson et al., 2010). (5) The status of the new language and the first language. When a language has low status or is perceived to be less valuable, it takes longer to reach a threshold level (Nikolov & Józsa, 2006).

Early L anguage Lear ning in Comple x Linguist ic Set t ings

19

(6) The materials available for language learning. When students have materials available for language learning, their pace of language learning can be faster (American Institute for Research, 2012; Enever et al., 2009; Enever, 2011). (7) The number of students in the class. Fewer students in a classroom give a faster language acquisition pace (Enever, 2011). (8) The hours of language learning. The more hours spent on language learning (not just exposure) the faster the pace of learning (Enever et al., 2008). (9) Scripts of the two languages. When the two languages have different scripts, it takes longer to learn the second language (Asfaha et al., 2009; Piper & Van Ginkel, 2016). (10) The language distance (i.e. how linguistically similar or different they are). When the languages are related, the language acquisition pace is faster than when they are not related (Chiswick & Miller, 2004). All these factors play a role in determining the appropriate pacing and sequencing for a given context (Pflepsen & Backman, 2012; Van Ginkel, 2014).

The Emerging Picture The information in this chapter has helped to gain some insights on language education in developing contexts. One thing that is clear, is that the situation is diverse and one solution will not fit all contexts. While much more research is necessary, there is a certain picture emerging: in a developing country, the longer children have a chance to learn a new language that will be used as a medium of instruction, the more chance they have to perform to their ability in education. It is also emerging that submersion and early exit models are not successful. For most children in these programmes, it is sink or swim. It is important to note the difference between submersion and immersion models. Immersion only tends to be effective when children speak a majority language. In most situations in Africa, children will speak a minority language, which is a significant factor and that makes an intended immersion programme, a submersion programme. Moreover, because in submersion and early exit models there is hardly any space for the minority children’s language or culture, they tend to have a negative effect on children (Caffery et al., 2014; Skutnabb-Kangas, 1988). For example, because the language and culture of the children is hardly given any space it harms their self-esteem, relationships, roots and sometimes race (Baker, 2006). As for the multilingual classroom, the data also seem to indicate that length depends on how ‘close’ or ‘far’ away the language is that will be used as a medium of instruction. The research from Ghana, Cameroon and Kenya show that just submerging the children in the new language is not helping

20

Par t 1: Over v iews of Research Findings

them to learn the language faster or better. The data from Cameroon showed that children who learn in their mother tongue even learn English better than those who are submerged in English. What is also evident is that much more research is necessary to understand what would be best for children in a linguistically diverse classroom. Furthermore, the success of language education depends very much on the context in which it is implemented. Understanding the context is therefore pivotal and will provide guidance for which models are most effective in which settings. It is quite possible that a multilingual country needs to have a flexible language policy that allows the home language to be used as sole medium of instruction until the new language has been learnt well enough, so that children can cognitively benefit from it when it is (also) used as a medium of instruction.

Economic Considerations for Implementing Multilingual Education Having said all the above, the implementation of a late exit programme is often not considered because it seems rather expensive to develop school materials in 80+ languages, train teachers in 80+ languages. This is a particularly pressing issue in developing countries and is important to address. This economic argument seems to be valid, but it would mean that many children will start education in languages they do not understand, which increases the risk of low learning outcomes. Therefore, it needs to be given thought before even looking at the academic advantages of mother tongue based multilingual education. It is believed that the most cost beneficial approach to education in complex multilingual settings is to transition minority language children as quickly as possible to a language of wider communication. That would prevent the cost of materials development in different languages, printing books in many languages and upgrading the teacher training programmes, which are indeed some of the extra costs (Alidou et al., 2006). However, there is also a cost of children not performing well in school, dropping out of school and repeating grades, which is not calculated in these ‘budgets’. To fully understand the cost of multilingual education, one has to consider all the costs and weigh them up against one another. Taking into account better learning outcomes, if few children repeated classes and fewer dropped out, then efficiency of education improves. Furthermore, because the students are better educated, they contribute to a better educated human capital, who can then earn more. There is more evidence that a wellimplemented multilingual education programme is, in the end, less costly than the largely unproductive ‘other language only’ education systems (Bender et al., 2005). Thus, when considering the economic argument, it is important to include all costs and benefits of a certain approach and not only

Early L anguage Lear ning in Comple x Linguist ic Set t ings

21

the immediate, the most visible cost, but also take into consideration longerterm outcomes of the investment.

References Alidou, H., Boly, A., Brock-utne, B. and Satina, Y. (2006) Optimizing Learning and Education in Africa – the Language Factor. Paris: ADEA. American Institute for Research (2012) Ethiopia English Early Grade Reading Assessment: Data Analytic Report. Washington. See http://pdf.usaid.gov/pdf_docs/pnady822.pdf (accessed 25 September 2015). Asfaha, Y.M., Kurvers, J. and Kroon, S. (2009) Grain size in script and teaching: Literacy acquisition in Ge’ez and Latin. Applied Psycholinguistics 30 (4), 709. Baker, C. (2006) Foundations of Bilingual Education and Bilingualism (4th edn). Clevedon: Multilingual Matters. Bender, P., Dutcher, N., Klaus, D., Shore, J. and Tesar, C. (2005) In Their Own Language … Education for all that Children Learn Better. Washington, DC: World Bank. See https:// openknowledge.worldbank.org/handle/10986/10331 (accessed 25 September 2015). Benson, C., Heugh, K., Berhanu, B. and Yohannes Mekonnen, A.G. (2010) The medium of instruction in the primary schools in Ethiopia: A study and its implications for multilingual education. In T. Skutnabb-Kangas and K. Heugh (eds) Multilingual Education Works; From the Periphery to the Centre (pp. 40–83). Hyderabad: Orient Blackswan Private Limited. Berry, V. (2013) English Impact Report: Investigating English Language Learning Outcomes at the Primary School Level in Rural India. London: British Council. Caffery, J., Coronado, G., Hodge, B. and Taylor-Leech, K. (2014) The Timor-Leste Mother Tongue Based Multilingual Education Pilot Project, A Strategic Evaluation. Australia: University of Western Sydney. Chiswick, B.R. and Miller, P.W. (2004) Linguistic Distance: A Quantitative Measure of the Distance Between English and Other Languages (No. 1246). Bonn. See http://ftp.iza.org/ dp1246.pdf (accessed 25 Setember 2015). Ellis, R. (1994) The Study of Second Language Acquisition. Oxford Applied Linguistics. Oxford: Oxford University Press. Enever, J. (ed.) (2011). ELLiE. Early Language Learning in Europe. London: British Council. See http://www.teachingenglish.org.uk/sites/teacheng/files/B309 ELLiE Book 2011 FINAL.pdf (accessed 25 September 2015). Enever, J., Damhuis, R., Lindgren, E., Lopriore, L., Lundberg, G., Mihaljevic Djigunlvic, J. and Mestres, E.T. (2008) Tussentijdse Resultaten 2007–2008. The Netherlands: Expertisecentrum Nederlands. Enever, J., Moon, J. and Raman, U. (eds) (2009) Young Learner English Language Policy and Implementation: International Perspectives. Reading: Garnet Publishing Ltd. Gadelii, K.E. (2004) Annotated Statistics on Linguistic Policies and Practices in Africa. Paris: UNESCO. Graham, B. and Van Ginkel, A.J. (2014) Assessing early grade reading: The value and limits of ‘words per minute.’ Language, Culture and Curriculum 27 (3), 244–259. Heugh, K. (2012) The case against bilingual and multilingual education in South Africa. PRAESA, Occasional Papers. Cape Town: PRAESA. See http://paulroos.co.za/wpcontent/blogs.dir/22/files/2012/07/Paper6.pdf (accessed 25 September 2015). Heugh, K., Benson, C., Bogale, B. and Yohannes, M. (2007) Final Report Study on Medium of Instruction in Primary Schools in Ethiopia. Addis Ababa. See http://mlenetwork.org/ content/final-report-study-medium-instruction-primary-schools-ethiopia-heughet-al-2007 (accessed 25 September 2015).

22

Par t 1: Over v iews of Research Findings

Institute of International Education (2012) Enhancing the Quality of English Language Education in Ethiopia Report on a Future Search Conference (pp. 1–47). Addis Ababa: Institute of International Education. Johnstone, R. (2002) Addressing ‘the Age Factor’: Some Implications for Languages Policy. Strasbourg: Council of Europe. See http://www.coe.int/t/dg4/linguistic/Source/ JohnstoneEN.pdf (accessed 25 September 2015). Leherr, K. (2009) National Literacy Acceleration Program (NALAP ) Baseline Assessment. Washington DC: USAID. See pdf.usaid.gov/pdf_docs/PNADW581.pdf (accessed 25 September 2015). Mihaljevic Djigunlvic, J. (2014) L2 learner age from a contextualised perspective. Studies in Second Language Learning and Teaching 3 (3), 419–441. MLE-Team (2013) Report on Rajbanshi MTB_MLE program Year 2012–2013. Unpublished. Muñoz, C. (2008) Age-related differences in foreign language learning. Revisiting the empirical evidence. International Review of Applied Linguistics in Language Teaching 46 (3), 197–220. Muñoz, C. (2014) Starting age and other influential factors: Insights from learner interviews. Studies in Second Language Learning and Teaching 3 (3), 465–484. Nikolov, M. (ed.) (2009) The Age Factor and Early Language Learning. Berlin: Mouton de Gruyter. Nikolov, M. and Józsa, K. (2006) Relationships between language achievements in English and German and classroom-related variables. In M. Nikolov and J. Horvath (eds) URPT 2006: Empirical Studies in English Applied Linguistics (pp. 197–224). Pécs, Hungary: Lingua Franca Csoport, PTE. Nyaga, S. and Anthonissen, C. (2012) Teaching in linguistically diverse classrooms: Difficulties in the implementation of the language-in-education policy in multilingual Kenyan primary school classrooms. Compare: A Journal of Comparative and International Education 42 (6), 863–879. Ouane, A. and Glanz, C. (2010) Why and How Africa Should Invest in African Languages and Multilingual Education: An Evidence- and Practice-based Policy Advocacy Brief. Hamburg: UNESCO Institute for Lifelong Learning. Pflepsen, A. and Backman, S. (2012) Malawi Teacher Professional Development Support (MTPDS) The Influence of Language on Learning: Recommendations on Planning for Language Use in Education. Washington, DC: USAID. Piper, B. and Zuilkowski, S. (2015) Assessing reading fluency in Kenya: Oral or silent assessment? International Review of Education 61 (2), 153–171. Piper, B. and van Ginkel, A. (2016) Reading the script: How the scripts and writing systems of Ethiopian languages relate to letter and word identification. Writing Systems Research, 1–24. Online publication 30th September 2016. Piper, B., Schroeder, L. and Trudell, B. (2016) Oral reading fluency and comprehension in Kenya: Reading acquisition in a multilingual environment. Journal of Research in Reading 39 (2), 133–152. Probyn, M. (2006) Language and learning science in South Africa. Language and Education 20 (5), 391–414. Skutnabb-Kangas, T. (1988) Multilingualism and the education of minority children. In T. Skutnabb-Kankas and J. Cummins (eds) Minority Education (pp. 36–67). Clevedon: Multilingual Matters. Taylor, S. and Coetzee, M. (2013) Estimating the Impact of Language of Instruction in South African Primary Schools: A Fixed Effects Approach. Stellenbosch: University of Stellenbosch. Thomas, W.P. and Collier, V. (1997) School effectiveness for language minority students. National Clearinghouse for English Language Acquisition (NCELA) Resource Collection Series 9, 1–96.

Early L anguage Lear ning in Comple x Linguist ic Set t ings

23

Van Ginkel, A.J. (2014) Additive Language Learning for Multilingual Settings. Washington, DC: USAID. See pdf.usaid.gov/pdf_docs/PA00JW1R.pdf (accessed 25 September 2015). Walter, S.L. (2015) Mother tongue-based education in developing countries: Some emerging insights. Unpublished research report. Walter, S.L. and Chuo, K.G. (2012) The Kom Experimental Mother Tongue Education Pilot Project Report for 2012. See http://www. mlenetwork.org/sites/default/files/The Kom MLE Project 2012.pdf (accessed 25 September 2015). Walter, S.L. and Trammell, K. (2010) The Kom Experimental Mother Tongue Education Project Report for 2010. Washington, DC. See http://www.mlenetwork.org/content/ kom-experimental-mother-tongue-education-project-report-2010 (accessed 25 September 2015).

3

Considering the Complexities of Teaching Intercultural Understanding in Foreign Languages Patricia Driscoll

‘Learning a foreign language is a liberation from insularity and provides an opening to other cultures. A high-quality languages education should foster pupils’ curiosity and deepen their understanding of the world.’ National Curriculum (2014: 212)

The new national curriculum in England includes for the first time foreign languages (FL) as part of the statutory requirements at Key Stage 2 (7–11 years). From September 2014 all state schools are legally required to provide FL to learners from the age of 7 to 14. Extending statutory learning by four years offers the potential for enhanced linguistic proficiency and deeper levels of cultural understanding. The strong cultural statement quoted above, taken from the FL programme of study in the national curriculum, highlights the importance of culture and its role in language teaching. Cultural learning, however, is rarely given prominence in primary language lessons in England (Cable et al., 2010), a scenario reflected in primary and secondary schools across Europe. The recent Council of Europe’s guide on plurilingual and intercultural education (Beacco et al., 2015), indicates that intercultural learning and cultural enrichment have been chiefly disregarded as a central objective of language learning in the Common European Framework of Reference (CEFR) (Council of Europe, 2001) across member states. Beacco et al. (2015) argue that despite the success of the CEFR for language teaching and assessment in all sectors of education across Europe, the cultural dimension is generally marginalised in schools. This chapter focuses on cultural learning associated with primary foreign languages. Informed by a range of evidence, the chapter explores the mismatch between the clearly articulated cultural purpose of language teaching in policy documents and curriculum guidance materials and actual 24

Comple x it ies of Teaching Intercultural Underst anding in FL s

25

practice in schools (Driscoll et al., 2013). I argue that given the rise in migration across Europe and the rapidly increasing multicultural dimension of primary schools, it is timely to review cultural development in languages education and reconceptualise it within a whole-school interconnected approach. Planning a coherent approach to cultural and intercultural development, which draws upon children’s experiences across curriculum subjects as well as their experience outside of school, has more chance of liberating young learners from insularity, opening their minds to other cultures and deepening their understanding of themselves and the world. This chapter is located within the English education system, itself culturally bound with educational traditions, values, purposes and pedagogy. I hope, however, that the chapter will have broader relevance to other education systems and primary FL communities. This chapter is informed by a mixed method longitudinal research study comprising observation of lessons and interviews with teachers and head teachers, as well as focus group discussions and questionnaires to children across Key Stage 2. Data were gathered in 40 primary schools across England for three years. The study was designed to investigate primary language practice and to assess children’s progress over time in the four skills of listening, speaking, reading and writing through specifically designed language achievement tests, as well as national tests at the end of primary schooling. The findings indicated that despite the importance teachers and head teachers attributed to intercultural and cultural learning in FL, lessons generally included mere snippets of cultural information giving a shallow understanding of the target cultures and very few opportunities for cultural reflection and analysis (Cable et al., 2010). The deeper purposes of intercultural learning conveyed by the teaching staff when interviewed were unfulfilled. Distinctions need to be made between intercultural and cultural development, understanding or competence. As the prefix ‘inter’ suggests, learning or the development of competencies occur in the space between or among cultures. FL is a major mechanism within the curriculum for imparting an understanding about the cultures and social practices in another country/ countries. It acts as a window to the wider world. At a basic level, the aim of ‘intercultural’ understanding or learning implies (1) a process of selfrealisation and reflection about one’s own culture which leads to the development of a sense of a cultural self; (2) an understanding of another or ‘foreign’ culture; and (3) the capacity to function across at least two different cultural settings. The FL teacher has three interconnected tasks: (1) to provide opportunities for learners to reflect upon and analyse their own cultural reference points which inform their identity; (2) to learn about the cultural characteristics, social practices and attitudes of another country; and (3) to adapt their behaviour and attitude when in the target culture. Intercultural understanding therefore, comprises a complex set of educational aims requiring multifaceted teaching tasks.

26

Par t 1: Over v iews of Research Findings

Jerome Bruner (1996) suggests that by focusing on antinomies or contradictions within complex educational aims, we can learn useful lessons and keep issues in balance. An ongoing dilemma for primary schools is finding a sufficient supply of well-qualified teachers to teach FL. A major contradiction arises therefore from the profile of many primary teachers with limited experiences of either the target language or associated cultures and the complex cultural requirements of the subject. The comprehensive questionnaires administered by Driscoll et al. in 2004 and Wade and Marshall in 2009, indicate there has been limited change in the relatively low levels of proficiency and limited educational experiences of living abroad for primary languages teachers in England (Driscoll et al., 2004; Wade & Marshall, 2009). For the FL teacher to provide a curriculum that promotes cultural insights and the capacity to function in at least one of the countries where the target language is spoken, they themselves need to have a level of knowledge and experience. I argue in this chapter that in light of this contraction, we need to reconfigure intercultural understanding so that the priority is not on the foreign culture, but on learners as individuals, their reflections and ongoing selfrealisation. The in-depth qualitative interviews we conducted for the threeyear longitudinal study (Cable et al., 2010), reveal that primary teachers are keen to offer a range of cultural experiences for their learners in a variety of ways through different subjects inside and outside of school. Primary teachers know their children and they are perfectly placed to help them reflect upon their own culture at home and their cultural experiences at school and in the community. Through designing concrete activities and experiences that are rooted in the diversity of the local world, primary teachers can help learners to develop their own intercultural skills, tolerance, empathy and the capacity to adapt their behaviour. Before considering these issues in more depth, I shall briefly consider some key arguments for including cultural and intercultural understanding in FL learning and explore the aims and purposes of the primary curriculum within which FL is located.

The Cultural Dimension Cultural dexterity became a key strand in FL learning with the rise in importance of fluent communication for trade, banking, academic pursuits and general travel. In order to prevent miscommunication, it is argued, an interlocutor needs to understand not only the linguistic structures and vocabulary, but also how to interact in ways that are culturally acceptable to those from the target culture (Byram, 1989, 2009). This argument is based, to an extent, on the concept of a homogenised national and regional culture where all individuals in country X think, behave and interact in similar ways. Embedded within learning a FL are linguistic and cultural differences, although not as separate entities. Kramsch argues (1993, 1998) that language

Comple x it ies of Teaching Intercultural Underst anding in FL s

27

itself is a cultural product of the community and heritage, and an individual speaks with their own voice as well as with the established reference points of their native society. According to this view, language has a role in promoting and preserving culture at the same time as it continually adapts to express the needs of people and changing societies. Rather than heritage, policy documents highlight the societal benefits of language learning and the importance of developing cultural knowledge, skills, dispositions, behaviour and attitudes for economic wellbeing and workforce mobility (Council of Europe, 1982; European Commission, 1995; Nuffield Foundation, 2000). FL is considered strategically as well as economically important and an enabling skill to support both white collar and blue collar workers (Tinsley, 2013). Speaking English undoubtedly has advantages as the language of international communication and the most dominant language of the internet (British Council, 2014). An ethnocentric approach, however, hinders international partnerships and trade, particularly when selling. In addition to the benefits of intercultural competence for the economy, promoting a culturally open mind set for social harmony at home is emphasised in a number of reports about the most needed languages in the UK (Tinsley, 2013: British Council, 2014). A widely used framework for intercultural competence and communication conceptualised by Michael Byram (1997) called the five savoirs, comprises knowledge, skills and attitudes needed for successful interaction with people from other cultures. Designed to be taught by linguists for older learners, Byram and Doyé (1999) stress that the foundations for intercultural competence should be laid in primary schools. By developing the competencies within the savoirs, teachers can help children to foster an openness towards ‘otherness’ and a capacity to reflect upon cultural differences, as well as question the assumptions of their own cultural group. These competencies include, savoir, meaning knowledge of the cultural norms of self, other individuals and societal interaction. Savoir s’engager, meaning critical cultural awareness, which, according to Byram is developed through reflection and analysis of culture. Savoir apprendre/faire, which refers to the capacity for interaction and discovery and savoir comprendre, which refers to the ability to interpret something from another person’s point of view. Finally, savoir être refers to attitudes such as curiosity and openness that promote open communication, rather than those coloured by suspicion or intolerance. Byram’s framework underpins the teaching and assessment of cultural development in the CEFR and national policy in a number of countries (Byram et al., 2002; Council of Europe, 2001; Hennebry, 2014). According to the Council of Europe’s guide for plurilingual and intercultural education (Beacco et al., 2015: 10) intercultural competence is: The ability to experience otherness and cultural diversity, to analyse that experience and to derive benefit from it. Once acquired, intercultural competence makes it easier to understand otherness, establish cognitive

28

Par t 1: Over v iews of Research Findings

and affective links between past and new experiences of otherness, mediate between members of two (or more) social groups and their cultures, and question the assumptions of one’s own cultural group. The concept of intercultural communicative competence implies an ongoing process rather than a final product. Within these ambitions of cultural decentring, it is not just that a learner is aware of their own cultural judgements about rules of conduct, communication, behaviour and attitudes, but also that they are able to bracket them to reduce cultural discord that may impede communication. Drawing upon Canale and Swain (1980) and van Ek (1986), Byram and Doyé (1999) argue that proficient intercultural speakers need linguistic competence, sociolinguistic competence, which they define as: ‘the ability to understand meaning (…) including what is taken for granted and left unsaid’ and discourse competence defined as: ‘the ability to overcome difficulties of production and understanding’ (Byram & Doyé, 1999: 141). Furthermore, they suggest the ability to fill in the gaps in one’s own knowledge (strategic competence) and the attitudes, dispositions and insight to overcome misunderstandings and differences in interpretations of the same phenomenon. Early awareness of these competences could arguably begin in primary schools given the optimum conditions of highly trained staff, who themselves are interculturally dextrous with a range of high quality resources and sufficient curriculum time to explore the nuances of communication. In the meantime, while we wait for these optimum conditions, a more holistic ‘primary’ approach to intercultural learning is required.

Aims and Purposes of Education The first aim of the new National Curriculum for England relates to the purpose of education, stating that it is designed to provide pupils with ‘an introduction to the essential knowledge that they need to be educated citizens’ (National Curriculum, 2014: 6). The second aim highlights the importance of the whole school experience rather than just subjects taught in lesson time. It states that the national curriculum ‘is just one element in the education of every child’ (p. 6) and the wider school curriculum should also be well-designed for learning and skill development. In relation to subjects, there are specific requirements for cultural learning outlined in the English, art and design, design technology, history and FL curriculum. There is no indication of how cultural learning should be explored in the wider curriculum, however, the quality of cultural development overall within subjects and the wider curriculum is inspected and graded by the Office for Standards in Education (Ofsted), as part of the government’s assessment of a school’s spiritual, moral, social and cultural development (SMCS). Following an inspection, schools are awarded a grade by the inspectors ranging from

Comple x it ies of Teaching Intercultural Underst anding in FL s

29

‘outstanding’, ‘good’, ‘requires improvement’ to ‘inadequate’. According to Ofsted (2016: 35, No. 136) the definition of cultural development is demonstrated by pupils’: (1) understanding and appreciation of the wide range of cultural influences that have shaped their own heritage and those of others; (2) understanding and appreciation of the range of different cultures within school and further afield as an essential element of their preparation for life in modern Britain; (3) knowledge of Britain’s democratic parliamentary system and its central role in shaping our history and values, and in continuing to develop Britain; (4) willingness to participate in and respond positively to artistic, musical, sporting and cultural opportunities; (5) interest in exploring, improving understanding of and showing respect for different faiths and cultural diversity and the extent to which they understand, accept, respect and celebrate diversity, as shown by their tolerance and attitudes towards different religious, ethnic and socioeconomic groups in the local, national and global communities. The Ofsted framework comprises expectations about learners’ understanding of themselves (point 1); knowledge, behaviour and attitudes towards, cultural diversity (point 2); cultural heritage and values (point 3): higher order cultural engagement (point 4); and intercultural awareness (points 1, 2 and 5). The Ofsted expectations for cultural development suggest ongoing short term and long term educational aims. They also suggest a notion of a heterogeneous national group that hold the same values, which are British. The problematic concept of British values is undefined, rather the emphasis is drawn to the importance of understanding, respecting and being open and tolerant towards diversity. For pupils to demonstrate understanding and respect, they need opportunities and experiences to explore, discuss, analyse and reflect upon ‘otherness’. Alexander (2010) argues that social and cultural representation is a core purpose of primary education. This overarching goal is promoted through the curriculum as subjects are taught through a selective lens. Historical events, for example, can be a matter of some debate, but in general, history is represented in the curriculum in a selective way intended to stimulate allegiance and national pride rather than alienation, discord or shame. Primary teachers are trained to promote the cultural threads inherent in the national curriculum and schools are part of the central social infrastructure for transmitting the national cultural legacy. On the one hand, therefore, schools are tasked with ensuring respect for diversity, cultural difference and otherness; on the other hand they are a mechanism to endorse, promote and foster the nationally accepted social and cultural structures.

30

Par t 1: Over v iews of Research Findings

Through contrast, FL teachers can illuminate cultural differences, provide an opening to other cultures and enrich pupils’ sphere of influence, thereby liberating them from insularity. By comparing attitudes, perceptions and understandings to (and of) events and customs from the past and present day in a positive way, FL can help learners to identify with a different social or cultural group and in doing so motivate learners to engage with ‘otherness’, which has implications for social harmony at home. There is, however, more of a tendency for FL to skim superficially through the historic footprint, landscape, landmarks and architecture of a country or point towards a homogenised anthropological view of modes of behaviour, food and festivals. According to Grenfell (2000) the emphasis on communicating in authentic settings in secondary school modern foreign languages places the learner in the role of tourist with associated cultural references related to the tourist viewpoint. Situating the primary learner as a foreign tourist is at best premature, as children rarely travel or study abroad without their own family, or, if they do, then they travel in an educational group with whom they maintain their own cultural norms. As such, they have very limited opportunity for interaction with unknown ‘others’ when travelling. Furthermore, assuming that all children in one class or school conform to one culture contests all we know about cultural differences in our schools, not only in terms of bilingual and plurilingual families, but also in relation to sociocultural differences. The mixed methods research design of the Cable et al. (2010) study, particularly in relation to the combining of data from children’s questionnaires and focus group interview data, allowed us to penetrate some of the complexity of investigating perceptions to learning about and attitudes towards cultural differences. Approximately 140 children were surveyed in each year from the age of seven followed by focus group interviews in each year group. The quantitative data indicated that younger children (aged seven and eight) reported enjoying learning about the country more than the older children (aged nine and 10). However, 94% of children aged 10 and 11 reported that they would like to visit the country, even though some did not enjoy learning about it in lessons. The focus group interviews revealed a more complex scenario. Children spoke of their enjoyment and interest in cultural experiences at school such as learning about different customs, sampling traditional food and going on school trips, but some of the older children reported an overall sense of boredom with lack of challenge in the FL curriculum and the repeated diet in terms of resources and activities. Older children, therefore, did indeed enjoy learning about culture less, but because of the pedagogy rather than because of their attitudes towards others. Individual differences were also highlighted through the interviews rather than through the surveys, for instance, one learner said: half of my family is Indian and I think learning French and lots of other languages sort of brings you more into life … you know what’s

Comple x it ies of Teaching Intercultural Underst anding in FL s

31

going on in the world; you know how people live… we learnt a lot about different places in France; so not just one place (Year 6 Focus group 2008–09) There is a tendency in primary FL to teach children about cultural differences between one country and another as though they were two homogenised cultural groups (Cable et al., 2010). In other words, the English do this, and the French, Spanish and Germans do that. It is difficult to see how assuming that all children in one class adhere to a notional national norm with one language and one set of beliefs, attitudes and behaviours, in order to compare them with another homogenised group in another country, will lay foundations for tolerance, empathy and respect for other cultures. The notion of sameness may work on a superficial level, but if changing attitudes and developing certain dispositions and approaches to interaction are the central goal, then creating an ‘us’ and ‘them’ flies in the face of logic. While the notion of national and regional identity may be important in creating a sense of belonging, and while it is important that children learn and adhere to societal and school rules, conceptualising ‘otherness’ through national differences may suggest a judgement about something ‘wrong’ or ‘odd’. Lo Bianco (2014) challenges the term ‘foreign’ in language education, arguing that globalisation, communication, interconnectedness, unprecedented expansion of migration and population shifts have converted most countries into multi- or pluri-cultural and lingual societies. Arguably, it is time to rethink intercultural development in primary schools and place a greater emphasis on children’s personal reflection, their own cultural reference points and developing a sense of pride in their own individual cultural identity. Drawing upon the existing cultural learning experiences of children in the class as a foundation for cultural learning requires a shift from a nostalgic notion of an unchanging culture deeply steeped in the soil of nations to a living dynamic concept of culture that evolves as individuals and societies change. Rather than focusing solely on cultural differences between home and abroad, the important skills of intercultural understanding, outlined in the savoirs, could be explored through a focus on the social and cultural references of members of a class and school, whether children are multilingual or not, as well as the diverse cultural reference points in schools abroad.

Validating Children’s Sense of Self through Cultural Reflection Children do not start school culturally neutral. The majority of learners inhabit a space with their family or guardians and if they attend a playschool or nursery, they meet a small number of other children and adults.

32

Par t 1: Over v iews of Research Findings

Within this small, local learning society they learn to recognise customs and values placed on different symbols and approaches to interaction. Donaldson (1978) suggests that in general, young learners have the same processes of interpretation as adults and the capacity to make sense of direct human interaction, which Bruner (1983) argues, enables children to manage highly complex scenes of human interaction regularly. As soon as they start school, children join a larger community with a great number of people. They begin to learn a new cultural code, new knowledge, skills and different approaches to interaction. The cultural influences and codes of behaviour learnt in some home environments are more aligned to the school culture. Some children therefore, have to modify their behaviour more than others. All children, however, have to adapt to the new learning environment and adopt new practices. Many of the rules and codes are made explicit as the teacher manages the class and teaches; however, codes of conduct are rarely open to discussion, rather, they are presented as the ‘right’ way to behave. In general schooling, children are rarely given opportunities to compare their own cultural domains with others within the class or critique the codes of conduct they are being taught. How then, can the foundations for intercultural competence be laid in primary school if learning does not start from the child’s own experience? Intercultural understanding essentially requires being open to different practices and suspending judgement in an attempt to understand the world from a different perspective. Is it not better therefore, to start reflecting upon difference within the safe space of classmates and friends? The concept of funds of knowledge for capitalising on children’s social and cultural learning experiences outside of school is useful here. Funds of knowledge refer to the knowledge, skills and ways of doing and living, which connect home, school and community (Moll et al., 1992; Gonzalez et al., 1995). Based on lived experiences within the family, they are according to Moll et al. (1992: 133) ‘historically accumulated and culturally-developed bodies of knowledge and skills essential for household and individual functioning and well-being’. Through these lived experiences, people develop values, beliefs, attitudes and behaviours including their approach to social interaction. González and Moll (2002) argue for a pedagogy that is more closely linked to students’ lives, their histories and their community. Drawing upon a study that involved students and families from Latin American backgrounds, they investigated how local funds of knowledge can be used as a foundation for future learning in order to validate students’ identity and enhance their engagement. Using ethnographic research methods, including participant observation, interviewing, life history narratives and field notes, teachers and high school students investigated the dynamic processes of learning in the daily practices of student households. The findings of their study indicated that over time, the status of local cultural and social capital grows and the funds of knowledge become a valuable resource

Comple x it ies of Teaching Intercultural Underst anding in FL s

33

for reconceptualising the academic discourse in the classroom. Links between the home and school also improved and teachers’ professional practice was transformed through participation in the research study. Esteban-Guitart and Moll (2014) extend the concept of ‘funds of knowledge’ to understand identity, which they refer to as ‘funds of identity’. They suggest that through investigating the practices, beliefs, knowledge and ideas that are culturally developed and socially imparted, insights can be gained about how individuals perceive themselves and make meaning. Rather than self as a metaphysical concept, they argue, self can be empirically linked to what one does. Identity, rather than solely in the mind, is embedded in the artefacts and activities of the social and cultural contexts. They suggest that through drawings, self-portraits and discussion with children, their funds of knowledge can be explored and insights gained into how, if at all, their funds of knowledge differ from those of adults. Designing a school curriculum to reflect funds of knowledge developed at home may be too radical in light of a national curriculum constructed around subject disciplines; however, it makes perfect sense to begin cultural and intercultural development in primary school around a child’s own cultural funds of knowledge and identity. While it may be impractical for all FL teachers to conduct ethnographic research in children’s homes, finding a space on the curriculum to help children research the artefacts, activities, resources and approaches to interaction of their home and community as a basis for sharing with others in the class is not. There are more than a million learners in UK schools who speak a language other than English (National Association of Language Development in the Curriculum, 2015) and the number is increasing as migrants and refugees arrive in schools from across the world. Valuing the diversity of cultures within one class and school can help children settle and help teachers reveal the rich layers of cultural experiences of children in the class. Conteh’s (2012) longitudinal case studies contribute a further dimension. Using linguistic ethnographic methodology, she tracked children’s learning experiences at home, in the community and at school to investigate bilingual children’s concept of funds of knowledge. Her findings suggest that using the concept of funds of knowledge, bilingual children’s underachievement can be addressed. She argues that classrooms need to be places where ‘threads, strands and fibres of global, national and local discourses are dialectically mediated and dialogically played out in the interactions between teachers and learners’ (Conteh, 2012: 114). Teaching about the threads and strands of the target culture, whether information about the landscape and festivals, customs when greeting others or slight differences in the organisation of schooling, could become part of a coherent programme which celebrates diversity at home and abroad. Culture is both dynamic and responsive. Rather than straying into stereotypes, FL could become the vehicle for extending the discourse of inclusion. Britain, for example, is classed as a black tea drinking nation. It is rated fifth in the world

34

Par t 1: Over v iews of Research Findings

of tea drinking countries (Wikipedia, 2015). There are, however, many Britons, who prefer coffee and some drink neither tea nor coffee. Information that tea consumption is extremely high in Britain, belies the fact that teadrinking is rapidly reducing (Sturgess, 2013). Primary schools are increasingly connected to other schools locally and nationally through a variety of links and partnerships. Almost half of the schools in the Cable et al. (2010) study had developed links associated with FL learning with at least one school abroad. These links offered opportunities for children to practise language through sharing curriculum resources and information about their school and their lives. A number of schools also organised face-to-face meetings and educational trips to their partner school abroad and some schools extended these links for joint curriculum planning and cross-curricular projects. Significantly, the mixed-method study conducted by Cable et al. (2010) facilitated analysis of multiple perspectives which provided a more comprehensive understanding of learner’s views about cultural experiences within FL. As discussed above, the focus group discussions both contextualised the responses from the questionnaire administered to pupils and contributed an individual learner perspective. The British Council’s, School’s-Online website depicts a number of projects where schools in different countries have developed cross-cultural curriculum themes to enable children to see first-hand different approaches to knowledge, skill-development, customs and interaction. These projects do not require linguistic knowledge and they indicate an increased motivation to expand cultural opportunities in schools. Providing a coherent cultural programme where learning objectives are planned rather than ad hoc, requires teachers to conceptualise culture within a pedagogic frame.

Primary Teachers’ Toolkit The predominant model for teaching foreign languages in primary schools in England is a teacher who is not a linguist and who is unlikely to have an in-depth knowledge of the target country (Cable et al., 2010; Driscoll et al., 2004; Wade & Marshall, 2009). The Driscoll et al. (2004) study highlighted the limited amount of cultural learning associated with FL across England. Only 22% of the 2966 responding schools considered positive cultural benefits arose from FL learning. The qualitative evidence, however, drawn from interviews and discussion groups with a range of key stakeholders in 15 case study primary schools in different local authorities, showed vast regional differences. Where there was a perceived need and where teachers had expertise, the cultural dimension was at the heart of FL learning. In one case study school, for example, based in a multilingual area with over 70 languages represented in the school, the aim of teaching FL centred on intercultural learning and language awareness. The school offered a range of

Comple x it ies of Teaching Intercultural Underst anding in FL s

35

stimulating cultural activities in 12 languages including Mandarin Chinese and Welsh. The local authority provided a wealth of resources for developing children’s cultural understanding and tolerance. For this study, funded by the government, the qualitative evidence, although offering an element of triangulation, played a secondary role to the quantitative evidence. At that time, the Department for Education were interested in providing a baseline of practice from which to measure growth rather than understanding individual differences. Only 20% of primary FL teachers are specialists who are likely to have had an extended sojourn in a country where the target language is spoken as part of their educational studies (Wade & Marshall, 2009). The assumption embedded within policy that primary FL teachers are equipped to teach intercultural understanding by drawing upon their own experience of the target country is overly optimistic (Driscoll et al., 2013). Indeed, it is a major contradiction in policy to expect primary FL teachers to facilitate intercultural understanding in a similar way to secondary school modern foreign language teachers. Primary schools are perfectly placed to explore cultural differences, motivate learners to enjoy ‘otherness’ and help them to develop social and cultural skills. The primary workforce are expert in teaching many if not all subjects across the curriculum. Their professional lives focus on developing the ‘whole child’, yet cultural learning in FL still focuses entirely on teaching the culture associated to the target language, rather than including references about the target culture within a larger programme of cultural learning. According to Cable et al.’s (2010) study, primary age children in England, who are learning a foreign language in school, learn a body of phrases, a bank of nouns in different topic areas and a few verbs each year. Teachers generally drew attention to well-known characteristics such as landmarks or celebrations when teaching the language. Most teachers in the case study schools also drew attention to the commonalities and differences between the two language structures, such as adjectival agreements or verb construction. Eric Hawkins (1981, 1984) argues that developing an awareness of language by comparing different languages side by side creates a firm foundation for subsequent language learning. This may be the case for language, but there is little evidence to suggest that a firm foundation is being laid for subsequent intercultural learning, despite the increased availability of multimedia resources. The limited availability of intercultural training to help primary teachers to develop their understanding and skills leads to limited systematic planning for cultural development across year groups. As a consequence, the cultural curriculum is not designed around learner’s developing social and cultural capacities or their interests (Driscoll et al., 2013). A large number of teachers have to rely on their own ‘toolkit’ of experiences and many feel ill-equipped to teach intercultural understanding, not only in England but across the world (Driscoll et al., 2004; Languages and

36

Par t 1: Over v iews of Research Findings

Cultures in Europe, 2007). In 1997, international primary language experts met in Graz, Austria, to discuss the needs for children, teachers and schools of the new phenomenon of early language learning. At that time, there were a limited number of well-informed academics in primary FL pedagogy and very few empirical studies to inform the discourse. Twenty years later, the situation has changed and we now have a growing body of research studies, well-informed academics and FL modules that have been designed as part of initial training programmes alongside other foundation subjects, such as geography. Irrespective of their own language competence, student teachers are introduced to the basics of teaching the foreign language, but there are still limited modules offering opportunities for intercultural development during pre-service training. For a short period, the English government funded FL primary specialist or enhanced programmes for student teachers, which included a four-week intense intercultural and teaching placement programme both in the UK and abroad. Driscoll and Rowe (2012) investigated the professional benefits of the short teaching placement abroad for student teachers. The mixed method research study involved 72 postgraduate students from Canterbury Christ Church University and 50 postgraduate students from Liverpool Hope University, who sojourned for four weeks in France, Spain, Italy and Germany, where they undertook an assessed teaching placement in a local primary school. During their placement abroad, students were required to complete reflective journals, which probed their emerging understanding of pedagogic practices within a contrasting cultural context and their response to the differences that they perceived. On their return, students completed evaluations and a questionnaire. The questionnaire used Likert items to rank how significant the students felt the impact of the placement abroad had been in altering their perceptions. Students were prompted to expand their responses in the spaces provided for qualitative comments. The questionnaire yielded a 78% response rate. In order to add further validity to the findings, semi-structured interviews and discussions were organised with small focus groups of students. Student teachers’ comments are therefore drawn from questionnaire, focus groups, evaluation and reflective journal data. Analysis involved a framework of frequency of occurrence, clustering of themes, a chain of evidence and identifying the relationship between different themes. Overall, the study found that a short placement abroad developed students’ capacity for critical cultural awareness through their reflection on differences in pedagogic principles and practice and offered a unique opportunity, through a lived experience of juxtaposition and comparison, to re-evaluate and question taken-for-granted practices and theoretical positions. These programmes are no longer funded and in-service opportunities have also diminished due to financial constraints on public funding in England. Primary teachers therefore, have far fewer funded opportunities to enhance their own intercultural skills within a professional or educational

Comple x it ies of Teaching Intercultural Underst anding in FL s

37

environment. Many young primary teachers have benefited from gap years or personal travels across the world, rather than educational sojourns, but to rely on these experiences for teaching such an important part of a curriculum appears a little random in relation to curriculum development. Embedded within the primary pedagogue’s role is the promotion of children’s social, cultural and life skills, their character and moral judgment. Teaching English or the language of instruction is a case in point. The widespread use of children’s literature, for example, is not purely to enhance literacy skills and reading enjoyment or as a stimulus for imagination and creative writing, but also it provides a means by which culture is represented. Anthony Browne’s picture book (1998) for example, Voices in the Park, explores the perceptions of four distinct characters as they encounter each other in the park. It is a complex tale of mistrust and judgement narrated through a simple story with rich colourful illustrations, which infuse the drama with humour. Issues of social class and gender are raised, as are perceptions of others and how we judge those who are not in our sociocultural sphere. Given the place of such activity in the current curriculum, promoting intercultural understanding is arguably an extension of primary teachers’ practice rather than a new set of skills. Teachers are increasingly using picture books to extend vocabulary and stimulate reading and writing activities, and in some cases to explore social, cultural and historical themes to expand learners’ understanding of the world beyond their own (Mourão, 2015). Songs, poems and music are also commonly used to help children learn the language (Cable et al., 2010); however, without a clear concept of intercultural understanding and associated training opportunities are lost.

Conclusion This chapter explores ways in which FL in primary schools could contribute to children’s overall cultural development. I draw upon a number of studies using a mixed method approach to suggest that a contradiction exists in teaching the cultural dimension created by requirements to teach about the cultures associated with the FL by teachers who are not familiar with the culture. I argue that rather than draw upon the notions of differences between national groups, cultural understanding should begin with the child as an individual. For children to develop a reflexive approach to their own cultural development and a curiosity about different cultural influences in their lives (savoir être), learning needs to start with their own cultural world. Developing skills of critical cultural awareness (savoir s’engager) requires reflection and analysis of cultural variance, which cannot be undertaken without experience and young children are most experienced about life at home and at school. Through age appropriate auto/narrative research and enquiry

38

Par t 1: Over v iews of Research Findings

developed in class they could develop a deeper understanding of themselves (savoir) and potentially a greater capacity to understand others (savoir comprendre). This chapter argues that through encouraging children to research, reflect upon and analyse their own cultural influences, through discussions about their ‘funds of knowledge’ (Gonzalez & Moll, 2002) and ‘funds of identity’ (Esteban-Guitart & Moll, 2014), they can review their own cultural reference points and their connections to the local, national and international world. Learning a foreign language could form a rich part of this overall cultural programme. Primary teachers in England spend a large part of every day with the children in their class, they know their children as individuals and they are significant persons of influence. I argue that the best introduction to intercultural understanding is to start with the experiences of ‘otherness’ in the class and the school. Primary teachers can teach the language well, but there is a tendency towards providing superficial snippets of cultural information, rather than an enriching programme that can ‘liberate learners from insularity or provide an opening to other cultures’ as the quote at the beginning of the chapter suggests. Exploring the diversity within local cultures, primary teachers could deploy their own understanding of the children in their class, their families, the community and the curriculum. By developing a reflexive and mindful approach to children’s own cultural influences, children could be helped to share their cultural narrative with other children in their school, in other partner schools in the UK and schools abroad, thereby problematising the ‘right’ version of cultural norms. Drawing upon the children’s experience, FL could drive a whole school systematic plan for cultural learning which would cultivate the character, dispositions, understanding and attitudes needed for the increasing multicultural societies at home and across the world.

References Alexander, R. (2010) (ed.) Children, Their World, Their Education. Final Report and Recommendations of the Cambridge Primary Review. Oxon: Routledge. Beacco, C., Byram, M., Cavalli, M., Coste, D., Egli Cuenat, M., Goullier, F. and Panthier, J. (2015) Guide for the Development and Implementation of Curricula for Plurilingual and Intercultural Education. Strasburg: Council of Europe, Language Policy Unit. See http:// www.coe.int/t/dg4/Linguistic/Source/Plurilingual%20and%20Intercultural%20Edu cation.pdf (accessed 12 March 2017). Browne, A. (1998) Voices in the Park. London: Dorling Kindersley Publishing, Picture Corgi Books: Random House Children’s Publishers UK. Bruner, J. (1983) Child Talk. New York: Norton. Bruner, J. (1996) The Culture of Education. London: Harvard University Press. British Council (2014) Languages for the Future. Which Languages the UK Needs Most and Why. UK: British Council. See https://schoolsonline.britishcouncil.org/ (accessed 12 March 2017). Byram, M. (1989) Cultural Studies in Foreign Language Education. Clevedon: Multilingual Matters.

Comple x it ies of Teaching Intercultural Underst anding in FL s

39

Byram, M. (1997) Teaching and Assessing Intercultural Communicative Competence. Clevedon: Multilingual Matters. Byram, M. (2009) Intercultural competence in foreign languages – The intercultural speaker and the pedagogy of foreign language education. In D. Deardorff (ed.) The Sage Handbook of Intercultural Competence (pp. 321–332). London: Sage Publications Ltd. Byram, M. and Doyé, P. (1999) Intercultural competence and foreign language learning in the primary school. In P. Driscoll and D. Frost (eds) The Teaching of Modern Foreign Languages in the Primary School (pp. 138–151). London: Routledge. Byram, M., Gribkova, B. and Starkey, H. (2002) Developing the Intercultural Dimension in Language Teaching, A Practical Guide for Teachers. Strasbourg: Council of Europe. See http:// www.coe.int/t/dg4/linguistic/source/guide_dimintercult_en.pdf (accessed 14 July 2014). Cable, C., Driscoll, P., Mitchell, R., Sing, S., Cremin, T., Earl, J., Eyres, I., Holmes, B., Martin, C. and Heins, B. (2010) Languages Learning at Key Stage 2: A Longitudinal Study Research Report No. 198. London: DCSF. See https://www.gov.uk/government/publications/languages-learning-at-key-stage-2-a-longitudinal-study-finalreport (accessed 17 June 2014). Canale, M. and Swain, M. (1980) Theoretical bases of communicative approaches to second language teaching and testing. Applied Linguistics 1 (1), 1–47. Conteh, J. (2012) Families, pupils and teachers learning together in a multilingual British city. Journal of Multilingual and Multicultural Development 33 (1), 101–116. Council of Europe (1982) Recommendation of the Committee of Ministers to Member States concerning Modern Languages. Recommendation (82) 18. Retrieved on 8th April 2017 from https://rm.coe.int/CoERMPublicCommonSearchServices/DisplayDCTMCont ent?documentId=09000016804fa45e Council of Europe (2001) Common European Framework of Reference for Languages: Learning, Teaching, Assessment. Cambridge: Cambridge University Press. National Curriculum (2014) Department for Education. Retrieved on 8th April 2017 from https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/ 335116/Master_final_national_curriculum_220714.pdf Donaldson, M. (1978) Children’s Minds. London: Croom Helm. Driscoll, P. and Rowe, J. (2012) Broadening the lens: An investigation of student teachers’ changing perceptions of pedagogy following a teaching placement in a primary school in mainland Europe. Education 3–13 40 (4), 417–431. Driscoll, P., Jones, J. and Macrory, G. (2004) The Provision of Foreign Language Learning for Pupils at Key Stage 2. DfES Research Report RR527. Nottingham: DfES Publications. Driscoll, P., Earl, J. and Cable, C. (2013) The role and nature of the cultural dimension in primary modern languages. Language, Culture and Curriculum 26 (2), 146–160. Esteban-Guitart, M. and Moll, L. (2014) Funds of identity: A new concept based on the funds of knowledge approach. Culture and Psychology 20 (1), 31–48. European Commission (1995) COM (95)590 White Paper on Education and Training: Teaching and Learning; Towards the Learning Society. Brussels, online. See http:// europa.eu/docu ments/comm/white_papers/pdf/com95_590_en.pdf (accessed 12 March 2017). González, N. and Moll, L.C. (2002) Cruzando el puente: Building bridges to funds of knowledge. Educational Policy 16 (4), 623–641. Gonzalez, N., Moll, L.C., Tenery, M.F., Rivera, A., Rendon, P., Gonzales, R. and Amanti, C. (1995) Funds of knowledge for teaching in Latino households. Urban Education 29 (4), 443–470. Grenfell, M. (2000) Modern languages beyond Nuffield and into the 21st century. Language Learning Journal 22 (1), 23–29. Hawkins, E. (1981) Modern Languages in the Curriculum. Cambridge: Cambridge University Press.

40

Par t 1: Over v iews of Research Findings

Hawkins, E. (1984) Awareness of Language. Cambridge: Cambridge University Press. Hennebry, M. (2014) Cultural awareness – Should it be taught? Can it be taught? In P. Driscoll, E. Macaro and A. Swarbrick (eds) Debates in Modern Languages Education (pp. 135–149). London: Routledge. Kramsch, C. (1993) Context and Culture in Language Teaching. Oxford: Oxford University Press. Kramsch, C. (1998) Language and Culture. Oxford: Oxford University Press. Languages and Cultures in Europe (LACE) (2007) The Intercultural Competences Developed in Compulsory Foreign Language Education in the European Union European Union EAC/25/06. Brussels: DG Education, Training, Culture and Multilingualism European Commission. See http://nellip.pixel-online.org/files/publications_PLL/18_Languages%20and%20 Cultures%20in%20Europe%20(LACE).pdf (accessed 12 March 2017). Lo Bianco, J. (2014) Domesticating the foreign ‘globalization’s effects on the place/s of languages. The Modern Language Journal 98 (1), 312–324. Moll, L.C., Amanti, C., Neff, D. and Gonzalez, N. (1992) Funds of knowledge for teaching: Using a qualitative approach to connect homes and classrooms. Theory into Practice 31 (2), 132–141. Mourão, S. (2015) English in pre-primary: The challenges of getting it right. In J. Bland (ed.) Teaching English to Young Learners: Critical Issues in Language Teaching with 3–12 Year Olds (pp. 51–70). London: Bloomsbury Publishing. NALDIC. National Association of Language Development in the Curriculum (2015) EAL Statistics. The Latest EAL Facts and Figures. See http://www.naldic.org.uk/researchand-information/eal-statistics (accessed 9 September 2015). Nuffield Foundation (2000) Languages: The Next Generation Final Report of the Nuffield Languages Inquiry. London: The Nuffield Foundation. Ofsted (2016) School Inspection Handbook. August 2016, Reference no: 150066. See https:// www.gov.uk/government/uploads/system/uploads/attachment_data/file/553942/ School_inspection_handbook-section_5.pdf (accessed 27 October 2016). Sturgess, E. (2013) No time for tea? How Britain became a nation of coffee drinkers. The Guardian, 19 December 2013. See http://www.theguardian.com/lifeandstyle/ wordofmouth/2013/dec/19/tea-coffee-drinkers-sales (accessed 12 October 2015). Tinsley, T. (2013) Languages: The State of the Nation Demand and Supply of Language Skills in the UK. London: British Academy. van Ek, J.A. (1986) Objectives for Foreign Language Learning. Volume I: Scope. Strasbourg: Council of Europe. Wade, P. and Marshall, H. with O’Donnell, S. (2009) Primary Modern Foreign Languages Longitudinal Survey of Implementation of National Entitlement to Language Learning at Key Stage 2. Research Report No. RR127. London: DCSF. Wikipedia (2015) List of countries by tea consumption per capita. See https://en.wikipedia. org/wiki/List_of_countries_by_tea_consumption_per_capita (accessed 26 August 2015).

4

Literacy Development in Children with English as an Additional Language (EAL) Victoria A. Murphy

Introduction Languages, like language learners, come in many varieties. In considering the construct of a ‘minority’ language, there are some important issues to ponder, one of which relates to the concept of relativity. I am not referring to the famous scientific theory of Einstein here (at least not directly) but rather, the notion that to consider whether a language is a minority language we need to think of it in relation to something else (i.e. other languages or other speakers). A language can be a so-called minority language because, in an ethnolinguistic sense, very few speakers in the world speak that language and it may even be in danger of dying out. For example, the language ‘Manx’ (or Manx Gaelic) is a language indigenous to speakers from the Isle of Man in the Irish Sea. Approximately 86,000 people live on the Isle of Man and of these, only a very small proportion speak Manx fluently (some estimates are as few as 50 people). Hence Manx is, ethnolinguistically speaking, a minority language (at least for now). The other way to think about ‘minority’ language relates to the proportion of speakers of different languages within given societies. If a child is growing up in England where the majority language is English, but has parents who speak Arabic, lives in a community where everyone around her speaks Arabic, and in essence develops Arabic as an L1, she will be considered a minority language learner because her home language is in the minority relative to the rest of the country (where English is spoken). It would be absurd to conceive of Arabic as a minority language in a general sense, since it is such an important world language and its dialects are spoken by millions of people across the globe (unlike Manx Gaelic). Nonetheless, in an English speaking country, Arabic speakers are in the minority. Thus, a minority language learner is either someone who speaks a language that is ethnolinguistically in the minority (like Manx) or someone 41

42

Par t 1: Over v iews of Research Findings

whose specific home language is in the minority relative to the language of the wider community/society. De Bot and Gorter (2005) have used the terms ethnic versus regional minority language learners to highlight this distinction where the speaker of Manx is a regional minority language learner and the speaker of Arabic in England is an ethnic minority language learner. The focus of this chapter is on children aged 3 to 12 years who are ethnic minority language learners in de Bot and Gorter’s (2005) sense, and in particular, I focus on some of the specific challenges faced by many such children in developing literacy skills in the majority language. The work presented in this chapter is by no means an exhaustive or systematic review of research literature that has addressed this important issue. Rather, it presents a kind of snapshot of some of the issues being addressed by some researchers, educators and policy makers concerning how best to educate children from linguistically diverse backgrounds. There are numerous important and challenging issues related to educational provision and practice for children who are minority language learners, which include how best to support the child’s first language (L1) in second language (L2) majority educational contexts, what kind of pedagogical practice is most supportive of emergent bilingual children’s bilingual competence, how to ensure that children from minority language backgrounds are valued and that their linguistic diversity is considered more of a resource (and less of a problem as unfortunately it so often is), and whether and to what extent we can modify assessment practices to reflect the reality that growing numbers of pupils are bilingual, even if so many schools are not. As this chapter is specifically focused on literacy skills, the discussion will be limited to some of the main factors known to influence developing literacy and for which we have some tangible evidence that children from minority language backgrounds are likely to struggle (e.g. vocabulary). Finally, while minority language learners are found throughout the globe and are learning many different majority languages, it is still unfortunately true that in terms of research on literacy development at least, English dominates as the most commonly discussed majority language in which minority language learners are developing literacy skills. Given the dominance of English in the literature, the main focus of discussion will be on children educated through the medium of English, in English speaking countries, but who come from home language backgrounds where English may not be present, or if it is, it is not the only language (in other words, children with English as an Additional Language (EAL)). Clearly, however, it is an empirical question whether, and perhaps more importantly to what extent, the same issues are relevant in other linguistic contexts. As with all areas of research, investigations of literacy development can be carried out using different methodologies, depending upon the specific focus of the study. Studies tend to either follow a quantitative, qualitative or mixed-methods approach. It is important to point out that no one approach is superior to any other. Rather, as with all research, the appropriateness of

Literac y Development in Children w ith English as an Addit ional L anguage

43

adopting quantitative or qualitative methodologies very much depends upon the nature of the research questions driving a particular study. Many of the studies discussed below that focus on specific L2 (English) outcomes in children with EAL adopt a quantitative methodology, as the primary focus of many of these studies is whether children with EAL differ from their nativespeaking (NS) peers. In addressing this type of question, researchers often adopt a more experimental or quasi-experimental design to make specific comparisons between the two groups in terms of target L2 outcomes. However (and not less importantly), other research has adopted more qualitative designs, which include a focus on issues such as the learning environment of the child, both in home and in school (e.g. Gregory, 2008; Gregory & Williams, 2000), or whether and to what extent particular pedagogical approaches which allow EAL children to use their home language might be beneficial (e.g. Hornberger & Link, 2012; Jonsson, 2013). These two approaches (quantitative and qualitative) are complementary and each offer different perspectives and insights into the task of developing literacy skills in a second language, as will be demonstrated below. The potential for combining the two approaches, adopting a mixed methods research framework, remains somewhat under-explored despite its potential for offering more comprehensive insights.

Minority Language Learners Before beginning to discuss literacy in minority language learners, however, it might be worth first asking the ‘so what’ question. In other words, why should we concern ourselves about the language and literacy development of ethnic minority language learners? Everyone knows that children are excellent language learners and as long as we start educating minority language learners early enough, given that they will be spending the entire school day within an English speaking environment and in an English speaking country, they are bound to pick up the necessary language skills to support their academic achievement. Right? Well, not necessarily. First, while it may be that there are no cognitive or developmental reasons to doubt that children can develop bilingually without cost to their languages or academic achievement, age alone will not guarantee success (Murphy, 2014). There are a whole host of variables which impinge upon the child’s ability to develop sufficiently advanced language and literacy skills in which to access the academic curriculum, some of which will be discussed below. Furthermore, if we examine international achievement studies we see a disturbing trend. The Programme for International Student Assessment (PISA) run by the Organisation for Economic Co-operation and Development (OECD) consistently identify that children from first and second generation immigrant families underperform relative to native speaking children on mathematics and reading. Indeed, PISA

44

Par t 1: Over v iews of Research Findings

results from the last few rounds of data collection (2003, 2006, 2009, 2012) have regularly shown that students from immigrant backgrounds underperform relative to native speakers, even when differences in socioeconomic status are factored out. The Progress in International Reading Literacy Study (PIRLS) illustrates the same findings. Figure 4.1 is taken from one of the OECD’s documents reporting on the PIRLS 2006 study. In this figure we see a clear trend that first and second generation ethnic minority children tend to underperform relative to other children on reading. Note too, however, that there is variability across countries, where some show little to no differences between the three categories of students depicted in Figure 4.1. This is an important point – namely that the minority language learner population is heterogeneous and it would be reckless to assume that just because a child is a minority language learner that they will necessarily have difficulties with their majority language literacy skills. Rather, the data depicted in Figure 4.1 illustrate an international trend. These findings are consistent with other work on English language learners (ELLs) in the North American context illustrating that ELL children are often over-represented in remedial support programmes (Paradis et al., 2011). This manifest difference between ethnic minority and native-speaking children is important, because as stated by the OECD ‘Integrating immigrant students into schools is a challenge for most countries; yet a country’s success in integrating immigrants’ children into society is a key indication of the efficacy of social policy in general and education policy in particular’ (OECD, 2013: 71). Hence, there are lessons to be learned about how we develop educational policy and curricula based on these and other studies.

Figure 4.1 PIRLS 2006: Reading scores by immigrant status at age 10 (OECD, 2012: 4)

Literac y Development in Children w ith English as an Addit ional L anguage

45

There are other issues as well that are worth considering. Jim Cummins (2000 and elsewhere) has written extensively about the destructive influence of negative attitudes towards many immigrant communities, attitudes that permeate and pervade the educational system and are reflected in attitudes and perceptions about how we should support the education of children from ethnic minority backgrounds. As Cummins notes, the L1 of putatively ‘low status’ communities is rarely included in a meaningful or effective way in classrooms and he advocates that teachers need to create ‘pedagogical spaces’ (Cummins, 2012: 1987) to use the L1 as a resource. These concerns are also closely tied to the question of identity. Given that most children from linguistically diverse backgrounds are not educated through the medium of their L1 – and importantly that teachers tend not to valorise the use of the home language in the classroom – the child is regularly sent the message that their linguistic culture is of no importance. This is one of the reasons why bilingual education programmes, such as those reviewed in Murphy (2014), are so important. The dual language immersion programme in particular, where children are educated through the medium of both their home language and the majority language, alongside majority language speaking children (e.g. Spanish-English dual immersion programmes in the US) have been consistently shown to be effective in promoting language and literacy in ELL/EAL children. For a child learning English as second, or additional language through formal education settings, seeing the English nativespeaking child learning their (minority) home language goes a long way to showing that child that their linguistic background is valorised. Unfortunately, it is not always possible to implement such programmes, particularly in countries where there are comparatively few homogenous L1 (home language) groupings. One such country is the UK, the context to which I now turn.

English as an Additional Language in the UK Children who have EAL are those as described above, children with a home language that is not English, yet who are living in and being educated in England, through the medium of the majority language – English. There are currently over a million pupils in England who are classified as EAL and who represent over 360 different home languages (NALDIC, 2014). There have also been significant increases in the numbers of EAL children over the years, where in 1997 the proportion of EAL pupils in England was 7.6% whereas it is 19.4% in 2015. Children with EAL in England are found throughout the country but there are areas where there are denser proportions of EAL children, such as in London and areas such as Birmingham, Manchester, and Bradford. The distribution of EAL pupils at the school level is highly skewed, where around one quarter (22%) of schools have fewer than 1% of EAL pupils.

46

Par t 1: Over v iews of Research Findings

There are other schools, however, which have 100% EAL pupils. Over one thousand schools in England (representing 8.4% of the schools in England) have more than 50% EAL pupils registered in the school and 762 of these are outside London (Strand et al., 2015). Table 4.1, taken from Strand et al. (2015), illustrates the achievement gap between EAL and non-EAL pupils broken down by age. Table 4.1 highlights a number of interesting issues for EAL pupils in England. First, by the end of the first year of schooling only 44% of children recorded as EAL achieve a ‘good level of development’ as measured by the Early Years Foundation Stage Profile (EFSP), relative to 54% of non-EAL pupils. The odds ratio analysis carried out by Strand et al. (2015) reveals that the odds of achieving a ‘good level of development’ for EAL pupils are 33% lower for EAL students relative to non-EAL pupils. As stated in Strand et al. (2015: 27): ‘We conclude that, perhaps not surprisingly, at the end of their first year of fulltime education children from homes where they may have had less exposure to English on average achieve lower results than those with FLE (first language English)’. However, what is also notable in Table 4.1, is that the gap between EAL and non-EAL children seems to narrow over time. At age 7, the odds ratio analysis is 0.73 and at age 11 it is 0.81 and by age 16 it is only 0.90. There are

Table 4.1 EAL versus non-EAL and achievement at age 5, 7 11 and 16: England 2013 Age

Stage

Domain

Measure

Source

EF %

EAL %

Odds Ratio

5

EYFSP

Reading Maths Overall

SFR 2013-47

73 71 54

63 62 44

0.63 0.67 0.67

7

KS1

SFR 2013-37 (Table 14)

11

KS2

16

KS4

Reading Maths Overall Reading Maths Overall English Maths MFL Overall Overall

At least expected level At least expected level Good level of Development (GLD) Level 2A+ Level 2A+ Average Re + Ma (2A+) Level 4B+ Level 4B+ Level 4B+ in RWM GCSE A*−C pass GCSE A*−C pass GCSE A*−C pass 5 + A*−C Incl. En & Ma EBacc achieved

57 53 55 77 74 64 68.8 71.2 32.3 60.9 22.5

48 46 47 69 72 59 64.6 71.8 47.5 58.3 24.4

0.70 0.76 0.73 0.65 0.90 0.81 0.83 1.03 1.90 0.90 1.11

SFR 2013-51 (Table 8b) SFR 2014-05

Notes: Source = DFE Statistical First Release (SFR) from which the data is drawn. RWM = Reading, writing and mathematics. MFL = Modern foreign Language. Source: Strand et al. (2015: 27).

Literac y Development in Children w ith English as an Addit ional L anguage

47

also interesting differences between reading and mathematics where EAL pupils’ scores in maths are higher than their scores in reading (in Key Stage 2 and Key Stage 4). This gap in performance between reading and mathematics is another reason to focus on literacy development, as discussed below.

Literacy Development of EAL Children As seen above, reading can be a challenge for many children with EAL (and other minority language learners around the world (e.g. Genesee et al., 2006; Verhoeven & Vermeer, 2006)). Solid literacy skills underpin all academic achievement. In the early primary years of education there is a strong focus on developing these literacy skills, but in the later primary and then secondary years, the focus is on using these skills in order to access the curriculum. As eloquently phrased by Chall (1983), by the end of primary school children need to be able to read to learn. If EAL children struggle with literacy, then this can have serious and negative consequences on their academic development. Research around the world, and within the UK, has demonstrated that children with EAL tend to lag behind native-speaking peers on measures of reading comprehension (e.g. Burgoyne et al., 2009, 2011; Hutchinson et al., 2003). It is worth examining more closely, therefore, what factors contribute to reading comprehension and identifying how we can best support the development of these skills.

Decoding skills in minority language learners Many different theories of reading have been proposed but arguably one of the most cited theories is Gough and Tunmer’s (1986) ‘Simple view of reading’. The basic tenet of this model is that the ability to extract meaning from text stems from two main (sets of) variables: word decoding and language comprehension skills. Word decoding is the ability to map phonemes on to graphemes and is often measured by single word and non-word reading tasks. Language comprehension includes listening comprehension skills and vocabulary (semantic) knowledge. A considerable amount of research has supported this conceptualisation of reading (e.g. Gough et al., 1996; Hatcher & Hulme, 1999; Muter et al., 2004). Word decoding skills themselves include a number of other abilities such as phonological and syntactic awareness, working memory and lexical access processes (Jongejan et al., 2007) and difficulties with these skills can have negative consequences on reading comprehension (e.g. Kame’enui & Simmons, 2001). Fortunately, single word decoding is not typically an area of concern for children with EAL. Lesaux et al. (2008) report on a large-scale meta-analysis of reading studies carried out in The Netherlands, Canada, the UK and the USA, and found that the word reading and phonological skills of

48

Par t 1: Over v iews of Research Findings

native-speaking and minority language learners were the same. This finding is also demonstrated in Hutchinson et al. (2003) and Burgoyne et al. (2009, 2011) who showed that there were no differences between EAL children and native-speaking children in England on measures of single word reading. There is credible evidence therefore, that word decoding skills, one of the pillars of the Simple View of Reading model, is not problematic for children with EAL. However, the other main component to this model, namely, vocabulary (semantic) knowledge, does consistently present itself as a challenge for many children with EAL.

Vocabulary knowledge in minority language learners A number of studies carried out in the L1 context have shown the importance of vocabulary knowledge for reading comprehension (e.g. Nation & Snowling, 2004; Nation et al., 2004, 2010). Unfortunately, children with EAL have been consistently shown to have difficulties with vocabulary knowledge (e.g. Bialystok et al., 2010). Within the context of England, Dockrell et al. (2010) have shown that EAL children from a range of different L1 backgrounds aged 3 to 5 had low scores on English language assessments tapping into verbal comprehension, naming vocabulary and sentence repetition. It is precisely this age group (i.e. early years) where we see the largest gaps between EAL and non-EAL pupils in England (see Table 4.1). The gap between EAL and non-EAL students in terms of vocabulary is also evident in older children, even after 10 years of formal education within English. Cameron (2002) compared native-speakers and children with EAL on two different vocabulary measures; Nation’s vocabulary levels test (Nation, 1990) and Meara’s Yes/No task (1992). The children with EAL had lower scores on the vocabulary assessments than the native-speaking children, even with some of the most frequent words on Nation’s (1990) levels test. More recent work has similarly shown the contribution of vocabulary knowledge to reading comprehension and how some children with EAL have lower vocabulary scores than native speakers. Babayiğit (2012) recruited EAL (mixed L1s) and native-speaking primary school students into her study and administrated a range of oral language (vocabulary), listening and reading comprehension assessments. As with other research, she showed that even after four years of formal schooling in England, the pupils with EAL tend to underperform relative to nativespeakers on measures of listening and reading comprehension and oral language (i.e. vocabulary). Furthermore, vocabulary was identified as a significant predictor of performance on reading comprehension tasks, again replicating previous studies. This work underscores the importance of developing vocabulary knowledge in children with EAL. However, one aspect of vocabulary knowledge that has not been properly addressed within this population is multiword phrases. Hence, I would argue, the problems some children with EAL have with vocabulary have not yet been fully explored.

Literac y Development in Children w ith English as an Addit ional L anguage

49

Multiword phrases as part of vocabulary knowledge Multiword phrases (MPWs) can be conceived in different ways (e.g. as idioms, collocations, phrasal verbs, etc.), but one unifying aspect to them is that they function as individual lexical items (Martinez & Schmitt, 2012; Wray, 2002). While much research has identified that MWPs can facilitate other aspects of linguistic knowledge and processing (Wray, 2002; Pawley & Syder, 1983), they have also been shown to pose problems for adult English L2 learners. In Martinez and Murphy (2011), for example, adult EFL students were given two sets of reading tasks aimed at assessing learners’ reading comprehension skills. The critical distinction between the two tasks is that one set of texts was formulated with more MWPs than the other. The actual vocabulary items across the two sets were matched (i.e. exactly the same words were in both sets of texts). The difference was in how these words were put together, where one set of texts contained more MWPs than the other. The adult learners in the Martinez and Murphy (2011) study were asked to answer a number of comprehension questions based on each text, and rate their self-perceived comprehension. There were higher accuracy scores on the comprehension questions for the text that had fewer MWPs. Interestingly, the EFL students also over-estimated their comprehension on the text with more MWPs relative to the text without many MWPs. In other words, while the text with MWPs elicited lower comprehension scores, the adult EFL participants in this study seemed to be unaware of their weaker comprehension of this text. This finding has been replicated in children with EAL in Kan and Murphy (in progress). EAL and non-EAL children in primary school (aged 9 to 11) were compared on the same task as in Martinez and Murphy (2011). The results indicated that both the non-EAL (native-speaking) and EAL children had higher scores on the text with no MWPs relative to the text with MWPs, just as in Martinez and Murphy (2011). Furthermore, the EAL and non-EAL pupils were the same on their reading comprehension scores for the text without MWPs, but the EAL pupils had lower comprehension scores than the non-EAL (native-speaking) children on the text with MWPs (Kan & Murphy, in progress). These findings show us that reading comprehension scores can be negatively influenced by the presence of MWPs in texts by both EFL adults and children with EAL, and even native-speaking children – a particular problem given the fact that MWPs are ubiquitous in discourse (Erman & Warren, 2000). Importantly, there is also some indication that learners may not even be aware of their lack of comprehension of texts which contain MWPs (Martinez & Murphy, 2011). As this area of vocabulary knowledge (MWPs) has not received much attention from researchers, there are very few tests available aimed at tapping into this knowledge. The research described in Smith and Murphy (2015) was an attempt to mitigate against that problem, where the aim was to develop a test of MWPs that would be suitable for use with primary school

50

Par t 1: Over v iews of Research Findings

Figure 4.2 The multiword phrase test (Smith & Murphy, 2015)

children, and then to administer that test to children with EAL and nativespeaking children to identify whether and to what extent there might be differences between these two groups on phrasal vocabulary. A test was developed where the test taker has to fill in the gaps of a target sentence by creating a two or three word multi-word phrase from a 2 × 3 matrix of constituent parts (see Figure 4.2). The target items on this test were verb + object MWPs matched in frequency and composed of comparably frequently occurring components. The test was administered to EAL and native-speaking children in years 3, 4 and 5 (aged 7/8, 8/9 and 9/10) in an effort to identify whether there were any interesting changes over time in children’s performance on this task. There were significant differences between children with EAL and monolingual English speakers on this task by year group. The native-speaking children seemed to develop in a more linear fashion where there were significant increases (i.e. higher scores) on the multiword phrase task (MPT) task across year groups. For the children with EAL, however, there were no differences between years 3 and 4. There were differences, however, between years 5 and 4, suggesting more of a step-function of development for these children. This finding could suggest a different developmental trajectory for MWP development in EAL relative to native-speaking children. Importantly too, the MPT accounted for 25% of the variance in a reading comprehension measure, even after controlling for non-verbal IQ and expressive and receptive vocabulary. These findings show therefore that MWP knowledge might develop differently for children with EAL and that this type of vocabulary knowledge is an important predictor in reading comprehension. McKendry and Murphy (in progress) have shown a similar finding with idioms. In this study children in year 5 (aged 10/11) with EAL in England (with mixed L1 backgrounds) were matched against children with English as a native language on measures of reading comprehension, non-verbal IQ and word decoding. They were administered an idiom comprehension task based on Cain and Towse (2008), which also manipulated whether the idioms were transparent, real (i.e. they were actual idioms in English or made up) and whether they appeared in supporting context or not.

Literac y Development in Children w ith English as an Addit ional L anguage

51

For example, the idiom ‘to have a bark that is worse than his bite’ was presented in two alternative ways: (1) In isolation, where the pupils were asked to choose from a set of four possibilities what the correct meaning was: What does it mean when someone says ‘to have a bark that is worse than his bite.’ — — — —

A: B: C: D:

To bark like a dog at people but never hit them. To shout loudly to encourage people. To make people do difficult things. To make threats but not to carry them out.

(2) The children read a contextualising paragraph and then chose from a set of four possibilities what the meaning of the idiom was: The school football team were training. First, they had to run around the pitch three times. Then the coach shouted that if they did not run faster they would have to run round three more times! The younger players looked quite scared at the thought of even more fast running. ‘Don’t worry about him’ said one of the older players, ‘his bark is worse than his bite.’ What does it mean when someone says ‘to have a bark that is worse than his bite.’ — — — —

A: B: C: D:

To bark like a dog at people but never hit them. To shout loudly to encourage people. To make people do difficult things. To make threats but not to carry them out.

The idioms were either transparent and real (e.g. ‘get away with murder’), transparent and novel (e.g. ‘caught between two fires’), opaque and real (e.g. ‘carry a torch’) or opaque and novel (e.g. ‘the turtle is shrouded’), and as indicated above they were either presented in context or in isolation. The results of the study revealed that children with EAL who had higher scores on reading comprehension measures produced more idiomatic responses (i.e. more likely chose the correct meaning) than those children with EAL with lower reading comprehension scores. Furthermore, more idiomatic responses were provided for transparent relative to opaque idioms, which is not surprising given that the meaning of transparent idioms are more accessible (which is why they are transparent in the first place). Interestingly, the EAL children overall showed a very similar pattern to the native-speakers where they were similarly influenced by context (higher scores when the idiom appeared in context). However, whether the idioms were real or novel had a much larger impact on the native-speaking (NS) students than the EAL pupils where NS pupils had much higher scores for real than novel idioms relative to EAL

52

Par t 1: Over v iews of Research Findings

children. This might be because, for the children with EAL, the idioms were all equally novel. Furthermore, expressive vocabulary seemed to be a stronger influence for the EAL pupils than the native-speaking children, suggesting that the native-speakers draw upon other skills, in addition to vocabulary, when completing the idiom task. Finally, McKendry and Murphy (in progress) also demonstrated that performance on an idiom comprehension measure predicts a significant amount of variance on reading comprehension measures. These findings concerning the relationship between multiword vocabulary (including idioms) in children with EAL indicate that many facets of vocabulary knowledge need to be taken into account when trying to characterise EAL children’s vocabulary knowledge, not just single word items. Furthermore, multiword vocabulary makes significant contributions to reading comprehension (i.e. explains a significant proportion of the variance) in reading comprehension measures. This finding in particular illustrates that it is not just single word vocabulary that is implicated in reading comprehension, but more ‘deeper’ aspects of vocabulary knowledge as well. However, these aspects of vocabulary do not always figure prominently in the classroom for pupils with EAL, highlighting the need to attend to vocabulary in the classroom. Indeed, in their review of interventions aimed at English language and literacy development for EAL pupils, Murphy and Unthiah (2015) showed that those interventions that focussed on some aspects of developing vocabulary knowledge yield positive results on subsequent language and literacy performance. This section has focused on reading comprehension skills of children with EAL and identified that across numerous studies, both within and outside the UK, some children with EAL have been shown to have specific difficulties with extracting meaning from text. This challenge with reading comprehension is contrasted against very well developed skills of single word reading, a consequence of good decoding skills. In other words, children with EAL do not typically struggle in developing strong phoneme-grapheme correspondences, enabling them to perform well (and sometimes better than native speakers) on measures of word reading accuracy. The problems that many children with EAL experience with reading comprehension seem to stem from the language comprehension element of the Simple View of Reading model. Indeed, listening comprehension has also been shown to be a challenge for some children with EAL (McKendry & Murphy, 2011) and numerous studies around the world have demonstrated that minority language learners tend to have smaller vocabularies than native-speaking children (e.g. Bialystok et al., 2010; Cameron, 2002). Research in reading comprehension in L1 children has consistently demonstrated the important contribution of vocabulary to skilled comprehension and the same is true for children with EAL as well. However, as the studies above have demonstrated, vocabulary does not just come in single discrete items, but can also include

Literac y Development in Children w ith English as an Addit ional L anguage

53

multiword phrases which are often neglected in classrooms. These multiword phrases constitute another dimension along which children with EAL are often different from their native-speaking peers. Given that the research above has shown that this aspect of vocabulary knowledge itself makes significant contributions to reading comprehension it would be worthwhile remembering to include them in vocabulary teaching activities.

Writing There has been far less research on writing skills in children with EAL than reading and oral language development. This gap in our understanding of the writing skills of pupils with EAL is a particular problem in light of the lags in academic achievement identified in the international student assessment studies described above, as well as the UK-based work showing achievement gaps (Strand et al., 2015). Writing is central for academic achievement and is the primary means by which children can show their understanding of different academic knowledge (Dockrell et al., 2014). Furthermore, writing is a challenge for many pupils (regardless of their linguistic background), where teachers often report finding effective instruction in writing difficult and assessing writing is also replete with complications (Dockrell et al., 2014). It is surprising, therefore, that writing research has received comparatively less attention than reading. As with reading, writing is considered to consist of multiple components, including working memory, transcription skills (e.g. handwriting, spelling) and executive functions that support the production of text (Berninger & Amtmann, 2003; Berninger & Winn, 2006). Writing takes time to develop and is constrained in the early stages by the lower-level processes of transcription. Once these processes become automatised the pupil can then develop more higher level skills of generating content and then organising (and articulating) content appropriately in text. There are also reciprocal relationships between the development of reading and writing instruction. Graham and Hebert (2011) for example, in a meta-analysis of research examining the relationship between writing instruction on reading, identified that strategies that required primary school pupils to write about what they read enhances their comprehension skills illustrating that teaching writing has a positive influence on improving reading. Furthermore, increasing the amount students are asked to write can also support reading skills in primary school pupils (Graham & Hebert, 2011). With such mutually supportive relationships between writing and reading, it would be fruitful to enhance writing instruction for children with EAL (and indeed all children). As indicated above, there is very little research on specific processes involved in EAL children’s writing and whether and to what extent children with EAL have different profiles of writing than native-speaking children. We do have enough of an indication from research, however, to suggest that EAL

54

Par t 1: Over v iews of Research Findings

pupils lag behind their non-EAL peers on different aspects of writing. Cameron and Besser (2004) carried out a comparison of EAL and non-EAL pupils (in England) on two compositions (fiction and persuasive). They found that the achievement gap between the two groups of pupils was nine percentage points on the national writing test Standard Attainment Tests (SATs). The EAL pupils made more grammatical mistakes and were less likely to use complex syntactic structures than the non-EAL students. Babayiğit (2015) has also examined EAL pupils’ writing. In her study, EAL and non-EAL pupils in year 5 (aged 10/11) in England were given a standardised assessment of writing requiring the children to write two paragraphs in response to a prompt (e.g. ‘my favourite game is…’). She found no differences between the two groups on spelling (replicating previous work), yet found that the nativespeaking English students scored higher on the written expression tasks than EAL pupils on measures of holistic quality, organisation, vocabulary and compositional fluency. A similar result was found in the comparison between EAL and non-EAL pupils’ writing in Murphy et al. (in progress). In this study, year 5 (aged 10/11) pupils in England were asked to complete a narrative writing task. The EAL and non-EAL pupils were matched on language age (as determined by the Clinical Evaluation of Linguistic Fundamentals (CELF). In other words, the two groups of children had similar levels of English vocabulary and syntactic skills. However, in matching the groups on these English measures, the EAL pupils ended up having higher nonverbal IQ scores. Despite this potential advantage, the EAL pupils still had lower scores than the nonEAL pupils on higher-level writing processes (such as organisation and ideas). Taken together, the few currently available studies examining writing skills in children with EAL show that, just as in their reading, they tend to lag behind their native-speaking peers on key aspects (higher level) features of writing. It is precisely these features that can lead to reduced performance on academic achievement. There is an urgent need, therefore, for considerably more work in this area, not just with children with EAL in the UK, but around the world. In summary, this section on the development of literacy skills in children with EAL has shown that while the EAL population is heterogeneous, there are clear trends showing that many children with EAL struggle with reading comprehension and higher level aspects of writing. There is strong evidence to show that vocabulary knowledge is a significant factor for both literacy skills.

The Importance of Multiple Methods in Research Much of the research described in the above sections adopted quantitative methodologies to focus on comparisons between children with EAL and NS peers. Other research, however, has adopted more qualitative methodologies,

Literac y Development in Children w ith English as an Addit ional L anguage

55

which offers a much needed complementary approach to the research described above given that using different methodologies allows researchers to focus on different types of questions. With respect to literacy development in EAL children, Gregory (2008) is one example of how adopting more qualitative approaches can lead to the development of important insights into the nature of the complexity of biliteracy development. Set within a Vygotskian, sociocultural framework, Gregory adopts qualitative methodologies including observation, interviews with mainstream and community teachers, taped reading interactions and interactions with parents, to provide a framework for observing, teaching and assessing young children who are learning to read in English, when their home language is not English. A major aim is to use these methodologies to help identify different strategies that children use in classrooms, which in turn can be used to help guide teachers in their practice. These methodologies allow Gregory (and other researchers) to ask important questions that have not been the focus of this chapter, such as the nature and influence of literacy practices in homes and communities, and how these might influence the development of English literacy skills for the EAL pupil in the classroom. Gregory (2008) has been able to show that while much of the research above views reading and writing as the development of a cognitive skill, it can also be conceived of as developing something beyond that, a set of reading practices that might be viewed very differently depending on cultural norms and how children bring these home literacy practices with them to school which may or may not be consistent with those adopted by the class teacher. Without more qualitative methodologies of observation, interviews and detailed (ethnographic) case studies, these important aspects of literacy development would remain misunderstood. In addition to understanding the social and cultural context through use of qualitative methodologies such as in Gregory (2008), other researchers have used these methods to investigate specific classroom practice. For example, Hornberger and Link (2012) report how a Spanish-speaking primary school student (six years of age) uses both Spanish and English in guided reading tasks to help develop and expand her communicative repertoire. By adopting methodologies which allow for specific and focussed observations of classroom and school behaviour, Hornberger and Link provide an interesting description of this one child’s use of two languages within school, and through these observations, illustrate how, when moving across two languages fluidly (i.e. ‘translanguaging’ see below), the child learning EAL can make meaning and communicate across different types of interactions, allowing for more engagement with language learning and teaching in both home and school contexts. The focus of this chapter has been on research which has adopted quantitative methodologies, but these two brief descriptions of Gregory (2008) and Hornberger and Link (2012) serve to both illustrate and remind the

56

Par t 1: Over v iews of Research Findings

reader that different methodologies allow for the investigation of different types of research questions – and both methodologies offer important insights into how children develop literacy skills in second languages. It seems possible that the investigation of some contexts might benefit from a mixed methods approach to provide a more comprehensive account of children’s EAL literacy development.

Implications This chapter has focused on a specific group of learners with English as an additional language. These are pupils who are being educated through the medium of English, in an English speaking country, but who have a home language that is not English. The research described in this chapter has indicated that for some (indeed many) children with EAL, developing well-developed literacy skills can prove to be a significant challenge. Interestingly, they do not seem to struggle with so-called lower level skills. Children with EAL have been consistently shown to have very well developed phonological awareness skills, which lead to excellent word decoding (single word reading) and spelling skills. However, a less well-developed English lexicon leads to problems in reading comprehension and also contributes to difficulties in writing. One of the ways some countries have attempted to support children with EAL is to develop bilingual education programmes to enhance their knowledge of the L1 (e.g. USA). Cummins (2000 and elsewhere) has shown that supporting L1 literacy skills can mutually support the development of their L2 (English). Hence the dual immersion programmes reviewed in Murphy (2014) have proven effective for EAL (or ELL pupils as they are known in the USA). This type of bilingual education programme works well in contexts like the USA where there are large proportions of ELL (EAL) pupils with the same L1 background (e.g. Spanish). However, a context such as England has been referred to as ‘super-diverse’ in relation to its linguistic makeup (Blommaert & Rampton, 2012; Vertovec, 2007). It is challenging at best to conceive of how one might develop a bilingual education programme when there might be more than 10 different L1s represented in the student population of a given school or class. One of the ways researchers are suggesting one can support students’ learning in this context is through translanguaging. García and Li Wei (2013) suggest that ‘translanguaging does not refer to two separate languages nor to a synthesis of different language practices or to a hybrid mixture. Rather translanguaging refers to new language practices that make visible the complexity of language exchanges among people with different histories’ (García & Li Wei, 2013: 21). Through translanguaging in the classroom, teachers can facilitate students’ use of their home languages in classroom-based interaction, in line with Cummins’ (2000) suggestions. There are some issues with respect to translanguaging, however, that require further exploration. One relates to

Literac y Development in Children w ith English as an Addit ional L anguage

57

the extent to which translanguaging is a separate construct from codeswitching and whether (assuming it is), it offers new and generaliseable theoretical constructs than previous conceptualisations of the use of two codes (codemixing, codeswitching) in the classroom. Recent discussions suggest it may not (Bhatt & Bolonyai, 2015). Furthermore, in order to unambiguously support translanguaging (assuming it is different from codeswitching), ideally it would be subjected to some form of quasi-experimental design where classes that facilitate translanguaging are compared against those which do not on some form of linguistic or indeed academic achievement outcome. Such empirical evidence has not yet been made available. Nonetheless, whatever it is called, facilitating the use of a students’ L1 in the classroom has for some time been considered an effective pedagogical and linguistic strategy to support minority language learners’ linguistic and academic development, and further research will be very useful in this area to provide clear(er) guidance on how best to support this practice. Other issues to consider within the context of this chapter relate to Initial Teacher Education (ITE) and Continuing Professional Development (CPD). In England, it is not possible for teachers to specialise in EAL pedagogy, and ITE programmes that have a very strong component on EAL learners and pedagogy are few and far between. This is at odds with the fact that the proportion of children with EAL in English schools is ever increasing. It would be most useful, therefore, if ITE programmes were developed to better equip future generations of teachers to accommodate the needs of all their pupils. Additionally, developing high quality (i.e. evidence-based) CPD for teachers aimed specifically at how to support EAL learners is required. Funding, too, is an issue. In England, the Ethnic Minority Achievement Grant (EMAG) was a ring-fenced pool of funds from which schools could draw on to support their ethnic minority pupils. This ring-fence was dismantled in 2012 and now less money is available to schools and teachers to support their ethnic minority pupils. These decisions are in conflict with the increasing proportion of children with EAL in English schools. Finally, the issue of assessment is of paramount importance here, since in the UK and the USA most children with EAL (ELL) are tested on the same high-stakes standardised tests, normreferenced against a monolingual population. Menken et al. (2014) have highlighted some of the difficulties with this approach and suggest that tangible validity problems arise when tests such as these are used for purposes beyond their intention (i.e. with pupils like EAL).

Conclusions In summary, while children with EAL constitute a heterogeneous group where many EAL pupils are matched or indeed even outperform their nativespeaking peers, numerous studies have demonstrated a clear trend for children

58

Par t 1: Over v iews of Research Findings

with EAL to underperform in academic outcomes. This lag has been attributed to a range of factors, but arguably the most significant one is weakened literacy skills relative to native-speaking pupils. Vocabulary (oral language) is a very strong predictor of literacy development and children with EAL have been shown to consistently lag behind their non-EAL peers on measures of vocabulary, which in turn has a negative influence on their developing literacy skills in their L2 (English). With increased migration and globalisation of commerce, the proportion of children with EAL in England, and indeed other minority language learners around the world, is only going to rise. Much work needs to be done, therefore, to ensure we meet our responsibilities to all of our societies’ children in meeting their full linguistic and academic potential.

References Babayiğit, S. (2012) The role of oral language skills in reading and listening comprehension of text: A comparison of monolingual (L1) and bilingual (L2) speakers of English language. Journal of Research in Reading 37 (1), 22–47. Babayiğit, S. (2015) The dimensions of written expression: Language group and gender differences. Learning and Instruction 35, 33–41. Berninger, V.W. and Amtmann, D. (2003) Preventing written expression disabilities through early and continuing assessment and intervention for handwriting and/or spelling problems: Research into practice. In H.L. Swanson, K. Harris and S. Graham (eds) Handbook of Learning Difficulties (pp. 345–363). New York: Guildford Press. Berninger, V.W. and Winn, W.D. (2006) Implications of advancements in brain research and technology for writing development, writing instruction, and educational evolution. In C. MacArthur, S. Graham and J. Fitzgerald (eds) Handbook of Writing Research (pp. 96–114). New York: Guildford Press. Bhatt, R. and Bolonyai, A. (2015) On the theoretical and empirical bases of translanguaging. Paper presented to the International Symposium of Bilingualism, 20th–24th May, Rutgers University: New Jersey. Bialystok, E., Luk, G., Peets, K.F. and Yang, S. (2010) Receptive vocabulary differences in monolingual and bilingual children. Bilingualism: Language and Cognition 13 (4), 525–531. Blommaert, J. and Rampton, B. (2012) Language and superdiversity. Max Planck Institute for the study of Religious and Ethnic Diversity Göttingen. Working Paper 12-09. Burgoyne, K., Kelly, J.M., Whiteley, H.E. and Spooner, A. (2009) The comprehension skills of children learning English as an additional language. British Journal of Educational Psychology 79, 735–747. Burgoyne, K., Whiteley, H.E. and Hutchinson, J.M. (2011) The development of comprehension and reading-related skills in children learning English as an additional language and their monolingual English-speaking peers. British Journal of Educational Psychology 81, 344–354. Cain, K. and Towse, A.S. (2008) To get hold of the wrong end of the stick: Reasons for poor idiom understanding in children with reading comprehension difficulties. Journal of Speech, Language and Hearing Research 51 (6), 1538–1549. Cameron, L. (2002) Measuring vocabulary size in English as an Additional Language. Language Teaching Research 6 (2), 145–173. Cameron, L. and Besser, S. (2004) Writing in English as an additional language at Key Stage 2 (Rep. No. 586). See http://www.naldic.org.uk/Resources/NALDIC/ Research%20and%20Information/Documents/RR586.pdf (accessed 1 October 2015).

Literac y Development in Children w ith English as an Addit ional L anguage

59

Chall, J. (1983) Learning to Read: The Great Debate. Wiley: New York. Cummins, J. (2000) Language, Power and Pedagogy: Bilingual Children in the Crossfire. Clevedon: Multilingual Matters. Cummins, J. (2012) The intersection of cognitive and sociocultural factors in the development of reading comprehension among immigrant students. Reading and Writing 25, 1973–1990. de Bot, K. and Gorter, D. (2005) A European perspective on heritage languages. The Modern Language Journal 89 (4), 612–616. Dockrell, J.E., Stuart, M. and King, D. (2010) Supporting early oral language skills for English language learners in inner city preschool provision. British Journal of Educational Psychology 80, 497–516. Dockrell, J.E., Connelly, V., Walter, K. and Critten, S. (2014) Assessing children’s writing products: The role of curriculum based measures. British Educational Research Journal 41 (4), 575–595. Erman, B. and Warren, B. (2000) The idiom principle and the open choice principle. Text – Interdisciplinary Journal for the Study of Discourse 20 (1), 29–62. García, O. and Wei, L. (2013) Translanguaging: Language, Bilingualism and Education. London: Palgrave. Genesee, F., Lindholm-Leary, K., Saunders, W.M. and Christian, D. (eds) (2006) Educating English Language Learners: A Synthesis of Research Evidence. Cambridge: Cambridge University Press. Gough, P. and Tunmer, W. (1986) Decoding, reading and reading disability. Remedial and Special Education 7, 6–10. Gough, P., Hoover, W. and Patterson, C. (1996) Some observations on a simple view of reading. In C. Cornoldi and J. Oakhill (eds) Reading Comprehension Difficulties: Processes and Intervention (pp. 1–13). Mahway, NJ: Lawrence Erlbaum Associates. Graham, S. and Hebert, M. (2011) Writing to read: A meta-analysis of the impact of writing and writing instruction on reading. Harvard Educational Review 81 (4), 710–744. Gregory, E. (2008) Learning to Read in a New Language: Making Sense of Words and Worlds (2nd edn). London: Sage. Gregory, E. and Williams, A. (2000) City Literacies: Learning to Read across Generations and Cultures. London: Routledge. Hatcher, P.J. and Hulme, C. (1999) Phonemes, rhymes and intelligence as predictors of children’s responsiveness to remedial reading instruction: Evidence from a longitudinal intervention study. Journal of Experimental Child Psychology 72, 130–153. Hornberger, N.H. and Link, H. (2012) Translanguaging in today’s classrooms: A biliteracy lens. Theory Into Practice 51, 239–247. Hutchinson, J.M., Whiteley, H.E., Smith, C.D. and Connors, L. (2003) The developmental progression of comprehension-related skills in children learning EAL. Journal of Research in Reading 26 (1), 19–32. Jongejan, W., Verhoeven, L. and Siegel, L.S. (2007) Predictors of reading and spelling abilities in first- and second-language learners. Journal of Educational Psychology 99 (4), 835–851. Jonsson, C. (2013) Translanguaging in and multilingual literacies: Diary-based case studies of adolescents in an international school. International Journal of Sociology of Language 224, 85–117. Kan, R.T.Y. and Murphy, V.A. (in progress). The influence of frequency and idiomaticity on second language reading comprehension in children. Kame’enui, E.J. and Simmons, D.C. (2001) Introduction to this special issue: The DNA of reading fluency. Scientific Studies in Reading 5 (3), 203–210. Lesaux, N.K., Geva, E., Koda, K., Siegel, L.S. and Shanahan, T. (2008) Development of literacy in second-language learners. In D. August and T. Shanahan (eds) Developing

60

Par t 1: Over v iews of Research Findings

Reading and Writing in Second-Language Learners: Lessons from the Report of The National Literacy Panel on Language-Minority Children and Youth (pp. 27–59). New York: Routledge/CALL. Martinez, R. and Murphy, V. (2011) The effect of frequency and idiomaticity on second language reading comprehension. TESOL Quarterly 45, 267–290. Martinez, R. and Schmitt, N. (2012) A phrasal expressions list. Applied linguistics 33, 299–320. McKendry, M.G. and Murphy, V.A. (2011) A comparative study of listening comprehension measures in English as an additional language and native English-speaking primary school children. Evaluation and Research in Education 24 (1), 17–40. McKendry, M.G. and Murphy, V.A. (in progress) Idiom comprehension in children with English as an Additional Language. Meara, P. (1992) EFL Vocabulary Tests. Swansea: Centre for Applied Language Studies. Menken, K., Hudson, T. and Leung, C. (2014) Symposium: Language assessment in standards-based education reform. TESOL Quarterly 48 (3), 586–614. Murphy, V.A. (2014) Second Language Learning in the Early School Years: Trends and Contexts. Oxford: Oxford University Press. Murphy, V.A. and Unthiah, A. (2015) A Systematic Review of Intervention Research Examining English Language and Literacy Development in Children with English as an Additional Language (EAL). Research Report for the Education Endowment Foundation, Bell Foundation and Unbound Philanthropy: London. Murphy, V.A., Kyriacou, M. and Menon, P. (in progress) Profiling writing challenges in children with English as an Additional language (EAL). Muter, V., Hulme, C., Snowling, M.J. and Stevenson, J. (2004) Phonemes, rimes and language skills as foundations of early reading development: Evidence from a longitudinal study. Developmental Psychology 40, 663–681. NALDIC (2014) EAL Statistics. See http://www.naldic.org.uk/research-and-information/ eal-statistics/eal-pupils Nation, P. (1990) Teaching and Learning Vocabulary. Boston, MA: Heinle and Heinle. Nation, K. and Snowling, M. (2004) Beyond phonological skills: Broader language skills contribute to the development of reading. Journal of Research in Reading 27, 342 –356. Nation, K., Clarke, P., Marshall, C.M. and Durand, M. (2004) Hidden language impairments in children: Parallels between poor reading comprehension and specific language impairment. Journal of Speech, Hearing and Language Research 47, 199–211. Nation, K., Cocksey, J., Taylor, J.S.H. and Bishop, D. (2010) A longitudinal investigation of early reading and language skills in children with poor reading comprehension. Journal of Child Psychology and Psychiatry 51 (9), 1031–1039. OECD (2012) OECD Family Database: Percentage of Immigrant Children and their Outcomes. OECD, Paris. See: http://www.oecd.org/els/soc/49295179.pdf (accessed 13 March 2017). OECD (2013) PISA 2012 Results: Excellence through Equity: Giving Every Student the Chance to Succeed (Vol. II). PISA, Paris: OECD Publishing. http://dx.doi.org/10.1787/978926 4201132-en Paradis, J., Genesee, F. and Crago, M. (2011) Dual Language Development and Disorders: A Handbook on Bilingualism and Second Language Learning (2nd edn). Baltimore, MD: Brookes Publishing. Pawley, A. and Syder, F. (1983) Two puzzles for linguistic theory: Nativelike selection and nativelike fluency. In J.C. Richards and R.W. Schmidt (eds) Language and Communication (pp. 191–226). London: Longman. PISA (2003) https://www.oecd.org/edu/school/programmeforinternationalstudentassessmentpisa/34002454.pdf PISA (2006) http://www.oecd.org/edu/school/programmeforinternationalstudentassessmentpisa/pisa2006results.htm

Literac y Development in Children w ith English as an Addit ional L anguage

61

PISA (2009) http://www.oecd.org/pisa/pisaproducts/pisa2009keyfindings.htm PISA (2012) http://www.oecd.org/pisa/keyfindings/pisa-2012-results.htm Smith, S.A. and Murphy, V.A. (2015) Measuring productive elements of multi-word phrase vocabulary knowledge among children with English as an additional or only language. Reading and Writing 28, 347–369. Strand, S., Malmberg, L. and Hall, J. (2015) England as an Additional Language (EAL) and Educational Achievement in England: An Analysis of the National Pupil Database. London: Education Endowment Foundation. See https://educationendowmentfoundation.org. uk/uploads/pdf/EAL_and_educational_achievement2.pdf (accessed 13 March 2017). Verhoeven, L. and Vermeer, A. (2006) Literacy achievement of children with intellectual disabilities and differing linguistic backgrounds. Journal of Intellectual Disability Research 50 (10), 725–738. Vertovec, S. (2007) Super-diversity and its implications. Ethnic and Racial Studies 30, 1024–1054. Wray, A. (2002) Formulaic Language and the Lexicon. Cambridge: Cambridge University Press.

Part 2 Empirical Studies Using Mixed Methods

5

Verbal Working Memory and Foreign Language Learning in English Primary Schools: Implications for Teaching and Learning Alison Porter

Introduction Variable outcomes in adult instructed language learning are widely accepted (Kiss, 2009), but child foreign language instructed learning is often assumed to be more homogeneous (Mihaljević Djigunović, 2009). Empirical evidence shows that this is unlikely to be the case, instead demonstrating considerable variation in young learner linguistic outcomes across a range of contextual factors; for example, nature of provision and teacher expertise (Murphy, 2014) and individual characteristics such as age (Johnstone, 2002; Muñoz, 2006), language learning strategies (Cameron, 2001), first language (L1) literacy skills (Sparks et al., 2009), socioeconomic status (Mattheoudakis & Alexiou, 2009) and aptitude in the form of memory and analytic skills (Alexiou, 2009). These kinds of differences create heterogeneous early FL learning classrooms which present a considerable challenge for teachers (Nikolov & Mihaljević Djigunović, 2011), especially as foreign language (FL) policy and practice is often grounded in certain assumptions about younger children learning languages: first, that younger children are better at learning languages due to brain plasticity conceptualised in the Critical Period Hypothesis (Lenneberg, 1967) and second that a younger start will increase exposure to linguistic input over a longer time span, resulting in enhanced linguistic outcomes. Consequently, research that identifies and explores the implications of particular individual differences in young FL classrooms could be of interest 65

66

Par t 2: Empir ical Studies Using Mixed Methods

to both policy-makers and practitioners. Working memory is an important individual difference in adult second language (L2) learning and has recently been linked to language learning aptitude (Wen, 2011). Evidence is emerging that verbal working memory (VWM), a sub-component of working memory, is important in both adult and child L2/FL learning. This mixed-methods study aims to contribute to the emerging evidence base reflecting individual differences in young FL learning by investigating the complex role of VWM in the development of FL general proficiency and FL literacy skills. Both quantitative and qualitative data in the form of FL test results and weekly written vocabulary work are presented and analysed in order to construct a nuanced portrayal of VWM as an individual difference in FL learning. The study finds that VWM has considerable influence on learning outcomes in both FL oracy and literacy and therefore suggests that young FL classrooms are complex, heterogeneous contexts. Potential implications for young learner FL pedagogy are also considered.

Theoretical Background Working memory The most widely adopted working memory model is believed to consist of a central executive, a controlling attentional system, that supervises and controls other ‘slave systems’, a term used by Baddeley and Hitch (1974) to indicate the phonological loop (concerned with speech information) and the visuo-spatial sketchpad (visual images) (Baddeley, 1986, 1997; Baddeley & Hitch, 1974). These are supported in later working memory models by an episodic buffer, a limited capacity system which supports the binding together of visuo-spatial and phonological representations into a single, ‘episodic (…) multi-faceted code’ (Baddeley, 2003: 203). Working memory has a complex role in L2 and FL learning, but is generally accepted as contributing to the process through both central executive type functions and, more specifically, the processing and temporary storage of phonology. In its executive role, working memory is viewed as a ‘limited capacity system that stores and processes information simultaneously in real time’ (Hummel & French, 2010: 372) and is therefore believed to be an important factor in complex cognitive tasks. For example, it is deemed fundamental in reading acquisition due to the multiple processes involved in reading, such as the encoding and retrieval of phonological, orthographic and semantic information (Baddeley et al., 1998; Pae & Sevcik, 2011: 49). Equally, Révész (2012: 101) suggests that the working memory executive ‘controls complex cognitive operations’ and concerns the ability to ‘divide and switch attention’, to activate and inhibit processing routes and to regulate the flow of information from short-term and long-term memory systems. The conceptualisation

Verbal Work ing Memor y and FL Lear ning in English Pr imar y School s

67

and categorisation of new information, as well as the building of associations between new and previously stored information is also believed to be a function of the executive aspect of working memory (Pae & Sevcik, 2011: 47). Meanwhile, the crucial sub-component of working memory, referred to by Baddeley (2003) as the phonological loop but also known as verbal working memory (VWM), verbal short-term memory, phonological working memory, phonological short-term memory and phonological memory (Hummel & French, 2010: 373), is largely involved with the temporary storage and rehearsal of phonology (Pae & Sevcik, 2011: 48; Swanson et al., 2006: 247–248), with ‘active manipulation’ of this information during storage being handled by working memory (Swanson et al., 2006: 247–248). The phonological loop, then, functions as a ‘sub-set’ of working memory performance dealing with initial verbal input which can then be further manipulated (processed and referred to pre-stored information) by working memory (Pae & Sevcik, 2011: 48). However, distinctions between both these types of working memory vary, for Miller and Kupfermann (2009: 13–14), VWM concerns the ability to temporarily store and manipulate verbal or written input whilst processing incoming information and retrieving existing phonological information from the long term lexicon. Regardless of variation in conceptualisations of facets of working memory, the idea that it consists of separable constructs is supported by language learning research, which has shown no intra-individual correlation between scores on non-word repetition and backward digit span (Kormos & Sáfár, 2008: 267).

Working memory and second language/foreign language outcomes L2 research has shown verbal working memory is linked to the development of L2 general proficiency. In a study conducted by French (2006), VWM was deemed to be influential in both productive and receptive vocabulary learning (English L2). VWM also contributed to the development of English L2 morphosyntactic knowledge (e.g. tense, aspect, inflections, negation, question formation) after a five month intensive English as a second language (ESL) programme (French & O’Brien, 2008). Another study, however, reported no correlation between working memory measured by word span and L2 vocabulary learning measured by word recognition (Akamatsu, 2008). In addition to L2 general proficiency, VWM has been found to influence other L2 measures, but exact inter-relationships between L2 learning and VWM (including working memory) appear to be complex. For example, Révész (2012) posits that in Hungarian adult beginner learners of English, different components of working memory influenced particular language sub-skills. VWM, measured by digit span and non-word repetition, was linked to oral performance ability while ‘complex working memory capacity’, exemplified by reading span score was more likely to predict written test scores including grammaticality judgements (Révész, 2012: 122–123).

68

Par t 2: Empir ical Studies Using Mixed Methods

Backward digit span scores (representative of complex working memory) also correlated with L2 general proficiency measures of reading, writing, listening and speaking, whilst phonological short-term memory (VWM) did not (Kormos & Sáfár, 2008: 267). Findings were linked to the ability to marshal attention, believed to be central in noticing and therefore important for learning L2 words and grammar (Kormos & Sáfár, 2008: 268). Finally, Sunderman and Kroll (2009) found that their measure of ‘complex’ working memory, measured by a reading span task, also tapped into the ability to control attention and suppress the L1 which, in a study abroad context, lead to improved performance in L2 production accuracy. These findings have been noted by others who accord an additional ‘sifting’ function to working memory which is linked to conscious learning (Mitchell et al., 2013: 157) and includes ideas related to ‘noticing’ (conscious awareness), attention or ‘resistance to distraction’ (Juffs & Harrington, 2011: 160; Robinson et al., 2013) and the ability for non-bilingual participants to inhibit L1 competing information when producing unfamiliar FL words (Trude & Tokowicz, 2011). Finally, there is potential for working memory in general to inhibit L1 (and to a lesser extent) L2 competing information (Gass & Lee, 2011: 76). Four studies of particular interest support a role for both kinds of working memory in both child L2 and FL learning. Swanson et al. (2006) found that VWM (digit span) and working memory, operationalised as simultaneous storage and processing, predicted both level and rate of growth in bilingual children’s L2 reading skills (Spanish L1/English L2 – mean age 6.21 years). Service (1992) found that FL reading comprehension, listening comprehension and writing in beginner Finnish English as a foreign language (EFL) learners aged between 9 and 10 were influenced by VWM, whilst Service and Kohonen (1995) demonstrated a role for VWM in FL reading comprehension and vocabulary acquisition in Finnish EFL learners aged 9 to 12. Finally, VWM was linked to L2/FL (English) listening comprehension and vocabulary acquisition in Finnish 9 to 10 year olds (Dufva & Voeten, 1999). Interestingly though, as the ability to repeat multisyllabic non-words (i.e. VWM) has been found to vary greatly at inter-individual levels during childhood (Gathercole, 2006: 513–515), it seems possible that the role of VWM in younger FL learning has distinct implications for primary FL pedagogy. In summary, there is a range of evidence that supports a relationship between working memory and/or VWM and second language learning, but attempts to clarify the exact nature of this link are, to some extent, clouded by differentiation between types of working memory, how these might be isolated in test situations and the exact language sub-skills each facet of memory could influence. Finally, despite evidence for the influence of VWM in second language learning, it is important to note that its influence tends to decline as a result of increases in L2/FL proficiency (French, 2006: 125– 127; Hummel, 2009). It is therefore argued that phonological memory

Verbal Work ing Memor y and FL Lear ning in English Pr imar y School s

69

(VWM) tests are robust predictors only of early language (lexical) development (Juffs & Harrington, 2011: 158), but also that growth in L2 vocabulary size reduces the effect of phonological memory in L2 learning and may even be the ‘catalyst’ for a change in processing techniques (French, 2006: 127). This chapter will reflect on the following research questions: (1) Does VWM influence the development of FL literacy (measured by FL reading aloud and FL reading comprehension) in young beginner FL classrooms? (2) Does VWM influence the development of FL general proficiency (measured by FL receptive vocabulary and FL elicited imitation) in young beginner FL classrooms?

Methods The data for this small-scale study, which explored the role of VWM in younger FL learning, were collected over a period of 30 weeks in total. During this period of investigation, a principled pedagogy to simultaneously teach all four skills in French was practised over 23 weeks (outlined in the next section). In order to measure progression, FL tests were conducted at the start of the school year, at the end of the teaching programme when FL school instruction ceased (23 weeks later) and towards the end of the school academic year (at 30 weeks) – seven weeks after cessation of FL lessons. The influence of VWM as an individual difference on FL learning outcomes was also explored using data collected in September, which measured each child’s VWM score through non-word repetition (to be explored in more detail in the ‘Data collection’ section).

The principled pedagogy The first principle involved simultaneous oracy and literacy teaching, aiming to challenge current instructional advice that often recommends presenting the written word last (Jones & Coffey, 2006: 50). Its premise was grounded in re-conceptualisations of L2 literacy as a ‘repeated process’ (Koda, 2008: 74) and that, as a result of L1 literacy development, some readingrelated skills, for example phonological awareness, could be available to beginner FL/L2 learners. Nevertheless, it also recognised that ‘language distance’, such as differences in phonology and orthography, could affect crosslinguistic influence between developing L1 and L2 literacies (Genesee et al., 2008: 72–73) and have been found (through cognate distance) to strongly predict listening and reading comprehension scores in instructed FL learners aged 10 to 11 years (Lindgren & Muñoz, 2013). In practice this meant that, wherever possible, each lesson was sub-divided into an oracy and a literacy

70

Par t 2: Empir ical Studies Using Mixed Methods

component. In addition, the children learned and practised core vocabulary in both spoken and written form simultaneously. The next principle related to a focus on FL sounds that acknowledged the importance of phonological processing in reading acquisition (Goswami, 1999; Koda, 1992) and that L2 phonemic awareness is influential in L2 reading (e.g. Walter, 2008). So, in most lessons, there was an activity which focused on either production and/or discrimination of FL sounds, for example through identifying and practising the articulatory properties of novel FL sounds. Voice recorders were occasionally used for children to self- and peerassess sound production. The third principle proposed systematic and explicit phonics instruction. This built on existing research that called for FL phonics instruction to improve word decoding (e.g. Cable et al., 2010; Woore, 2007) and L1 reading instruction research that notes the importance of developing word recognition (e.g. Adams, 1990). Classroom activities involved practising linking FL sound/spelling links through games and word reading opportunities. Finally, the fourth principle allowed for experiencing FL sound and print through a range of meaning-focused activities such as stories, extended writing and spontaneous speech. This principle leant on both the consensus that learning to read should be integrated (phonics and text work) (e.g. Adams, 1990) and decades of language learning evidence that called for ‘active (language use) (…) for purposes that matter’ (Hawkins, 1996: 30) and creativity, imagination and risk-taking in FL curricula (Mitchell, 2002).

Participants The research was conducted in two mixed-year classes within two small rural primary schools in south-east England (number on roll in each school 0.05), ratings for all switches, including those not involving English –0.114 (p > 0.05). At least for the small sample presented here, intensity of open micro-switches seems not to correlate significantly with story comprehension.

Discussion We saw that affordances for receptive codeswitching presented in the language learning environment via the trilingual picture book software were associated with higher individual student comprehension outcomes in Study 1. This positive effect, though, seems to be difficult to pinpoint to specific behaviours. Having access to translations is beneficial, yet total switches, macro-switches, overall micro-switches (see Study 1), focused microswitches and open micro-switches (see Study 2) on their own do not seem to explain this beneficial effect.

Codesw itching Your Way to L anguage Lear ning?

121

The benefits that accrue to learners through access to the trilingual digital environment may not be due to the potential for receptive codeswitching itself. Potentially, the trilingual environment encourages learners to use all their linguistic resources for decoding the text. Possibly, students were more active in negotiation with their partner. As switching languages often involved discussing the need for receptive codeswitching due to noncomprehension, an increased communication about comprehension and noncomprehension and deriving from this a higher level of negotiation of meaning and/or form, might be the mechanism through which the trilingual treatment was beneficial for comprehension. Regarding the choice of data, logfile analysis and video ratings provide quite different glimpses into the strategies used by learners in tackling the story. Yet, neither the different measures derived via logfile analysis, nor the video ratings of codeswitching correlate with story comprehension measures. Logfile measures are ‘objective’, as long as there exists agreement about the formal definitions used as basis for quantifying these measures. Ratings, on the other hand, require human raters, who may exhibit imperfect concentration and personal bias, and are liable to misinterpret data. At the same time, they can take things into consideration that an automated logfile analysis cannot. Raters can – and were instructed to – hypothesise about the intentions of learners. Do they seem to be moved by curiosity and an urge to explore, as attested by laughter, and by discourse focused on experiencing different versions of the story? Or did they switch explicitly in order to understand, perhaps even as a response in a negotiation sequence? For a logfile analysis, a button clicked is a button clicked, while for a rater, there can be a distinction between ‘I don’t understand – switch to German’ (focused switch), ‘Come on, let’s listen to the Turkish version! Hihhihihi!’ (open switch) and ‘Oops, I accidentally clicked the wrong button. Wait a sec, I switch back.’ (micro-switch, neither focused nor open). The fact that both types of measures did not seem to directly impact comprehension makes it more plausible that this, at first sight, surprising result is not just a consequence of invalid measures of receptive codeswitching. Further work will show in how far logfile analysis and video ratings can complement each other in throwing light on other aspects of the learning process, for example, regarding vocabulary learning or grammar-related learning.

Appendices Appendix 7.1: Measures for story comprehension 2. Warum sagt Ruben ‘I wish I were at school’? [ ] weil ihm eingefallen ist, dass die Schule schon begonnen hat

122

Par t 2: Empir ical Studies Using Mixed Methods

[ ] Weil er seiner Lehrerin den Stein zeigen möchte [ ] Weil die Berufe zu schwer für ihn sind. 3. Wieso ist Ruben auf der Baustelle? [ ] weil er jeden Morgen zur Baustelle geht [ ] weil der Stein ihn auf die Baustelle gezaubert hat [ ] weil es dort weitere Steine gibt 4. Warum fragt Rubens Mutter ihn am Ende, ob Ruben krank sei? [ ] Weil Ruben zittert und hustet. [ ] Weil Ruben ein rotes Gesicht hat. [ ] Weil Ruben sagt, er möchte gerne zur Schule gehen Translation: 2. Why does Ruben say, ‘I wish I were at school’? [ ] because he remembers lessons have begun. [ ] because he wants to show the stone to his teacher. [ ] because the jobs are too difficult for him. 3. Why is Ruben on a construction site? [ ] Because he walks to the construction site every morning. [ ] Because the stone magically moved him there. [ ] Because there are more stones there. 4. Why does Ruben’s mother ask, toward the end of the story, if Ruben is ill? [ ] Because Ruben is shivering and coughing. [ ] Because Ruben’s face turned red. [ ] Because Ruben said he wants to go to school

Appendix 7.2: Relevant rating items for receptive code-switching (9) Rezeptives Code-Switching Es findet rezeptives Code-Switching statt. [Werden während der Instruktionsphase nur die Flaggen gedrückt, um die formale Bedienung zu lernen, wird dies nicht mitgezählt.] [ ] Ja [ ] Nein (10) Fokussiertes rezeptives Code-Switching Fokussierte rezeptive Code-Switches können beobachtet werden (Switches, die erkennbar der Bedeutungs- oder Formaushandlung dienen).

Codesw itching Your Way to L anguage Lear ning?

123

bezogen auf den englischen insgesamt Text (d.h. nicht ausschließlich zwischen dt. und türk.) gar nicht etwas: 1–3 ‘normale’ Sequenzen intensiv: 3 oder mehr Sequenzen, oder eine sehr intensive, langandauernde Sequenz

(11) Offenes rezeptives Code-Switching Offene rezeptive Code-Switches können beobachtet werden (Switches, die erkennbar Spaß machen oder Neugierde befriedigen sollen), bezogen auf den englischen insgesamt Text (d.h. nicht ausschließlich zwischen dt. und türk.) gar nicht etwas: 1–3 ‘normale’ Sequenzen intensiv: 3 oder mehr Sequenzen, oder eine sehr intensive, langandauernde Sequenz Translation: (9) Receptive code-switching Receptive code-switching takes place. [If flags are clicked during the instruction phase to test how the software works this is not counted.] [ ] Yes [ ] No (10) Focused receptive code-switching Focused receptive code-switching can be observed (Switches that recognisably serve negotiation of meaning or form).

124

Par t 2: Empir ical Studies Using Mixed Methods

Referring to the English text All (i.e. not exclusively referring to the German and Turkish text) Not at all Somewhat: 1–3 ‘normal’ sequences1 Intensive: 3 or more sequences, or a very intensive, long sequence (11) Open receptive code-switching Open receptive code-switching can be observed (Switches that recognisably serve having fun or satisfying curiosity). Referring to the English text All (i.e. not exclusively referring to the German and Turkish text) Not at all Somewhat: 1–3 ‘normal’ sequences Intensive: 3 or more sequences, or a very intensive, long sequence

Note (1) In analogy to other items, exactly three sequences were rated as ‘Somewhat’.

References Bündgens-Kosten, J. and Elsner, D. (2014) Rezeptives Code-Switching ein- und mehrsprachiger Lerner/innen in multilingualen Settings [Receptive code-switching of mono- and multilingual learners in multilingual settings]. Fremdsprachen Lehren und Lernen 42 (2), 56–73. Bündgens-Kosten, J., Elsner, D. and Hardy, I. (2016) Videoanalyse ein- und mehrsprachiger Lerner/innen bei der computerbasierten Textarbeit: die Rolle von Code-Switching und Negotiation im frühen Englischunterricht [Video analysis of mono- and multilingual learners during computer-based text work: The role of code-switching and negotiation while reading in early-English]. In U. Rauin, M. Herrle and T. Engartner (eds) Videoanalysen in der Unterrichtsforschung: Methodische Vorgehensweisen und Anwendungsbeispiele [Video Analysis in Teaching Research: Methodological Approaches and Practice Examples] (pp. 225–241). Weinheim/Basel: Beltz Juventa.

Codesw itching Your Way to L anguage Lear ning?

125

Butzkamm, W. (2000) Monolingual principle. In M. Byram (ed.) Routledge Encyclopedia of Language Teaching and Learning (pp. 415–417). London and New York: Routledge. Butzkamm, W. (2003) Die Muttersprache als Sprach-Mutter: Ein Gegenentwurf zur herrschenden Theorie [The mother tongue as language mother: an alternative to dominant theory]. Französisch Heute 34 (2), 174–192. Council of Europe (2014) Language Education Policy. See http://www.coe.int/t/dg4/linguistic/Division_EN.asp (accessed 8 June 2015). Elley, W.B. (1989) Vocabulary acquisition from listening to stories. Reading Research Quarterly 24 (2), 174–187. Elley, W.B. (1997) In Praise of Incidental Learning: Lessons from Some Empirical Findings on Language Acquisition. See http://files.eric.ed.gov/fulltext/ED412563.pdf (accessed 8 June 2015). Ellis, G. and Brewster, G. (2014) Tell it Again. The New Storytelling Handbook for Primary English Language Teachers. London: British Council Publishing. See https://www. teachingenglish.org.uk/article/tell-it-again-storytelling-handbook-primary-englishlanguage-teachers (accessed 23 July 2015). Elsner, D. (2013) Fostering Multilingualism with Computer-Based Multilingual Storybooks: The European Comenius Project MuViT. Presentation at the World CALL Conference on Global Perspectives on Computer-Assisted Language Learning. Glasgow, 10–13 July 2013. See https://prezi.com/pbjnxpu-ybnj/fostering-multilingualism-with-computer-based-multi lingual-storybooks-the-european-comenius-project-muvit/ Elsner, D. (2015) Kompetenzorientiert unterrichten in der Grundschule [Competenceoriented teaching in the elementary school]. Englisch 1–4. München: Oldenbourg. Elsner, D., Bündgens-Kosten, J. and Hardy, I. (2014) Awareness of multilingual resources: EFL primary students’ receptive and productive code-switching during collaborative reading. In J. Enever, E. Lindgren and S. Ivanov (eds) Conference Proceedings from Early Language Learning: Theory and Practice 2014 (pp. 41–49). Umea: Umea University. García, O. (2009) Bilingual Education in the 21st Century: A Global Perspective. New York: Blackwell/Wiley. Ghosn, I.-K. (2013) Storybridge to Second Language Literacy: The Theory, Research, and Practice of Teaching English with Children’s Literature. Charlotte, NC: IAP. Groot-Wilken, B. and Husfeldt, V. (2013) Die Testinstrumente und -verfahren des EVENING-Projekts; Eine empirische Betrachtungsweise [The test instruments and methods of the EVENING Project: An empirical perspective]. In O. Börner, G. Engel and B. Groot-Wilken (eds) Hörverstehen – Leseverstehen – Sprechen: Diagnose und Förderung von sprachlichen Kompetenzen im Englischunterricht der Primarstufe [Listening Comprehension - Reading Comprehension – Speaking: Diagnosis and Support of Linguistic Competences in Elementary Level English Teaching] (pp. 121–140). Münster: Waxmann. Gumperz, J.J. and Hernandez-Chavez, E. (1972) Bilingualism, bidialectalism and classroom interaction. In C. Cazden, V. John and D. Hymes (eds) The Functions of Language in the Classroom (pp. 311–339). New York: Teachers College Press. Hidi, S. and Harackiewicz, J. (2000) Motivating the academically unmotivated: A critical issue for the 21st century. Review of Educational Research 70 (2), 151–180. Lenhard, W. and Schneider, W. (2006) Ein Leseverständnistest für Erst- und Sechstklässler: ELFE 1–6 [A Reading Comprehension Test for First- and Sixth-Graders: ELFE 1–6]. Göttingen: Hogreve. Liebscher, G. and Dailey-O’Cain, J. (2005) Learner code-switching in the content-based foreign language classroom. The Modern Language Journal 89 (2), 234–247. Lohe, V. and Elsner, D. (2014) Developing language swareness in primary school children with Multilingual Virtual Talking Books: First results of the pilot study. International Journal of Computer- Assisted Language Learning and Teaching (IJCALLT) 4 (4), 30–47. Ma, J. (2008) Reading the word and the world’ – how mind and culture are mediated through the use of dual-language storybooks. International Journal of Primary,

126

Par t 2: Empir ical Studies Using Mixed Methods

Elementary and Early Years Education 36 (3), Special Issue: Teaching and Learning as Socio-cultural processes, 237–251. Nemoianu, A.M. (1980) The boat’s gonna leave: A study of children learning a second language from conversations with other children. Pragmatics and Beyond (vol. 1, pp. 1–116). Amsterdam: J. Benjamins. Quintero, E.P. and Rummel, M.K. (1998) American Voices: Webs of Diversity. Upper Saddle River, NJ: Merrill. Robertson, L.H. (2006) Learning to read ‘properly’ by moving between parallel literacy classes. Language and Education 20 (1), 44–61. Sert, O. (2005) Code-switching. The Internet TSL Journal. See http://iteslj.org/Articles/ Sert-CodeSwitching.html (accessed 23 July 2015). Schnepf, S.V. (2004) How different are immigrants? A cross-country and cross-survey analysis of educational achievement. IZA Discussion Paper No. 1398. See http://ftp.iza. org/dp1398.pdf Thomas, S.G. (2001) Language Contact, an Introduction. Edinburgh: Edinburgh University Press. Varonis, E.M. and Gass, S.M. (1985) Non-native/non-native conversations: A model for negotiation of meaning. Applied Linguistics 6 (1), 71–90.

8

Individual Differences and English L2 Learning in Two Primary Classrooms in France Heather Hilton

Introduction This chapter reports on an exploratory project that was organised to observe two groups of French children, both beginning English in the autumn of 2012: one first-grade and one third-grade class (54 children in all). Three weeks of English lessons were followed in each school and a range of information was collected, using a ‘mixed methods’ approach: emerging L2 listening and speaking skills, but also each child’s linguistic profile, motivation for learning English, L1 knowledge and skill, cognitive profile and personality. This chapter will summarise some of the findings from this rich database. There is a tendency among non-specialists – as well as specialists working in the ‘nativist’ paradigm – to consider child language acquisition as something that takes place effortlessly and ‘implicitly’, with the child unaware of the feat being accomplished. Notions of this type seem to underlie recent educational policy lowering the age at which children begin formal learning of a foreign language in public schools: younger just seems to be better for European policy-makers, parents, and politicians (Murphy, 2014: 1; Nikolov, 2013). ‘Usage-based’ or ‘emergentist’ research on child language learning, however (Bates & Goodman, 1999; Lieven & Tomasello, 2008; MacWhinney, 1999; Tomasello, 2003), shows that first language (L1) acquisition is a full-time job for infants and young children, requiring much effortful learning and many years of constant exposure and practice. The acquisition of a foreign language (L2) is qualitatively different, but equally complex, for a variety of obvious but frequently-overlooked reasons: far less contact time with the L2 in classroom contexts, less social urgency (schoolaged L2 learners already know how to communicate with their entourage) and therefore lower motivation, and of course, the problem of ‘competition’ 127

128

Par t 2: Empir ical Studies Using Mixed Methods

from the solidly-mastered L1 when using and learning the L2 (MacWhinney, 1987; Bates & MacWhinney, 1987). Language learning – like all human learning – is complex, and it is not surprising that L1 and L2 acquisition researchers operating within the emergentist paradigm have recently been influenced by Complex Systems and Dynamic Systems Theory (or, more precisely, Complex Dynamic Systems Theory), introduced into language acquisition research circles by Paul van Geert (1991, 2006, 2007) and used in sensory-motor developmental studies by Esther Thelen and Linda Smith (1994). These analytic methods (emanating from research in mathematics and physics) are certainly a means of enriching research in the social sciences considerably, although their actual implementation remains to be determined. This chapter will not present new forms of data analysis (see Hilton & Royer, 2014 for a modest attempt), but will simply look in structured fashion at multiple parameters in two primary English classrooms, and possible complex interactions between these parameters in emergent L2 knowledge and skill. These factors, as Paradis (2011) has pointed out, are either contextual – external to the learner (institutional, methodological, social), or individual – internal to the learner (cognitive, affective, conative). In a ‘mixed methods’ approach, the data gathered can be analysed both quantitatively and qualitatively.

Individual Differences in Primary Language Learning A considerable body of scientific literature exists on the role of individual variables in L2 classroom learning, and it would not be feasible to attempt to summarise these findings here (see Skehan, 1991; Dörnyei & Skehan, 2003 and the articles in Robinson, 2002 for excellent comprehensive reviews; and Murphy (2014: 131–158) for an up-to-date summary of research on individual differences in child L2 learning). In the project reported here, we focused on cognitive, conative (i.e. motivational) and social factors that are relevant to outcomes in classroom language acquisition, both in the native and a foreign language. The tendency over the past few years has been to abandon the somewhat fuzzy notion of ‘language learning ability’ (as measured by instruments such as the MLAT (Carroll & Sapon, 1955), in favour of more precise characteristics or capacities that underlie this larger construct (Skehan, 2002). We can, for example, consider cognitive variables linked with working memory (WM). Attention is essential for learning of all kinds (Baars, 1997; Baddeley, 1999: 78–82; Boujon & Quaireau, 1997; Schmidt, 2000: 16) and measures of an individual’s attentional capacity classically figure in

English L2 Lear ning in Two French Pr imar y Cl assrooms

129

psychometric tests measuring cognitive ability (digit span, reverse digit span, sustained attention tasks, etc.). Robinson (1995) and Schmidt (2000) have carefully described how ‘noticing’ new forms and structures in the L2 classroom is a prerequisite for learning them and psychologists have known for over a century (James, 1890, cited by Camus, 1998: 167) that attention is necessary for associative learning, such as the learning of new soundmeaning pairings in L2 vocabulary acquisition. As studies of dyslexic learners have shown, the phonological component of working memory (Gathercole & Baddeley, 1994) plays an important role in language processing and language learning; logically enough, phonological capacity in working memory has also been linked with the ability to learn new L2 words (Baddeley et al., 1998; Papagno et al., 1991). A test of phonological capacity in working memory might also help to identify dyslexic learners participating in the study, this being an important variable, of course, in any language acquisition context. Variables linked with long-term memory that may be relevant in L2 classroom learning are knowledge of and skill in the first language, as well as metalinguistic knowledge and skill; a child with a large vocabulary or more advanced listening skill in his/her L1 is at an advantage for most kinds of classroom learning (Anderson & Nagy, 1983; Nagy, 1988; see also articles in Freedle & Carroll, 1972; Lundsteen, 1979) and so may also quite logically be at an advantage for L2 learning. Various studies have shown an advantage for children with better metalinguistic skills (ability to manipulate syllables, phonemes or morphemes, for example) in learning to read (Gombert & Colé, 2000; Nagy & Anderson, 1995) and it may be interesting to see if this advantage extends to oral L2 learning. Despite the contribution of cognitive factors such as these, language teachers know that any slight attentional or phonological deficit can be overcome (or any advantage undermined) by affective and conative characteristics and attitudes. The key determinant in any learning situation is the individual’s motivation for the subject and the work proposed; without sufficient desires to learn, s/he will not invest sufficient time and attention for learning to take place (Baddeley, 1999: 77; Dörnyei, 2001: 7). Motivation is a complex construct (Dörnyei & Otto, 1998) and probably changes from hour to hour (Dörnyei, 2001: 1), but it is useful to identify which learners in a classroom are particularly keen or demotivated by the subject or the activities proposed. Personality is possibly an even more complex construct than motivation and we will only look at ‘social competence’ – that is, a child’s interest in the group, social confidence and attitude towards sharing (Dumais et al., 1997). There is also the biological variable of gender, which may play a role in classroom language learning, since young girls appear to be at a biological (or social) advantage for early verbal learning (Burman et al., 2008). Biology – or rather physiology – brings us back to the question that opened this chapter; whether the age of the learner is also a relevant variable.

130

Par t 2: Empir ical Studies Using Mixed Methods

The Seine & Marne Primary English Study The study presented here was set up in the context of a recent (2012) decision by the French Ministry of Education to begin foreign language tuition as early as the Cours préparatoire (CP) – the first year of primary education, undertaken the year each child turns six. A small research team based in Paris was able, due to the gradual implementation of this new scheme, to follow two groups of children in the same school system, one starting English as a foreign language at age six, another beginning in the Cours élémentaire 2 (CE2, third grade), at the age of eight. This coincidental result of changing language policy provided an interesting laboratory to test the popular ‘earlier is better’ hypothesis, as well as the variables contributing to the language learning process described above. The 54 Francophone children participating in the project all began studying English as a foreign language in the autumn of 2012: 25 first-graders and 29 third-graders, enrolled in the same regroupement scolaire in the Seine et Marne region east of Paris (in which small schools in neighbouring villages work together to offer a coordinated primary school curriculum, with each establishment covering two or three grades). The ‘research questions’ initially underlying the project were very general, since most of the researchers involved are looking into primary classroom language learning for the first time; we simply wanted to conduct a structured exploratory observation, in order to study the methodological choices adopted by the teachers and to compare language-learning outcomes for six- and eight-year old beginning language learners. An important aspect of the study was to consider individual characteristics (such as those listed in the previous section) that may influence learning pathways during the early phases of L2 acquisition, in complex interaction with methodological and institutional variables. The research questions that will be dealt with here are: • •

What differences can be observed in the learning pathways of six- and eight-year-olds in the early stages of English L2 learning? Are there advantages to either age group? Which individual and contextual characteristics seem to affect the L2 primary learning process?

Data Collection Procedure and Instruments In all, 10 researchers have been involved in the data collection for this project, as well as Patricia Sapina-Descrouet, the Pedagogical Consultant for primary language teaching in the Seine et Marne school system (who was present for all phases of data collection). Two researchers filmed one week of English lessons in each school, at regular intervals throughout the first year:

English L2 Lear ning in Two French Pr imar y Cl assrooms

131

early December, mid-February and late May. In all, just over four hours of English lessons were recorded in each school, providing qualitative information on lesson content and the teachers’ methodological choices. In addition to these samples of actual classroom activities, we gave each teacher a questionnaire to complete. The questionnaire was an expanded version of the teacher questionnaires used in the Early Language Learning in Europe (ELLiE) study (Enever, 2011). These questions were discussed orally in a recorded interview in February 2013. The questions concern the teacher’s perception of, and adherence to, the national syllabus for primary English teaching, her background and training, methodological choices and preferences, and motivation for, and confidence in, her English lessons. An analysis of the classroom context will be presented in the first part of the findings section. Information on each of the 54 children participating in our study was obtained through a battery of 12 tests and questionnaires, chosen as measures of the cognitive, social and affective factors mentioned above as relevant to outcomes in classroom learning and classroom language learning. Table 8.1 presents a list of 11 of the instruments used for data collection. Sustained attention was measured by the Symbol Search test from the WISC-IV battery (in which children must rapidly scan rows of small pictures and cross out all the animals); working memory span and working memory ability were measured by the classic Digit Span and Reverse Digit Span tasks (in which the child has to recall increasingly long strings of discrete numbers, in order and in reverse), also taken from WISC-IV. The phonological component of working memory was tested with a non-word repetition task in French, in which the child repeats plausible L1 nonsense words of increasing length. L1 knowledge and skill were measured with the vocabulary and listening tests from Khomsi’s (2001) Evaluation du langage oral and metalinguistic skill was tested with syllable and phoneme suppression tasks in L1 French. The six-year-olds were also assessed for social competence, using the relevant 40-item scale from the Profil socio-affectif questionnaire, which the teacher fills in for each child. The children’s motivation for their English lessons (and for school in general) was measured using a ‘smiley’ questionnaire adapted from the ELLiE project; the children responded to the questionnaire at the beginning and end of the first year, and once a year for the following two years. Finally, the children’s parents were asked to fill in a linguistic profile questionnaire, designed to identify bi- or plurilingual individuals in the two classrooms, as well as parental attitudes to learning English (a contextual factor that could influence the children’s motivation). We are fully aware that the sorts of variables reflected in the measures adopted for our project are themselves dynamic constructs, rather than fixed traits (Dörnyei, 2009) and that static measures of the type used here can only reflect the intellectual or conative state the child was in when the test was administered. These instruments were, however, the best adapted to the context in which the data were collected (due to heavy time constraints) and

132

Par t 2: Empir ical Studies Using Mixed Methods

Table 8.1 Individual variables measured in the Seine & Marne Primary English project Variable

Task

Instrument

Attentional capacity in working memory

Symbol search task* Digit span* Working memory span (reverse digit)* Non-word repetition*

Weschler (2007) WISC-IV

Phonological memory

Social competence

L1 listening* L1 vocabulary* Syllable suppression* Phoneme suppression* Questionnaire

Motivation for English

Smiley questionnaire

Linguistic profile

parent questionnaire

L1 verbal skill L1 verbal knowledge Metalinguistic skill

Casalis (2000) Répétition de logotomes. Khomsi (2001) Evaluation du langage oral. Ecalle (2007) THaPho Dumais et al. (1997) Social Competence Scale Adapted from Early Language Learning in Europe project questionnaire (Enever, 2011) Questionnaire adapted from Enever (2011)

*Test was administered by certified psychometrists.

provide data that enable us to explore possible interactions between individual learner characteristics and emerging English knowledge and skill, even if the picture could be much more detailed for each learner. Three psychometrists spent two full days in each school to administer the nine tests that are starred in Table 8.1; these are days in which individual children are taken out of class every 30 minutes or so – a very disruptive procedure for the teachers, which cannot be undertaken repeatedly. Of the instruments listed in Table 8.1, only three could be administered collectively (symbol search and the syllable and phoneme suppression tasks).

Information on emergent English knowledge and skill Our three week-long observational visits to the schools (December, February, May) were also used to collect information on the children’s emergent knowledge of English; four language acquisition researchers received the children individually to complete a small set of tasks in English, in addition to the motivation questionnaires. These tasks were created for the project, to measure knowledge of and skill with elements that are particularly relevant in the early stages of language learning: phonological perception and imitation, listening, and very simple production. A list of the tasks, with their characteristics, is provided in Table 8.2. Each task was proposed at two

English L2 Lear ning in Two French Pr imar y Cl assrooms

133

Table 8.2 Tasks used to measure emergent knowledge and skill in English Test

Targeted knowledge, skill

Number of items

Test administration

Phoneme discrimination

Ability to perceive English ‘minimal pairs’ (certain vowel and consonant distinctions)

33

English listening

Ability to reconstruct the meaning of simple utterances in English; containing items previously heard in class, as well as new combinations of familiar lexical (and grammatical) items Ability to describe or react to pictures illustrating objects and phenomena covered in classroom activities Ability to repeat English phrases and sentences of increasing length, containing known words and expressions, but also known words in new combinations, and one or two unrehearsed cognates (such as ketchup, cornflakes, banana)

15

Individual, in December and May; automated presentation (in Praat); children’s answer recorded by examiner Individual; paper presentation in February; laptop presentation in May; answers recorded by examiner

Simple production

Elicited imitation

(open)

Individual, in February and May; production recorded by examiner

16

Individual, in February and May; production recorded by examiner

different times, either December and May, or February and May, to give a longitudinal perspective on the learning process. A 33-item phoneme discrimination task was administered in December and May; the child hears two English monosyllables (usually CVC, like ‘but/bat’), and says whether they are the same or different. This task was programmed in Praat (Boersma & Weenink, 2015), but answers were keyed in by the researcher (so reaction times were not recorded). Skill in making sense of spoken English was measured with a ‘listen and point’ task, in which the child heard a short utterance and selected (from a set of pictures) the one corresponding to the meaning of the utterance. The examiner noted the children’s responses on a sheet of paper (again, no reaction time information). All of the stimuli for the receptive tasks were recorded and a laptop was used to play them to the child over a set of good-quality padded headphones.

134

Par t 2: Empir ical Studies Using Mixed Methods

The listening task was immediately followed by a speaking task, in which the child was asked to describe or comment freely on additional pictures, similar to those in the listening activity. This was followed by the elicited imitation task, in which the child repeated phrases and sentences of increasing length (from two to 10 syllables). Some of these sentences had been heard and repeated in class, others contained new combinations of familiar words or cognates. All production data were recorded directly onto the examiner’s laptop, using Audacity recording software, and a single sound file saved for each subject. The February productions have been orthographically transcribed and analysed and we will comment on these here. For the imitation task, following Tracy-Ventura et al. (2014), we counted the number of ‘correct’ syllables produced for each imitated utterance (all consonants and vowels recognisably reproduced), as well as the overall quality of each imitated utterance (taking prosody and fluency into account, as well as phonological accuracy) on a scale from 0 to 4. Two raters were involved in these qualitative judgments, with an interrater agreement of 94.6%. For the openended speaking task, numbers of words produced in all, as well as mean, length of utterance (in words) are used as production measures. Error rate is the number of errors produced per 100 words. For both tasks, the transcription conventions of the Child Language Data Exchange System were used (MacWhinney, 2000). Parental consent was requested – and 100% obtained – for all aspects of data collection and use. In a longitudinal follow-up to this initial round of data collection, we returned to the schools twice; in May 2014 and June 2015.

Findings The classroom contexts Although two different schools participated in our study, both are part of the same micro-network (villages only three kilometres apart), ensuring that the 54 learners involved have the same socioeconomic background (there are eight pairs of siblings, for example). All 25 first-graders were born in 2006; all 29 third-graders in 2004. Table 8.3 further illustrates the composition of the groups, with information obtained through linguistic profile questionnaires, as well as the filmed lessons and the teacher interviews. The two classes are composed mostly of monolingual children, with only three or four multilinguals (children who use a language other than French at home) in each class. There are also more girls (n = 31) than boys (n = 23). Table 8.3 also summarises some basic differences in the children’s learning contexts. The first-grade teacher majored in English at college, had lived for two years in the United States, was highly motivated, structured and confident in her approach to her English lessons (from teacher interviews),

English L2 Lear ning in Two French Pr imar y Cl assrooms

135

Table 8.3 Composition of the primary English groups, institutional and methodological variables 1st graders (n = 25) Group characteristics and numbers Girls/boys 15/10 French monolinguals/bi- or 21/4 plurilinguals School and classroom characteristics Teacher’s college major English Teacher is also main teacher English sessions per week 3 × 25 minutes (Monday, Tuesday, Thursday morning) Methodology and support Communicative materials approach, teacherdesigned materials (no textbook), regular use of storybooks

3rd graders (n = 29) 16/13 26/3

Economics not main teacher 2 × 40 minutes (Tuesday morning, Thursday afternoon) Communicative approach, textbook with audio recordings + teacherdesigned materials, interactive whiteboard

which she used to start off three out of the four school days every week; she was the main teacher for the group and thus worked with the same students all day. The third-grade English teacher was not the third-graders’ primary teacher, but came in from another classroom to work with the class for English only. These language lessons were set at two different times during the week, for longer sessions. The third-graders’ English teacher majored in Economics and Business in college, and volunteered to participate in the project precisely because she was hoping for guidance in organising and carrying out the English lessons (from teacher interviews and conversations throughout the year). She was very motivated for this subject, but reported herself as feeling ‘unsure’ of her abilities in oral English. At the beginning of the project, an interactive whiteboard had just been installed in the third-grade classroom and the English teacher and children adopted this resource enthusiastically. Both teachers used a wide variety of materials to supplement their lessons: real objects, paper or fabric puppets, activity cards, games and so on. The first-grade teacher also included picture books as a systematic feature of her lessons (at the end of each lesson, one child chose a story to listen to from a well-stocked classroom library). Both teachers used a communicative approach, with a functionally-organised syllabus (greeting, asking someone’s name, asking and telling your age, expressing likes and dislikes etc.), small group work, role plays and so on. The English lessons in both classrooms also contained a lot of teacher-student question-answer interaction, as well as much choral and individual repetition, to practice new words and

136

Par t 2: Empir ical Studies Using Mixed Methods

expressions. The first-grade teacher produced all the materials used for her English lessons; the third-grade teacher used commercially-available teaching materials (Voise et al., 2009), which she augmented with her own materials and interactive white board (IWB) activities.

Group comparisons Table 8.4 presents the group averages for the various measures taken during the first year of the project and the results of non-parametric comparisons between the six- and eight-year olds. Most of the findings concerning the tests of individual differences are predictable (since many of these are standardised tests at which eight-year-olds routinely outperform six-yearolds), but some of the group comparisons in column four are more unexpected. It is, for example, predictable (and completely normal) to find third-graders surpassing first-graders in the L1 listening and vocabulary tests (measures 1 and 2), but it is surprising to find the six-year-olds performing at equivalent levels in the Symbol Search and Reverse Digit Span tests of working memory and attentional capacity (measures 3 and 5). It is also surprising that the first-graders outperformed the third-graders on the non-word repetition task (measure 6); unless there are a number of dyslexic children in the third-grade class (and there are indeed two children in this group undergoing speech therapy), one would expect the scores to be equivalent. As a group, the first-graders also appear to have above-average metalinguistic skill (measures 7 and 8) for their age; as beginning readers, their performance on the phoneme suppression task is particularly impressive. These results on the psychometric tests lead us to suspect that the first-grade class is a highly-functioning group of youngsters, a hypothesis that was confirmed in conversations with teachers and administrators working with them. The group comparisons for the English tasks analysed here also reveal a few surprises. Both groups are equally motivated for their English lessons early in the school year, but the younger children are slightly more enthusiastic by the end of the year (measures 10 and 11); this may be due to methodological or institutional variables or to the fact that eight-year olds are more blasé when answering questionnaires. For the receptive English tasks (measures 12 and 13), we do not see a phonological advantage for the firstgraders (phoneme discrimination task), but their L2 listening performance is equal to that of the third-graders, an interesting – and perhaps unprecedented – finding. Our psychometric tests (measure 2) confirm the established fact that eight-year-olds are more skilful L1 listeners; it is therefore interesting to observe that this age advantage disappears in L2 listening. Various factors could account for the impressive performance of the six-year-olds on the L2 listening task: their excellent working memory and phonological capacities, the rich and high-quality input they were exposed to during their English lessons, which were more regular, and conducted by a linguistically

English L2 Lear ning in Two French Pr imar y Cl assrooms

137

Table 8.4 Class means for measures of individual differences and English knowledge and skill, with group comparisons (column 4)

Individual differences (tests in French) (1) French vocabulary

1st graders

3rd graders

Group comparisons (Mann-Whitney)

15.9

17.5

3rd > 1st* (U = 195.0, p < 0.01) 3rd > 1st*** (U = 43.0, p < 0.0001) 1st = 3rd (U = 283.0, p = 0.17) 3rd > 1st* (U = 184.5, p < 0.01) 1st = 3rd (U = 264.5, p = 0.08) 1st > 3rd** (U = 170.0, p < 0.001) 1st ≈ 3rd (U = 247.5, p = 0.053) 1st = 3rd (U = 308.0, p = 0.40) (ranging from 154–225)

(2)

French listening

13.2

17.2

(3)

Symbol Search (sustained attention)

11.8

10.8

(4)

Digit Span (WM capacity)

5.8

7.3

(5) Reverse Digit Span (WM span)

4.6

5.3

(6)

Non-word repetition (phonological memory)

94.1

86.9

(7)

Phoneme suppression (metalinguistic skill)

5.0

5.4

(8)

Syllable suppression (metalinguistic skill)

5.4

5.7

(9)

Social competence (first204.6 graders only) Tasks in English and motivation for English (2012–2013) (10) Motivation for English in 21.5 December Range 12–24 (11) Motivation for English in May 22.3 Range 18–24 (12)

–

20.0 9–24 20.0 7–24

English phoneme discrimination (December)

21.3

20.4

Range

12–28

14–24

1st = 3rd (U = 254.5, p = 0.18) 1st > 3rd (m) (U = 206.0, p = 0.03) 1st = 3rd (U = 273, p = 0.25) (Continued)

138

Par t 2: Empir ical Studies Using Mixed Methods

Table 8.4 Class means for measures of individual differences and English knowledge and skill, with group comparisons (column 4) (continued)

(13)

(14)

(15)

(16)

(17)

(18)

English listening (Dec + May combined score) Range English production (number of utterances) Range English production (MLU in words) Range English production (error rate per 100 words) Range Elicited imitation (percentage of correctly imitated syllables, Feb.) Range Elicited imitation (quality score, 0–4)

1st graders

3rd graders

Group comparisons (Mann-Whitney)

15.1

15.8

10–22 9.1

9–21 13.5

1st = 3rd (U = 281.0, p = 0.40)

3–14 1.3

0–25 2.7

1–3.9 2.1

0–4.5 11.5

0–14.3 70.1 52–89

0–44 63.8 47–84

3.0

2.7

3rd > 1st** (U = 115.0, p = 0.001) 3rd > 1st*** (U = 69.5, p < 0.0001) 3rd > 1st*** (U = 84.0, p < 0.0001) 1st > 3rd* (U = 146, p = 0.01) 1st > 3rd(m) (U = 169, p = 0.05)

p < 0.05 *p < 0.01 **p < 0.001 ***p < 0.0001

(m)

confident teacher, with advanced competence in English (excellent phonology, prosody, accuracy and fluency). The results on our February production measures also show interesting group differences for L2 speaking. In the more open speaking task (measures 14–16), the eight-year-olds produce more and longer utterances; this greater productivity naturally leads them to commit more errors of English syntax or morphology at this very early stage in their L2 tuition. An average mean length of utterance (MLU) of 1.3 words shows that the six-year-olds were essentially naming objects or phenomena in the pictures, usually with a single word (‘dog’, ‘boy’, ‘five’). On the imitation task, however, the six-yearolds clearly outperform the eight-year-olds, in terms of numbers of recognisable syllables produced (line 17), and the slightly better overall quality of their imitated sentences (line 18). This tendency towards greater phonological and prosodic accuracy could, again, be attributed to this group’s excellent working memory abilities (including phonological WM), to the large

English L2 Lear ning in Two French Pr imar y Cl assrooms

139

quantity of correctly-pronounced English heard during their lessons (storyreading tasks, instructions, etc.), to their younger age – or most probably to a complex mixture of all three. Comparisons of the male and female learners did not reveal any group differences on the measures of emergent English skill, although the girls were more motivated about their English lessons in February than the boys (U = 191, p = 0.01*), a difference which disappears in May. We did not find an advantage for the small group of bilingual children participating in the study (n = 9), compared to the 45 monolingual learners. Spearman correlations run on the measures in Table 8.4 indicate possible interactions between: metalinguistic skill (the phoneme suppression task) and success on the English listening tasks (rS = 0.527**); phonological memory and grammatical accuracy (with a negative correlation (−0.368*) between the non-word repetition task and the error rate in English production). An unexpected negative correlation was also found between L1 listening skill and L2 imitation skill (rS = −0.368*); this could simply reflect the first-grader advantage for imitation already observed (since the six-year-olds are also at a disadvantage for L1 listening). Correlations between the measures of emergent English skill hint at an interesting possible relationship between imitation ability and grammatical accuracy, with a negative correlation between imitation success and the rate of errors in L2 speaking (rS = −0.301, p < 0.05m). Somewhat surprisingly, the social competence scores (obtained for the first-graders) did not correlate with any of their English scores.

Individual profiles To appreciate possible interactions between individual differences and L2 primary learning, complex systems theory encourages the researcher to eschew correlations and group comparisons, in favour of detailed studies of individuals, in particular those whose language performance is either extremely good or extremely poor (Verspoor et al., 2013). Our project data enables us to look at children with outlier performance in the English tasks, and examine various factors (as measured by the psychometric tests and questionnaires) that might help understand their performance. Imitation is certainly a crucial and underestimated skill in languagelearning and it is interesting to look at the two children with the best results on the English imitation task in February (over 80% of the total number of syllables in the task correctly reproduced). Both of these children are girls (one six, one eight), and both, as it turns out, are among the small sub-group of bilingual participants in the project. The first-grade imitation expert – whom we will refer to by her initials, PS – is the child with the highest score for both groups on the sustained attention task; she also has a ‘perfect’ motivation score, both in February and in May, and a high score on the social competence scale. Both PS and eight-year-old IB obtained perfect scores on the L1

140

Par t 2: Empir ical Studies Using Mixed Methods

non-word repetition test; sound phonological memory no doubt underlies their impressive ability to imitate a new language. Unlike PS, IB’s score on the symbol search task is low, but her digit span score is literally ‘off the charts’ (the maximum observed for all the learners), with 13 numbers recalled correctly in order. Interestingly enough, neither PS nor IB have particularly high L1 listening scores, but their L2 listening scores (identical in February and equivalent in May), are well above average, with IB obtaining one of the best scores in the third-grade group. Interestingly, PS does not do particularly well on the L2 phoneme discrimination task and both girls are only moderately productive in the speaking task, producing 10 and 11 words in all. It is interesting to note that both girls also preferred not to answer (rather than point at random) when they had not understood during the first English listening task; this may indicate a cautious side to their personality that is also reflected in their limited productivity in a testing situation. The two children with the greatest difficulty imitating L2 English utterances (fewer than 50% of the target syllables reproduced recognisably) are both third-graders and both boys, one of whom is bilingual (like the L2 imitation experts). The non-word repetition score of AN reveals that he is probably dyslexic (although this did not come through in the parent questionnaire); MP has an average score on this measure, but lower-than-average scores on the working memory span measures. Despite his phonological difficulties, AN reports high motivation for English, both in February and May; in February, MP has the secondlowest motivation score of all; it increases in May, but remains well below the group average. It is very interesting that despite their difficulties with the imitation task – which taxes working memory, of course – both of these boys have enthusiastic and creative production behaviour, with mean utterance length of over three words, and valiant attempts to combine the words they know in original statements: ‘a girl go a pink’ [for a girl wearing a pink shirt], ‘it’s a book yellow and red’, ‘three dog black’, ‘look, baby’). The most successful L2 listener in our subject pool is a six-year-old girl, AA, who also has very high results on the L2 auditory discrimination task and good imitation skill, but – like PS – is not particularly productive in her early L2 speech. The child with the lowest listening score is a third-grade girl, LR, who also has the lowest score on the symbol search measure of sustained attention – a combination which probably highlights the relationship between working memory and task demands. LR also has a very low non-word repetition score (and therefore a problem with phonological memory); she is, however, quite productive (above the group average), despite relatively low scores on the motivation questionnaires. The most productive L2 speaker, AR (who produced close to 100 words, with an average utterance length of over four words), is a third-grade boy; he is also one of the bilingual subjects, with good metalinguistic skills and strong L2 phoneme discrimination ability, who scored very high on the symbol search test of sustained attention. At the opposite extreme, another eight-year-old boy, MCA, declined to speak at all during the

English L2 Lear ning in Two French Pr imar y Cl assrooms

141

first production task; this is a child with excellent attentional and working memory capacity, good metalinguistic skill and phonological memory, and a high English listening score; he also, however, preferred not to answer listening questions in case of uncertainty, so is probably, like the expert imitators, PS and IB, not keen to produce if he is not sure of what to say.

Conclusions, Methodological and Pedagogical Considerations This brief summary of part of a large dataset certainly corroborates the hypothesis that primary classroom language learning is a complex process, in which variables of all kinds interact in sometimes unexpected ways. Group comparisons can give some idea of the effect of variables such as age, classroom methodology or motivation in L2 learning, but our closer look at a few individuals at the extremes of these groups shows that there is not one particular variable or set of variables that determines success or difficulty in this L2 learning context (Paradis, 2011; Sylva, 2016). Cognitive aptitude – as measured by the psychometric tests of working memory (capacity, span, non-word repetition) – appears to be linked to enhanced skill at L2 imitation, which, in turn, may well contribute to the emergence of grammatical competence, since we found a weak negative correlation between successfullyimitated syllables and error rates in spoken production. A high percentage of correctly-recalled syllables in the sentence imitation task depend on getting not only the content words right, but also the function words and grammatical morphemes. Cognitive aptitude may also be linked to strong performance on the L2 listening task, but it is hard to say whether the cognitive advantage is for listening, or rather for execution of the task itself. Qualitative observations from our longitudinal follow-up in 2014 and 2015 have revealed an unexpected set of institutional variables, which overwhelm more subtle issues of a slight cognitive or linguistic advantage for individuals in our groups of child learners. Regrettably, the promising group of six-year-old beginners were not offered a structured English programme in second grade (2013–2014); their motivation for English dropped and they were unable to execute the speaking task requested of them in May 2014. The good news is that their English listening skill remained amazingly intact, and even increased slightly; this is possibly an effect of age on the listen-and-point task, but it does indicate some long-term benefits of high-quality L2 tuition as early as first grade (in line with findings reported by Graham, 2016 and Unsworth et al., 2015). In the 2014–2015 school year, the younger group (now in third grade) was ‘back on track’ with regular English lessons and our 2015 data collection indicates that they are now very keen to speak in English. The older group of learners continued, enthusiastically, with the same English teacher in fourth grade and were quite productive in 2014; oddly enough, their

142

Par t 2: Empir ical Studies Using Mixed Methods

English listening skill did not improve during this second year, due, perhaps, to a very limited lexical syllabus and to the Francophone phonological model provided by their teacher. Sadly, this older group also experienced a gap in its foreign language tuition, their fifth grade teacher (2014–2015) having decided his level in English was insufficient to continue the programme. A longitudinal study of this type can help us to understand some of the multiple factors that shape learning outcomes in primary classroom language learning. The preceding paragraph points up surprising irregularities in the institutional implementation of the primary L2 curriculum in France and the importance of continuity in maintaining learner motivation and emergent language skill (corroborating large-scale findings by Enever, 2011 and Nikolov, 2013). I have come away from this three-year observational period with doubts as to the relevance of the functionally-organised communicative or ‘actional’ approach (Council of Europe, 2001) at a young beginner level, in contexts where children only have an hour or so of classroom contact time with the language per week. Over the two years where each group has actually had English lessons, the functional syllabus seemed to return incessantly to the same language functions that were covered in year 1 (greetings, giving your age, the date, counting, expressing food-related likes and dislikes); the tasks and activities undertaken do not appear to enable the children to extract morpho-syntactic regularities from the input (‘*it’s a book blue’ still being produced; confusions in the use and understanding of ‘I’ and ‘you’, etc.). Yet the enthusiasm and the potential of every child participating in the project can act as an inspiration, to encourage us to propose methodology, materials and programmes better adapted to the complexities of this particular learning situation.

Acknowledgements This project was made possible by a PARI grant from the Université Paris 8 and by the unconditional cooperation of the Board of Education of the Seine et Marne département, the teachers and school directors and the participating families. Special thanks go to Patricia Descrouet-Sapina, Béatrice Grellier, Sabine Brito, Marie-Claude Mercy and all of the children who welcomed us into their classrooms.

References Anderson, R.C. and Nagy, W.E. (1983) Reading comprehension and the assessment and acquisition of word knowledge. Advances in Reading Research 2, 231–256. Baars, B.H. (1997) In the theatre of consciousness: Global workspace theory, a rigorous scientific theory of consciousness. Journal of Consciousness Studies 4 (4), 292–309. Baddeley, A.D. (1999) Essentials of Human Memory. Hove: Psychology Press. Baddeley, A., Gathercole, S. and Papagno, C. (1998) The phonological loop as a language learning device. Psychological Review 105 (1), 158–173. Bates, E. and Goodman, J.C. (1999) On the emergence of grammar from the lexicon. In B. MacWhinney (ed.) The Emergence of Language (pp. 29–80). Mahwah, NJ: Lawrence Erlbaum.

English L2 Lear ning in Two French Pr imar y Cl assrooms

143

Bates, E. and MacWhinney, B. (1987) Competition, variation, and language learning. In B. MacWhinney (ed.) Mechanisms of Language Acquisition (pp. 157–193). Hillsdale, NJ: Lawrence Erlbaum. Boersma, P. and Weenink, D. (2015) Praat: Doing Phonetics by Computer. See http://www. praat.org/(accessed 20 June 2015). Boujon, C. and Quaireau, C. (1997) Attention et Réussite Scolaire. Paris: Dunod. Burman, D.D., Bitan, T. and Booth, J.R. (2008) Sex differences in neural processing of language among children. Neuropsychologia 46 (5), 1349–1362. Camus, J.-F. (1998) L’attention. In J.-L. Roulin (ed.) La Psychologie Cognitive (pp. 137–206). Rosny: Bréal. Carroll, J.B. and Sapon, S.M. (1955) Modern Language Aptitude Test. New York: The Psychological Corporation (Harcourt Brace Jovanovich). Casalis, S. (2000) Répétition de Logotomes. Lille: Université de Lille. Council of Europe (2001) The Common European Framework of Reference for Languages. Cambridge: Cambridge University Press. Dörnyei, Z. (2001) Motivational Strategies in the Language Classroom. Cambridge: Cambridge University Press. Dörnyei, Z. (2009) Individual differences: Interplay of learner characteristics and learning environment. In N.C. Ellis and D. Larsen-Freeman (eds) Language as a Complex Adaptive System (pp. 230–248). Oxford: Wiley-Blackwell. Dörnyei, Z. and Otto, I. (1998) Motivation in action: A process model of L2 motivation. Thames Valley University Working Papers in Applied Linguistics 4, 43–69. Dörnyei, Z. and Skehan, P. (2003) Individual differences in second language learning. In C.J. Doughty and M.H. Long (eds) The Handbook of Second Language Acquisition (pp. 589–630). Oxford: Blackwell. Dumais, J.E., LaFreniere, P.J., Capuano, F. and Durning, P. (1997) Profil Socio-affectif. Paris: ECPA. Ecalle, J. (2007) THaPho: Test ‘d’Habiletés Phonologiques. Dardilly: Mot à mot. Enever, J. (ed.) (2011) ELLiE: Early Language Learning in Europe. United Kingdom: British Council. Freedle, R.O. and Carroll, J.B. (1972) Language Comprehension and the Acquisition of Knowledge. Washington, DC: Winston. Gathercole, S.E. and Baddeley, A. (1994) Working Memory and Language. Hove: Lawrence Erlbaum. Gombert, J.E. and Colé, P. (2000) Activités métalinguistiques, lecture et illettrisme. In M. Kail and M. Fayol (eds) ‘L’Acquisition du Langage (Vol. 2) (pp. 117–150). Paris: Presses Universitaires de France. Graham, S. (2016) Learning French in the primary school: The role of teaching and teacher factors. Paper presented at the Young Language Learners Symposium, 6 June 2016, Oxford, England: Department of Education, University of Oxford. Hilton, H.E. and Royer, C. (2014) An analysis of complexity in primary school L2 English learning. EUROSLA Yearbook 14, 65–78. James, W. (1890) Principles of Psychology. New York: H. Holt and Co. Khomsi, A. (2001) Evaluation du Langage Oral. Paris: ECPA. Lieven, E. and Tomasello, M. (2008) Children’s first language acquisition from a usagebased perspective. In P. Robinson and N.C. Ellis (eds) Handbook of Cognitive Linguistics and Second Language Acquisition (pp. 168–196). London: Routledge. Lundsteen, S.W. (1979) Listening: Its Impact at All Levels on Reading and the Other Language Arts. Urbana, IL: ERIC Clearinghouse on Reading and Communication Skills. MacWhinney, B. (1987) The competition model. In B. MacWhinney (ed.) Mechanisms of Language Acquisition (pp. 249–308). Hillsdale, NJ: Lawrence Erlbaum. MacWhinney, B. (ed.) (1999) The Emergence of Language. Mahwah, NJ: Lawrence Erlbaum.

144

Par t 2: Empir ical Studies Using Mixed Methods

MacWhinney, B. (2000) The CHILDES Project: Tools for Analyzing Talk. Mahwah, NJ: Lawrence Erlbaum Associates. Murphy, V.A. (2014) Second Language Learning in the Early School Years: Trends and Contexts. Oxford: Oxford University Press. Nagy, W. (1988) Teaching Vocabulary to Improve Reading Comprehension. Urbana, IL: International Reading Association. Nagy, W.E. and Anderson, R.C. (1995) Metalinguistic Awareness and Literacy Acquisition in Different Languages. Technical Report #618, Champaign, IL: College of Education, University of Illinois. Nikolov, M. (2013) Early foreign language learning: Is it child’s play? Plenary lecture presented at the EUROSLA 23 Conference, 30 August 2013, Amsterdam: Universiteit van Amsterdam. Papagno, C., Valentine, T. and Baddeley, A. (1991) Phonological short-term memory and foreign-language vocabulary learning. Journal of Memory and Language 30, 331–347. Paradis, J. (2011) Individual differences in child English second language acquisition. Linguistic Approaches to Bilingualism 1 (3), 213–237. Robinson, P. (ed.) (1995) Attention, memory, and the ‘noticing’ hypothesis. Language Learning 45 (2), 283–331. Robinson, P. (2002) Individual Differences and Instructed Language Learning. Amsterdam: John Benjamins. Schmidt, R. (2000) Attention. In P. Robinson (ed.) Cognition and Second Language Instruction (pp. 3–32). Cambridge: Cambridge University Press. Skehan, P. (1991) Individual differences in second language learning. Studies in Second Language Acquisition 13 (2), 275–298. Skehan.P. (2002) Theorizing and updating aptitude. In P. Robinson (ed.) Individual Differences and Instructed Language Learning (pp. 69–94). Amsterdam: John Benjamins. Sylva, K. (2016). Supporting bilingual learners in English speaking pre-schools. Paper presented at the Young Language Learners Symposium, 6 July 2016, Oxford, England: Department of Education, University of Oxford. Thelen, E. and Smith, L.B. (1994) A Dynamic Systems Approach to the Development of Cognition and Action. Cambridge, MA: MIT Press. Tomasello, M. (2003) Constructing a Language. Cambridge, MA: Harvard University Press. Tracy-Ventura, N., McManus, K., Norris, J.M. and Ortega, L. (2014) “Repeat as much as you can”: Elicited Imitation as a measure of oral proficiency in L2 French. In P. Leclercq, A. Edmonds and H. Hilton (eds) Measuring L2 Proficiency: Perspectives from SLA (pp. 143–166). Bristol: Multilingual Matters. Unsworth, S., Persson, L., Prins, T. and de Bot, K. (2015) An investigation of factors affecting early foreign language learning in the Netherlands. Applied Linguistics 36 (5), 527–548. van Geert, P. (1991) A dynamic systems model of cognitive and language growth. Psychological Review 98, 3–53. van Geert, P. (2006) Dynamic systems methods in the study of language. Presented at Apprentissage des Langues Premières et Secondes (conference), 24 January 2006, Paris: Ministère de la Recherche. van Geert, P. (2007) Dynamic systems in second language learning: Some general methodological reflections. Bilingualism: Language and Cognition 10 (1), 47–49. Verspoor, M., Chan, H., and Lowie, W. (2013) The dynamic development of L2 sentence complexity. Presented at the ADYLOC Modeling Workshop, 8 April 2013, Paris: UMR 7023, Centre National de la Recherche Scientifique. Voise, A.-M., Perez, F., Furgerot, J.-M., Guyonnet, K. and Dunn, N. (2009) Speak and Play CE2. Les Mureaux: SED. Weschler, D. (2007) WISC-IV: Wechsler Intelligence Scale for Children (French version). Paris: ECPA.

9

Content and Language Integrated Learning (CLIL): A Panacea for Young English Language Learners? Karmen Pižorn

Introduction In the last decade, many countries in the EU and other parts of the world have implemented content and language integrated learning (CLIL) in their state schools (Lorenzo et al., 2010) based on the rationale that by integrating foreign language (FL) and content learning, foreign language learning will automatically improve with at least no cost to content learning. However, empirical CLIL research does not convincingly demonstrate the foreign language benefits of CLIL, and many CLIL research studies suggest that benefits may be attributable to a number of factors (Bruton, 2011, 2013). This is especially the case when young foreign language learners are considered. There have not been many studies conducted to explore CLIL and its advantages and disadvantages in early foreign language classrooms across the world. Thus governments, when lowering the age of the compulsory foreign language onset, are not provided with many clear-cut research outcomes on which foreign language teaching approach to implement in their primary schools. Furthermore, foreign language teaching is a complex dynamic process, which requires a complex dynamic systems perspective towards doing research (Dörnyei, 2014). CLIL as a dual-focused teaching process adds to this complexity and is considered to have a number of multi-dimensional features (Dalton-Puffer, 2008). To explore such complicated processes and outcomes, researchers need to utilise both qualitative and quantitative methods (Yin, 2006). The integration of qualitative and quantitative methods within a systematically developed mixed methods design may prove to be a useful tool for conducting research in applied linguistics. 145

146

Par t 2: Empir ical Studies Using Mixed Methods

Critical Review of Research on Primary CLIL CLIL was developed as a European initiative, inspired by other models such as French Immersion in Canada or CBI (content-based instruction) in the US. Its popularity began to grow with the publication of the White Paper by the European Commission (1995). It incorporates any activity in which ‘a foreign language is used as a tool in the learning of a non-linguistic subject in which both language and the subject have a joint role’ (Marsh, 2002: 58). A conceptual framework of CLIL has been gradually developed, however, primarily in relation to the needs and cognitive development of older (secondary) school learners and not openly considering the features of young learners explicitly. Influenced by the work of Mohan (1986), Coyle (1999) designed the 4Cs conceptual framework focusing on the interrelationship between content (subject matter), communication (language), cognition (learning and thinking) and culture (social awareness of self and ‘otherness’). Coyle (2008: 550) highlights the importance of the integrating the 4Cs as follows: ‘It goes beyond considering subject matter and language as two separate elements but rather positions content in the “knowledge for learning” domain (integrating content and cognition) and language, a culturebound phenomenon, as a medium for learning (integrating communication and intercultural understanding)’. According to Coyle et al. (2010), the 4Cs framework is built on six principles: (1) subject matter is about the learners constructing their own knowledge and developing relevant skills – and not about transmitting the content in a limited period of time (Eurydice Report, 2006; Lantolf, 2000; Vygotsky, 1978); (2) in constructing their own knowledge and skills, learners need to be able to understand the language of teaching/learning the content (Met, 1998); (3) thinking processes require analysis in terms of their linguistic demands to facilitate development (Bloom, 1984; McGuiness, 1999); (4) learners should learn the foreign language in context (Krashen, 1985; Swain, 2000); (5) learners need to interact for learning to take place (Mohan, 1986; Pica, 1991; van Lier, 1996); and (6) intercultural understanding stimulates the development of alternative approaches to the learning and teaching of content and language (Byram et al., 2001; Cummins, 2004). A close observation of these six principles shows that it would be inappropriate if these principles were over-generalised to all age groups, especially to young language learners. For example, principle four may not be automatically transferred into a young foreign language classroom, as young learners are not able to learn/acquire a language in any kind of de-contextualised situations; instead, what young language students need is a number of playful, game-like activities, appropriately scaffolded, with plenty of multi-sensory teaching and an energetic teacher with excellent classroom managerial skills. To set a context for a young learner is by far not the only prerequisite for a

CL IL: A Panacea for Young ELL s?

147

successful lesson. The other principle that may be disputed with regard to a young foreign language classroom is principle five. The research shows that many children go through the so-called silent period and a forced interaction, which children are pushed into while not ready to start speaking, could even have a negative impact on their speaking and fluency skills development. Therefore, more research is needed to reflect on the particular characteristics of young language learners in CLIL programmes when applying the above six principles. There are a number of CLIL programmes, which differ in variables such as starting age, duration, intensity and compulsory status (Coyle, 2007). Clegg (2003) differentiates between language-led CLIL programmes, which borrow (parts of) subjects and emphasise language development, and subjectled programmes, which in extreme versions may provide no explicit language teaching. In the CLIL Compendium (Marsh & Maljers, 2001), European models take a distinctive structure as they are positioned on a specific language continuum, which takes into consideration different social and contextual variables (Baetens Beardsmore, 2009). Let us consider two CLIL programmes as opposite extremes from this continuum. Slovene-Hungarian primary and lower-secondary bilingual schools, which were set up in the 1960s for the Hungarian minority living in the eastern part of Slovenia, practise a type of CLIL that can be defined as a total immersion, since all subjects are half taught in Slovene and Hungarian each lesson. This means that schools need to employ two teachers (a Hungarian and a Slovene mother tongue speaker) for each school subject. On the other hand, the implementation of foreign languages to the first triad (ages 6 to 8) in Slovene primary schools is based on a CLIL approach, but is restricted to the regular English language lessons due to the Slovene Constitution, which requires that the language of instruction in the Slovene school system is Slovene (except for the minority schools). Given the diversity, such a flexible and comprehensive approach to implementing CLIL may have both strengths and weaknesses (Coyle, 2008). Some researchers see its strength in integrating content and language learning in varied, dynamic and relevant learning environments built on ‘bottom-up’ initiatives as well as ‘top-down’ policy (Coyle, 2006). On the other hand, its potential weakness lies in the interpretation of this ‘flexibility’, meaning that any kind of approach incorporating content in language lessons makes them a CLIL programme. Furthermore, an oversimplified perspective of CLIL implies that any variation on this type of programme will enhance both the target language competence and content capacities to a much greater extent than a regular communicatively based language teaching (Bruton, 2013; Gajo, 2007). CLIL research studies, particularly those in favour of CLIL programmes, have also been heavily criticised for their lack of: (1) transparency in selecting students for CLIL and non-CLIL programmes; (2) transparency in the interpretation of research results, depending very often on researcher interests;

148

Par t 2: Empir ical Studies Using Mixed Methods

(3) providing appropriate size samples of participants; (4) incorporating pretesting stages of CLIL and non-CLIL groups; (5) involving and comparing similar CLIL and non-CLIL groups of students in language proficiency, motivation, etc.; and (6) rigorous research substantiated by evidence on young foreign language learners involved in CLIL programmes in Europe and worldwide (Eurydice, 2012; Wei & Feng, 2015). It is therefore essential to consider Meyer et al. (2015: 546) warning, which he iterated together with his co-authors in a recent study: ‘In order for CLIL to earn its rightful place in the pedagogic arena of contemporary and future curricula, it has to demonstrate rigorous theoretical underpinning, substantiated by evidence in terms of learning outcomes and capacity building. The emergence of CLIL as a distinct field of enquiry suggests that whilst relevant research in the field of bilingual education can inform and guide CLIL enquiry, new avenues of research need to be explored.’

The Aims of the Study The current study focuses specifically on identifying the idiosyncrasies of implementing CLIL during regular English lessons at primary school level in a Slovene town of approximately 30,000 inhabitants and to provide a more comprehensive picture of its potential benefits and weaknesses by analysing quantitative and qualitative data acquired during the experiment. The present study sets out to contribute to existing CLIL studies involving young English as a foreign language (EFL) learners by addressing the following research questions: (1) How does a CLIL approach enhance the English language proficiency of 10-year old EFL learners in a country with a rather high exposure to the English language? (2) How does a CLIL approach impact on young EFL learners’ attitudes to English language instruction and English language learning? (3) How do young EFL learners perceive a CLIL approach?

Methodology Participants The participants consisted of four year 5 classes. There were 68 learners in the experimental group, 38 male and 30 female students. The control group involved 58 learners, 29 male and 29 female students. Altogether, 126 Slovene EFL 10-year-old learners participated in the study that was conducted over three months at two primary schools in the same town in Slovenia. All students had been learning English as a compulsory school subject for about a year and a half and had never participated in a CLIL

CL IL: A Panacea for Young ELL s?

149

programme. However, it should be emphasised that exposure to English language media is rather high, as all TV programmes in English are subtitled, except cartoons, which are often dubbed. The experimental group (CLIL group) and the control group (non-CLIL group) consisted of two classes each and were taught by two primary classroom teachers specialised to teach English to young learners. Both females had 15 to 20 years teaching experience and their language proficiency was at B2/C1 level. The teacher of the experimental group had a thorough training in CLIL, while the other teacher was trained in the communicative language teaching approach.

The procedure The experiment took three months (March to May 2013). The control group (two year 5 classes) had regular English lessons three times a week following the goals set by the Slovenian National English Language Curriculum for Primary School using an accredited textbook. The experimental group (two year 5 classes) was taught the content of non-language school subjects for year 5, such as science and technology, art, music etc. in English during their regular English language instruction. No English language textbooks were used. Prior to the experiment, teachers and school advisors thoroughly studied all the school subject curricula for year 5 in order to select relevant goals and content areas, which could be taught in English in CLIL lessons guided by the 4Cs framework. Two content areas were selected: animals and weather, which were used to meet the content goals of the following school subjects: science and technology, society (history and geography), music and art. These two topics were selected for several reasons. First, both topics are regularly studied in science and technology and society in year 5. Second, animals proved to be the most favoured reading topic by year 5 Slovene students in another study conducted by Bodlaj (2013). Third, both topics provide a number of possibilities for the students to construct their own meaning and develop relevant skills without a teacher being the only transmitter of the knowledge. Finally these two topics were also taught during regular English language lessons in the control group (see Appendix 9.1 for an example of a weather lesson plan).

Research design Prior to and at the end of the experiment, the students of both groups took a standardised English language proficiency test The Cambridge Young Learners English test (CYLE), level Movers. A similar approach to testing foreign language proficiency was used in the CLIL study designed by Alonso et al. (2008). According to the publishers, the CYLE test is aligned to Common European Framework of Reference for Languages (CEFR) Level A2. It is freely available online, but has not been used by students and teachers in Slovenia,

150

Par t 2: Empir ical Studies Using Mixed Methods

hence making it suitable for use in the current study. The test consisted of five listening comprehension tasks (25 test items), and six reading comprehension and writing tasks (40 test items). The oral tasks were not administered due to the lack of resources. The research team did not have financial resources to design, pilot and standardise test tasks, or to employ external interlocutors and oral examiners who would administer oral test tasks with each student separately so that reliable results could be obtained. The students’ attitudes to English language learning and the English language instruction were measured through a questionnaire consisting of two questions: Do you like learning English? and How much do you like English language lessons? The questionnaire was administered before and after the experiment to both groups of students. In the experimental group, the students were asked to assess each of the 22 CLIL tasks immediately after they had been completed. The students had to complete a four-level Likert scale on how much they liked each of the tasks. They also had to write relevant arguments for their decisions. At the end of the experiment, a group of four students was interviewed to identify the main issues related to the experiment. Their answers were transcribed and coded, and developed into an evaluation questionnaire consisting of 12 statements. These referred to the use of ICT, posters, oral presentations, online dictionaries, online information search engines, alternative assessment instruments, and the opportunity to learn and revise the content of other school subjects.

Results Analysis and interpretation of results How does a CLIL approach enhance the English language proficiency of 10-year old EFL learners? The learners’ listening comprehension skills were assessed by Cambridge Movers in both groups before and after the experiment. As Table 9.1 shows, results for the two groups were equivalent for the listening comprehension Table 9.1 Listening comprehension test results before and after the experiment

Listening comprehension test before the experiment Listening comprehension test after the experiment

Group

Mean

SD

EG CG EG CG

14.213 13.836 19.19 17.78

5.7901 4.688 6.687 5.481

Notes: SD – standard deviation; EG – experimental group; CG – control group.

CL IL: A Panacea for Young ELL s?

151

Table 9.2 The percentage of learners who improved their results on the listening comprehension test Listening comprehension

EG (%)

CG (%)

Total

No improvement Improvement Total

13.4 86.6 100

24.1 75.9 100

18.4 81.6 100

Notes: EG – experimental group; CG – control group.

test and we may assume that there were no systematic differences between the two groups. According to the results in Table 9.2, 75.9% of students in control group (CG) improved their listening comprehension skills while in experimental group (EG) 86.6% did so. Although there is 10.7% difference between the groups, the chi-squared test does not confirm any significant difference between the experimental and the control group (χ2 = 2.37, p = 0.123). The learners’ reading comprehension and writing skills in English were also assessed by Cambridge Movers in both groups before and after the experiment. As Table 9.3 shows, the two groups were very much equivalent in their results on the reading comprehension and writing test and we may assume that there were no systematic differences between the two groups. According to the results in Table 9.4, 31% of students in CG improved their reading comprehension and writing skills while in EG, 57.4% did so. Table 9.3 Reading comprehension and writing test results

Pre-reading comprehension and writing test Post-reading comprehension and writing test after the experiment

Group

Mean

SD

EG CG EG CG

19.7 16.8 19.8 16.8

10.3 8.7 10.8 7.7

Notes: SD – standard deviation; EG – experimental group; CG – control group.

Table 9.4 The percentage of learners who improved their results on the reading comprehension and writing test

No improvement Improvement Total

EG (%)

CG (%)

Total

42.6 57.4 100

69.0 31.0 100

54.8 45.2 100

Notes: EG – experimental group; CG – control group.

152

Par t 2: Empir ical Studies Using Mixed Methods

There is 26.4% difference between the groups and the chi-squared test confirms a significant difference between the experimental and the control group (χ2 = 8.752, p = 0.003).

How does a CLIL approach impact on young EFL learners’ attitudes to English language instruction and English language learning? Prior to the experiment, the learners’ attitudes in both groups were measured by a short questionnaire, which included two questions: (1) How do you feel about learning English (in general)? (2) How do you feel about learning English at school? After the experiment, the students filled in a similar questionnaire consisting of two questions: (1) Did you like English lessons during the last three months? (2) If so, please rank how much you enjoyed them during the last three months. Considering the results in Table 9.5, the two groups expressed very similar attitudes to English language instruction; therefore, we may assume that there were no systematic differences between the two groups. After the experiment, 58.8% of the experimental group learners reported that they liked English lessons very much (an increase from 50.0% before the experiment), while for the control group learners there was no reported change. After the experiment, there was a drop of 17.7% among the experimental group students who reported that they quite liked English lessons, while the drop among the control group students was less than half that of the experimental group (8.7%). The number of experimental group learners who did not like English lessons had increased by 8.8% after the experiment, a result very similar to the difference expressed among the control group learners (8.6%). Table 9.6 illustrates how many students in each group expressed more positive attitudes to English language lessons at the end of the experimental period, while the learners’ improvement in their attitudes is shown in Table 9.7. Of the experimental group, 29.4% of learners improved their attitudes to English language instruction, while in the control group only 1.7% improved their attitude. The chi-squared test confirmed that there is a significant difference between the experimental and the control group (χ2 = 17.277, p < 0.0001). Table 9.5 Learners’ attitudes to English language instruction prior to the experiment

I like it very much. I quite like it. I don’t like it. Total

EG (%)

CG (%)

Total

50.0 45.6 4.4 100

46.6 46.6 6.9 100

48.4 46.0 5.6 100

Notes: EG – experimental group; CG – control group.

CL IL: A Panacea for Young ELL s?

153

Table 9.6 Learners’ attitudes to English language instruction after the experiment

I like it very much. I quite like it. I don’t like it. Total

EG (%)

CG (%)

58.8 27.9 13.2 100

46.6 37.9 15.5 100

Total 53.2 32.5 14.3 100

Notes: EG – experimental group; CG – control group.

Table 9.7 Learners’ improvement in their attitudes to English language instruction

Improvement No improvement Total

EG (%)

CG (%)

Total

29.4 70.6 100

1.7 98.3 100

16.7 83.3 100

Notes: EG – experimental group; CG – control group.

The next aim of the study referred to the students’ attitudes to English learning and the results before the experiment had taken place are reported in Table 9.8. Prior to the experiment, more than half of the students in both groups stated that they liked learning English very much and only about 5% did not like it. Overall, we can say that both groups shared a very similar attitude to English language learning, for example, there was a 1.7% difference between the students of both groups who liked English learning very much and a 0.9% difference between those who quite liked learning English. According to the results in Table 9.9, 26.5% of students in EG improved their attitudes to English language learning while in CG 3.4% did so. There is a 23.1% difference between the groups and the chi-squared test confirms a significant difference between the experimental and the control group (χ2 = 12.42, p = 0.000). Table 9.8 The students’ attitudes to English language learning prior to experiment EG (%)

CG (%)

60.3 35.3 4.4 100

58.6 36.2 5.2 100

I like learning English very much. I quite like learning English. I don’t like learning English. Total Notes: EG – experimental group; CG – control group.

Total 59.5 35.7 4.8 100

154

Par t 2: Empir ical Studies Using Mixed Methods

Table 9.9 Learners’ improvement in their attitudes to English language learning

Improvement No improvement Total

EG (%)

CG (%)

Total

26.5 73.5 100

3.4 96.6 100

15.9 84.1 100

Notes: EG – experimental group; CG – control group.

How do young EFL learners perceive a CLIL approach? In order to gain some insights into this question all students of the experimental group were asked to express their attitudes to each of the 22 CLIL activities (see Table 9.10). Activity 1, watching a cartoon about Madagascar, was ranked as the most preferred activity among all the students (mean = 3.9). There was only one Table 9.10 The students’ attitudes to 22 CLIL activities No.

CLIL activity

Mean

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)

A cartoon on Madagascar Information search on the internet Searching for animals on the web zoo map Watching a video clip on weather Using an online dictionary Weather song A handout ‘How many sunny days?’ Playing a game ‘What’s the weather like in …?’ Reading a story about an argument between the wind and the water Designing a weather forecast Designing a poster on animals Making a plan for one’s own zoo and describing it Following daily weather conditions and filling in the weather table Learning a song ‘There is thunder’ Learning rhymes about winds Listening to the radio weather forecast for Europe and drawing weather symbols on the map of Europe A handout ‘Weather report’ Learning the rhyme about an elephant Watching a video clip and drawing symbols on the world map Reading the article ‘The wind brings the weather’ Reading the article ‘Clouds and the rain’ Reading English books on animals

3.9 3.7 3.5 3.5 3.4 3.4 3.4 3.4 3.3 3.3 3.3 3.2 3.1 3.1 3 3

(17) (18) (19) (20) (21) (22)

3 2.9 2.9 2.9 2.9 2.8

CL IL: A Panacea for Young ELL s?

155

student who did not like this activity. The second most preferred activity was the Information search on the internet (mean = 3.7). The third place is shared by two activities, Searching for animals on the web zoo map and Watching a clip on weather (mean = 3.5). The activities that students liked least involved Reading English books (mean = 2.8), two activities involving reading English articles (mean = 2.9) and Drawing symbols according to the visual and audio input (mean = 2.9) as well as Learning rhymes (mean = 2.9). For each activity, students had to give arguments/reasons why they liked or disliked it. In this way the researchers were able to obtain qualitative data on the students’ perceptions of each activity, as well as other important elements of the (foreign language) learning process. For example, Table 9.11 shows how students perceived three different activities, the first with a mean of 3.5 (out of 4.09), the second with an average mean (3.2) and the last with the lowest mean of all activities (2.8). The explanations provided by students for each of 22 activities were further categorised into three dimensions following the proposed model developed by Marsh et al. (2001): (1) communicative (language) dimension; (2) content dimension; and (3) cognition dimension, omitting the dimension of culture given the young students’ metalinguistic and metacultural age development and their lack of intercultural experiences as well as the time constraints in the administration of the study. Most of the arguments for why the students liked the CLIL activities may be associated with the dimension of content or learning (43%). This explanation is closely followed by the dimension of cognition (41%), while the dimension of language (communication) relates to only 16% of the reasons why students liked or disliked a particular activity (see Figure 9.1). After the experiment, the CLIL group students’ perceptions were assessed by a questionnaire consisting of 12 statements, which the students had to tick according to their level of agreement. Table 9.12 shows how the students perceived individual issues of the experiment. If the 12 statements are divided into three groups, we may observe that one third of the statements (1–4) referred to the use of multimedia and collaborative work, which the students liked using/practising and would like to do more often. The second third (5–8) encompasses activities that involved successful processes of learning and assessment, while the final third of statements (9–12), which the students agreed with the least, involved the perceived level of the students’ learning outcomes. As can be observed, the students seem to be rather critical of their learning achievements and are fairly reluctant to express their appreciation for revising the content of nonlinguistic subjects through EFL (29.9%). The chi-square test for equal proportions (Table 9.13) shows that the number of students who appreciated a CLIL approach amounts to 77.6%, which is statistically significantly higher than the hypothetical 50% (χ2 = 30.2; p < 0,05). From this, it can be assumed that most of the students appreciated the CLIL approach.

156

Par t 2: Empir ical Studies Using Mixed Methods

Reading English books

Making a plan for one’s own zoo and describing it

Watching a video clip on weather

Why did you like the activity?

Why didn’t you like the activity?

No answer

Table 9.11 Students’ arguments why they liked or disliked the activities

% 4.4 3.0 3.0 1.5 1.5

% 0

1.5

Answer I like watching TV. Interesting. I revised the vocabulary. Fun. I was successful. I’d like to go to the Weather room. You could use your own imagination. I learnt a lot.

% 28.3 23.9 22.4 6.0 4.5 1.5

Answer Boring. I didn’t understand well. I don’t like watching TV. Childish. Too fast speech.

19.7

I didn’t feel like it.

6.1

19.7

4.6

It was fun. It was interesting. We could choose animals and where in the zoo they would live. We could draw and write.

9.1 7.6 7.6

I didn’t like it because we had to write. Boring. I was lost. I didn’t like it because I don’t like zoos.

1.5

I’d like to be a zookeeper.

1.5

Because of new words I made many mistakes. I didn’t like it because we had to draw.

I liked it because I like animals. It was not difficult. We could do the task during the holidays. I like designing plans. I learnt a lot about animals. I learnt many new words. I like reading English books/ texts. It was interesting.

1.5

It was boring. I don’t like reading. I didn’t understand all.

15.2 10.7 6.0

There was more than in the textbook. I revised my knowledge.

4.6

4.5 3.0 1.5

1.5

1.5 1.5 1.5 22.8 12.1 10.6 12.1 1.5 1.5

We didn’t get enough information. I didn’t read it all.

1.5

I didn’t feel like it. I didn’t find out anything new.

1.5 1.5

1.5

1.5

CL IL: A Panacea for Young ELL s?

157

16% 41% 43%

Cognion

Content

Communicaon

Figure 9.1 The distribution of communication, content and cognition dimensions

Table 9.12 The students’ perceptions of the CLIL teaching approach Mean (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)

I’d like to watch more videos in English lessons. I liked having lessons in the computer classroom. I prefer making contributions in a group to preparing them on my own. During English lessons, I’d like to be able to use the internet more. While designing a poster, I learnt a lot about animals. While listening to my schoolmates’ presentations, I learnt a lot about animals. I liked learning and revising new subject matter through songs and rhymes. I prefer being assessed by giving a presentation on a particular topic to being orally tested. While designing a poster, I learnt to search for data in English books. I learnt to use online dictionaries. I learnt to search for data on the internet. In English lessons, I like revising the subject matter of other school subjects (e.g. science, maths).

3.6 3.6 3.5 3.4 3.3 3.3 3.3 3.2 3.1 3.1 2.9 2.9

Table 9.13 Chi-square test for equal proportions

CLIL approach not appreciated CLIL approach appreciated Total

Observed (n)

Expected (n)

Observed (%)

15

33.5

22.4

50.0

52

33.5

77.6

50.0

67

67

100

Expected (%)

100

Chisquare

p

20.433

0.000

158

Par t 2: Empir ical Studies Using Mixed Methods

Discussion of the Results Although there is some evidence that CLIL may prove very efficient in producing proficient foreign language learners (Lasagabaster & Sierra, 2010), there are also a number of research studies showing that CLIL students do not necessarily outperform non-CLIL ones (e.g. Bruton, 2011). CLIL students may even perform much worse, despite a much bigger language input provided for CLIL group students (Muñoz, 2006). The results of the current study show a significant improvement in the reading comprehension and writing skills for the CLIL group, but no statistical significance in the improvement of their listening comprehension skills. It is a rather complex endeavour to find relevant reasons for such results, as the foreign language learning process is not limited to school instruction only, but also to the out-of-school environment, especially in a country with a high potential provision of EFL exposure (Enever, 2011; Lindgren & Muñoz, 2013). However, it seems reasonable to discuss some potential underlying causes for the obtained results given the considerable amount of qualitative and quantitative data collected. One of the reasons for the significantly better reading comprehension performances of the CLIL students might be due to the higher level of reading input during the experiment. During English lessons, the control group was restricted to reading the texts in one of the accredited English language textbooks that tend to only include one short reading text per unit. These reading texts may be used either as lead-in stimuli or revision tools for a selected topic. For example, a text about a big town in the UK might motivate students to become interested in finding out more about the urban way of life and thus stimulate learners to explore vocabulary related to life in towns and cities. The course book texts are simplified and tend to be used in order to contextualise a specific grammatical or vocabulary feature. The subject matter of the texts is not designed to be studied for the purposes of solving some strictly content-related tasks but, on the contrary, to answer questions that assess mostly lexical and grammatical elements of the foreign language. On the other hand, the reading texts used in the CLIL group were authentic or semi-authentic, thus they were not written specifically for classroom use. They were also much longer than the texts in the textbooks and many of them were selected by the students themselves. As Table 9.12 shows, 60.6% of the CLIL group students appreciated reading authentic texts on animals due to their enhancement/self-perceived improvement in learning (cognition) and their positive attitudes and interests (like reading English books/it was interesting). However, more than one third of students (37.9%) did not appreciate reading these types of texts. The biggest group found the texts boring, some did not like the activity because they did not like reading in general or because they did not understand everything. Hui (2015) reports on a similar finding in Hong Kong English language immersion schools in

CL IL: A Panacea for Young ELL s?

159

which students whose English language proficiency was not at the appropriate level found the lectures boring and tiring, as they were forced to listen to the teacher but were not able to understand the content. It is therefore plausible that students accounted for their lack of understanding by describing it as boredom and dislike of reading. As for the results in the listening comprehension tasks, where no significant differences were observed between the experimental and control group, it cannot be claimed that the input of listening texts was significantly bigger in one of the groups. Both groups were exposed to the teacher using English for most of the teaching time. CLIL and non-CLIL group students listened to the recorded audio and videotexts during the lessons for a similar amount of time (on average about five minutes recorded listening input per lesson). Given the particular models of learning selected for each of these two groups, the input listening texts used during the instruction time reflected this with regard to authenticity. The non-CLIL group was exposed to the adapted listening texts only, while the CLIL group listened to or viewed off-air programmes that had not originally been designed for teaching purposes. As it was not possible to control the students’ out-of-school listening and watching of English language texts, which are freely available to all students as long they have a TV set at home and/or a computer with an internet connection (77% of Slovene households had an internet connection in 2014), it is impossible to predict how much the students’ listening skills were improved as a result of out-of-school exposure. However, a number of studies have shown that listening comprehension is enhanced by students’ exposure to foreign language input (Lindgren & Muñoz, 2013). When comparing the percentage of students who improved their language skills (reading, writing and listening), it is worth noting that all students improved their reading and writing skills by more than half (54.8%), whereas their listening skills improved by less than one fifth (18.4%) on average. It is, therefore, important to note that CLIL does not necessarily improve (young) foreign language students’ language proficiency across all the language skills to the same degree. Immersion programmes from second language contexts or bilingual school programmes where the language input tends to be relatively high cannot be simply transferred or transposed to CLIL programmes that take place during regular foreign language lessons, as these are highly context-specific (e.g. see: Marsh, 2002; Marsh et al., 1998; Wolff, 2002) and their generalisability from one situation to another is thus severely limited. As Gallardo del Puerto et al. 2009: 65) suggest: ‘[…] most of the immersion conditions […] bear little resemblance to the study of English through CLIL programmes in Europe, particularly in terms of the sociolinguistic and sociocultural context in which the L2 is learned and the authenticity of the input.’ Another research question this study attempted to answer refers to the impact of CLIL on the students’ attitudes to foreign language instruction

160

Par t 2: Empir ical Studies Using Mixed Methods

and the foreign language learning process. The current study reveals that there was a significant difference between the groups after the experiment in favour of the CLIL group. A further analysis of the qualitative data shows that the CLIL group appreciated cognitive and content-related activities much more than the language activities. This is, however, in opposition with the results of the most preferred activities in CLIL classes. The students namely least appreciated activities that involved most cognition, such as ‘revising subject matter of other school subjects in English’. This is a rather paradoxical finding that should alert researchers about the young students’ development/competences of their meta abilities (metacognitive, meta-linguistic, meta-cultural etc.) and their understanding of learning processes.

Conclusion CLIL as a teaching approach in early foreign language education should not be taken for granted and implemented in regular language lessons with the assumption that foreign language proficiency will automatically improve for all students. Some researchers (Apsel, 2012; Bruton, 2011) claim that CLIL is exclusive and mostly attracts high achievers. We believe that this is an imprecise claim that does not consider CLIL, as does any kind of (language) teaching approach, needs to take into account students’ age, lesson time, different abilities, competences, motivation and interests. For example, in this study, a CLIL activity was not appreciated by 15% of the CLIL group students because they had to observe a natural phenomenon regularly and fill in a chart, but they forgot to do it. This suggests that the students were unsuccessful not necessarily because of their inappropriate level of foreign language proficiency, but because an unpredicted element interfered. From this, it is evidently rather too simplistic to propose that it is only the level of foreign language proficiency that matters in CLIL instruction. On the other hand, the current study does support the theory of the need for comprehensible input (Krashen, 1987), as with all CLIL activities that the students disliked we can find the following explanations, such as did not understand (all) the words/text. Some reported of being lost or bored, which are typical answers of students who are not able to comprehend the (language and/or content) input at the set teaching pace (Hui, 2015). As the current study shows the CLIL group outperformed the non-CLIL group in reading/writing, but not in listening comprehension skills. One of the most likely reasons for such results might be the increased input of authentic reading texts and tasks the students had to read and complete. If this were the main reason we may wonder what would have happened if the reading load had been intensified in the non-CLIL (communicative language teaching – CLT) group. Would the CLT students have improved their reading

CL IL: A Panacea for Young ELL s?

161

skills as much as the CLIL group students or would the CLT reading tasks characterised by rather low cognitive demands and inauthentic purposes for reading make them lag behind? The quantitative data show that students enjoy being able to have fun, learn (a lot/something new), become interested in something, use their imagination etc. On the whole, cognitive and content dimensions are much more appreciated than aspects of language. Another question that needs further research can be framed as: Do the teaching activities which were appreciated by most students correspond to the activities which enhance their foreign language proficiency most? Surprisingly, three least appreciated activities among CLIL group learners involve reading processes; however, the results show that the reading input was an important factor in increasing their reading comprehension skills. Thus, another important finding of this study refers to the development of the students’ self-reflection competences on their own learning processes, which, in the case of young language learners, is far less likely to be fully formed, thus, students’ perceptions of their learning cannot be considered as decisive factors in designing CLIL programmes.

References Alonso, E., Grisaleña, J. and y Campo, A. (2008) Plurilingual education in secondary schools: Analysis of results. International CLIL Research Journal 1 (1), 36–49. Apsel, C. (2012) Coping with CLIL: Dropouts from CLIL streams in Germany. International CLIL Research Journal 1 (4), 47–56. Baetens Beardsmore, H. (2009) Language promotion by European supra-national institutions. In O. García with contributions by H. Beatens Beardsmore (eds) Bilingual Education in the 21st Century: A Global Perspective (pp. 197–217). Chichester: Wiley-Blackwell. Bloom, B. (1984) Taxonomy of Educational Objectives. Boston: Allyn and Bacon. Bodlaj, T. (2013) Bralna pismenost v angleškem jeziku učencev 5. razreda osnovne šole: diplomsko delo. [Reading Literacy in EFL among Year Students of the Primary School: A BA thesis]. Ljubljana: University of Ljubljana, Faculty of Education. See http://pefprints. pef.uni-lj.si/id/eprint/1618 (accessed 20 September 2016). Bruton, A. (2011) Is CLIL so beneficial, or just selective? Re-evaluating some of the research. System 39, 523–532. Bruton, A. (2013) CLIL: Some of the reasons why and why not. System 41, 587–597. Byram, M., Nicols, A. and Stevens, D. (eds) (2001) Developing Intercultural Competence in Practice. Clevedon: Multilingual Matters. Clegg, J. (2003) The Lingue E Scienze Project: Some outcomes. L’Uso Veicolare della lingua straniera in apprendimenti non linguistici. Centro Diffusione Comunitaire Quaderni 6. Coyle, D. (1999) Supporting students in content and language integrated learning contexts: Planning for effective classrooms. In J. Masih (ed.) Learning through a Foreign Language: Models, Methods and Outcomes (pp. 46–62). United Kingdom: Centre for Information on Language Teaching and Research, London. Coyle, D. (2006) CLIL in Catalonia, from Theory to Practice. APAC Monographs, 6. Gerona: APAC. Coyle, D. (2007) Content and language integrated learning: Towards a connected research agenda for CLIL pedagogies. The International Journal of Bilingual Education and Bilingualism 10 (5), 543–562.

162

Par t 2: Empir ical Studies Using Mixed Methods

Coyle, D. (2008) CLIL – a pedagogical approach from the European perspective. In N. Van Deusen-Sholl and N.H. Hornberger (eds) Encyclopedia of Language and Education. Second and Foreign Language Education (2nd edn), (Vol. 4) (pp. 97–111). New York: Springer Science and Business Media LLC. Coyle, D., Hood, P. and Marsh, D. (2010) Content and Language Integrated Learning. Cambridge: Cambridge University Press. Cummins, J. (2004) Using IT to create a zone of proximal development for academic language learning: A critical perspective on trends and possibilities. In C. Davison (ed.) Information Technology and Innovation in Language Education (pp. 105–126). Hong Kong: Hong Kong University Press. Dalton-Puffer, C. (2008) Outcomes and processes in content and language integrated (CLIL) learning: Current research in Europe. In W. Delanoy and L. Volkmann (eds) Future Perspectives in English Language Teaching (pp. 7–23). Heidelberg: Carl Winter. Dörnyei, Z. (2014) Researching complex dynamic systems: ‘Retrodictive qualitative modelling’ in the language classroom. Language Teaching 47 (1), 80–91. Enever, J. (ed.) (2011) ELLiE. Early Language Learning in Europe. London: British Council. European Commission (1995) Teaching and Learning: Towards the Learning Society. White Paper on Education and Training (Vol. 95COM). Brussels: European Commission. Eurydice Report (2006) Content and Language Integrated Learning (CLIL) at School in Europe. Brussels: Eurydice European Unit. See: http://www.indire.it/lucabas/lkmw_file/ eurydice/CLIL_EN.pdf (accessed 20 May 2016). Eurydice (2012) Key Data on Teaching Languages at School in Europe. Brussels: Education, Audiovisual and Culture Executive Agency. See: http://eacea.ec.europa.eu/education/ eurydice/documents/key_data_series/143en.pdf (accessed 15 May 2016). Gajo, L. (2007) Linguistic knowledge and subject knowledge: How does bilingualism contribute to subject development? The International Journal of Bilingual Education and Bilingualism 10 (5), 563–581. Gallardo del Puerto, F., Lacabex, G. and Lecumberri, M.L.G. (2009) Testing the effectiveness of content and language integrated learning in foreign language contexts: The assessment of English pronunciation. In Y. Ruiz de Zarobe and R. Jimenez Catalan (eds) Content and Language Integrated Learning: Evidence from Research in Europe (pp. 63–80). Bristol: Multilingual Matters. Hui, P. (2015) Boredom and fear in the undergraduate classroom: The medium of instruction controversy in Hong Kong. Inter-Asia Cultural Studies 16 (2), 253–262. Krashen, S.D. (1985) The Input Hypothesis. London: Longman. Krashen, S.D. (1987) Principles and Practice in Second Language Acquisition. New York: Prentice-Hall International. Lasagabaster, D. and Sierra, J.M. (2010) Immersion and CLIL in English: More differences than similarities. ELT Journal 64, 376–395. Lantolf, J. (ed.) (2000) Sociocultural Theory and Second Language Learning. Oxford: Oxford University Press. Lindgren E. and Muñoz, C. (2013) The influence of exposure, parents, and linguistic distance on young European learners’ foreign language comprehension. International Journal of Multilingualism 10 (1), 1–25. Lorenzo, F., Casal, S. and Moore, P. (2010) The effects of content and language integrated learning in European education: Key findings from the Andalusian sections evaluation project. Applied Linguistics 31, 418–442. Marsh, D. (ed.) (2002) CLIL/EMILE! The European Dimension: Actions, Trends and Foresight. Brussels: Potential Public Services Contract DG EAC. European Commission. See: http://www.sukrzsza.sk/docs/clil-emile.pdf (accessed 2 March 2016). Marsh, D. and Maljers, A. (2001) CLIL Compendium. Brussels: Directorate-General for Education and Culture of the European Commission (Socrates/Lingua). See http:// www.clilcompendium.com (accessed 2 March 2016).

CL IL: A Panacea for Young ELL s?

163

Marsh, D., Marsland, B. and Maljers, A. (eds) (1998) Future Scenarios in Content and Language Integrated Learning, EuroCLIC. den Haag: European Platform for Dutch Education. Marsh, D., Maljers, A. and Hartiala, A.K. (2001) Profiling European CLIL Classrooms: Languages Open Doors. Jyväskylä: University of Jyväskylä. See: http://lakk.bildung.hessen.de/netzwerk/faecher/bilingual/Magazin/mat_aufsaetze/clilprofiling.pdf (accessed 9 February 2016). McGuiness, C. (1999) From Thinking Skills to Thinking Classrooms: A Review and Evaluation of Approaches for Developing Pupils’ Thinking. Research Report 115. DfEE: HMSO. Met, M. (1998) Curriculum decision-making in content-based language teaching. In J. Cenoz and F. Genesee (eds) Beyond Bilingualism: Multilingualism and Multilingual Education (pp. 35–63). Clevedon: Multilingual Matters. Meyer, O., Coyle, D., Halbach, A., Schuck, K. and Ting, T. (2015) A pluriliteracies approach to content and language integrated learning – Mapping learner progressions in knowledge construction and meaning-making. Language, Culture and Curriculum 28 (1), 41–57. Mohan, B. (1986) Language and Content. Reading, MA: Addison-Wesley. Muñoz, C. (2006) Accuracy orders, rate of learning and age in morphological acquisition. In C. Muñoz (ed.) Age and the Rate of Foreign Language Learning (pp. 107–126). Clevedon: Multingual Matters. Pica, T. (1991) Classroom interaction, negotiation, and comprehension: Redefining relationships. System 19, 437–452. Swain, M. (2000) The output hypothesis and beyond: Mediating acquisition through collaborative dialogue. In J. Lantolf (ed.) Sociocultural Theory and Second Language Learning (pp. 97–114). Oxford: Oxford University Press. Van Lier, L. (1996) Interaction in the Language Curriculum: Awareness, Autonomy and Authenticity. New York: Longman Group Ltd. Vygotsky, L.S. (1978) Mind in Society. Cambridge, MA: Harvard University Press. Wei, R. and Feng, J. (2015) Implementing CLIL for young learners in an EFL context beyond Europe. Grassroots support and language policy in China. English Today 31 (1), 55–60. Wolff, D. (2002) Fremdsprachenlernen als Konstruktion: Grundlagen für eine konstruktivistische Fremdsprachendidaktik. Frankfurt/Main: Lang. Yin, R.K. (2006) Mixed methods research: Are the methods genuinely integrated or merely parallel? Research in the Schools 13 (1), 41–47.

Part 3 Longitudinal Perspectives Using Mixed Methods

10 The Dynamics of Motivation Development among Young Learners of English in China Yuko Goto Butler

Introduction Despite the increasing popularity of teaching English to young learners (conventionally defined as children from age 5 to age 12 or the end of primary school) worldwide, we do not fully understand what motivates these children to learn English. A number of researchers have expressed concern that although children show very high motivation to learn English initially, their motivation tends to wane by the end of their primary school years (e.g. Carreira, 2006; Jin et al., 2014; Kim, 2011; Mihaljević Djigunović, 2015; Nikolov, 1999). Others report that young learners’ motivation remains relatively stable throughout primary school (e.g. Heinzmann, 2013). And whether or not children’s motivation to learn English diminishes as they get older, the source of that motivation remains unclear. A number of factors – both contextual and individual – likely contribute in complicated ways to their motivational development. In this chapter, I report on a longitudinal study that used mixed methods to examine the dynamic trajectories of young learners’ development of motivation while paying close attention to specific contextual factors and individual differences. In particular, the study sheds light on the role of young learners’ socioeconomic status (SES) in their motivation to learn English as an international language of power. The study was conducted in an eastern coastal city in China that experienced rapid economic expansion and substantial disparities in SES among school-age children, a phenomenon that has been observed in many parts of East Asia in recent years.

Theoretical Framework Studies on motivation in language learning have undergone drastic changes in recent years. Gardner’s socioeducational model (Gardner, 1985), 167

168

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

which originated in Canada, has been very influential in the field. However, researchers, particularly those who work in the context of English learning as a lingua franca, have recently challenged that model’s major factor for learning a language, namely integrative motivation (i.e. one’s desire to be closer to or part of the targeted language community), for not being equally important across contexts. Among learners of English as a lingua franca (ELF) in the globalising world, the target language community may not be clearly identified. Instead, other types of motivation, such as Yashima’s (2002) international posture – one’s desire to communicate internationally – are considered more significant drives for students to learn English as a lingua franca. Another influential model of motivation is Deci and Ryan’s (1985) selfdetermination theory (SDT). Two major constructs underlying this model are intrinsic motivation and extrinsic motivation. Intrinsic motivation refers to one’s motivation for the sake of enjoyment. Extrinsic motivation refers to one’s motivation driven by utilitarian reasons. Assuming that intrinsic motivation is critical for learning and creativity, the SDT model focuses on how social and environmental factors that support our basic psychological needs – to feel autonomous, to have a high degree of self-perceived competence and to feel connected to significant individuals such as teachers and parents (relatedness) – facilitate our intrinsic motivation. Acknowledging that many of the behaviours that learners engage in are not necessarily intrinsically motivated, the model also considers how external motivators (external regulations in SDT), such as rewards and feedback, can be internalised as learners become more self-determined. In the past, extrinsic motivation was considered detrimental to intrinsic motivation; however, researchers no longer presuppose such a relationship between them (e.g. Ryan & Deci, 2000). More recently, Dörnyei (2003) has challenged the static view of learners’ motivation adopted by most studies, arguing instead for a more contextualised and dynamic approach to understanding motivation (the process model of motivation). Similarly, selfdetermination theorists have come to consider that intrinsic motivation does not necessarily reside in individuals but ‘exists in the relation between individuals and activities’ (Ryan & Deci, 2000: 56). A learner can be intrinsically motivated for certain activities but not others, or for the same activities but in different times or contexts.

Motivation Studies among Young Learners of Foreign Language Although there is plentiful research on motivation among adult language learners, when it comes to young language learners’ motivation, research is hard to come by. This may be in part due to a belief that young learners’ motivation and other affective dispositions, such as self-concepts and

Dynamic s of Mot ivat ion among Lear ners of English in China

169

attitudes, are not well developed and thus unstable. But new evidence indicates that young learners’ motivation and other affective dispositions are formed early and are more stable than once believed (e.g. Heinzmann, 2013). Considering that (a) a growing number of children are learning a second or foreign languages (including ELF) and (b) fostering their motivation and positive attitudes toward the target language is a major objective in many ELF programmes across the world, a better understanding of young learners’ motivational changes during the early years of their ELF learning would have significant practical and theoretical implications. A handful of studies available on motivation among young learners of ELF used cross-sectional or retrospective designs (e.g. Carreira, 2006; Kim, 2011; Nakata, 2009) and thus they have relatively limited capacity to examine developmental changes. However, two recent longitudinal studies (Heinzmann, 2013; Mihaljević Djigunović, 2015, 2016) do merit consideration. In a three-year study, Heinzmann (2013) investigated German-speaking Swiss primary school students’ motivation and attitudes toward English and French (from grades 3 to 5, ages 9/10 to 11/12, n = 552). Based on a survey of their motivation (intrinsic, extrinsic, effort and value), attitudes, anxiety, achievement self-concept and other affective dispositions (ethnographic vitality of the target language and stereotypes), a series of statistical analyses were performed. Heinzmann found that the Swiss children showed a very high motivation to learn English overall and that their motivation was ‘surprisingly stable with few substantial ups and downs’ (Heinzmann, 2013: 119). While the children exhibited a slight decline in intrinsic motivation, their lingua franca orientation (which is considered to be a type of extrinsic motivation) increased over the three years. When examining motivation at the individual level, the study also found that motivational changes at the individual level were minimal despite individual variations at the initial stage; in other words, ‘most primary school children go through the same motivational trajectory’ (Heinzmann, 2013: 118). Moreover, structural equation modelling (SEM) indicated that young learners’ self-concept (composed of perceived ease of English learning, competence, and expectation in Heinzmann’s study), turned to be the most influential direct and indirect variable on motivation. Whereas Heinzmann (2013) relied solely on quantitative analyses, Mihaljević Djigunović (2015) employed a mixed methods approach in her study. As part of a Croatian national longitudinal project (n = 208), she followed two groups of students for three years: young beginners (learners who started learning English from grade 1, ages 6/7) and older beginners (learners who started learning English from grade 4, ages 9/10). Their motivation, attitudes and self-concepts were first examined and compared quantitatively between the two groups primarily based on coded student interview data. Later, selected students’ cases (n = 6) were analysed qualitatively while incorporating additional data based on teacher interviews and classroom

170

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

observations. The quantitative analysis revealed that at the group level, the younger beginners showed higher motivation than the older beginners, but both groups showed decline in motivation over the three years. It is important to note that motivation in Mihaljević Djigunović was very holistically analysed; it was based on a three-level Likert scale on two items (‘How do you like learning English this year?’ and ‘How do you like learning new English words this year?’) along with an oral interview response to ‘Which is your favourite school subject this year?’ Her case study, however, revealed substantial individual but ‘idiosyncratic’ (Mihaljević Djigunović, 2015: 31) differences in the trajectory of motivation and self-concept. Interestingly, contextual differences did not make meaningful differences in the children’s motivational trajectories over the three years. The researcher concluded that the young learners’ motivation was ‘multifaceted’ rather than ‘monolithic’ (Mihaljević Djigunović, 2015: 32). Based on the same Croatian national longitudinal project mentioned above, Mihaljević Djigunović (2016) examined the trajectories of 24 young learners’ motivation and self-concept over time (from grade 5 to 8) and in relation to their speaking performance. As with her 2015 study, Mihaljević Djigunović measured self-concept based on the students’ response to an interview question, ‘Do you think you are just as good at English as your classmates, or worse, or better than they are?’ The results indicate that the motivation of the young learners declined from grade 5 to grade 7 and went up at grade 8 (note that the students were older than those in Mihaljević Djigunović, 2015), whereas their self-concept increased from grade 5 to 6 but gradually declined after grade 6. Correlational analyses showed that, whereas motivation did not correlate with the students’ oral performance, selfconcept did at all grade levels. In sum, the limited research available indicates inconsistent results in terms of developmental changes in motivation and other affective dispositions (e.g. self-concept). The inconsistent results may be partially due to the different measurements used to capture motivational constructs, as well as different contextual and individual factors among the studies. In any event, the mechanisms or reasons behind such developmental changes are not yet clear.

Socioeconomic Status and Young Learners of ELF Increasing disparities by SES among school-age children, as well as children’s unequal access to educational opportunities, are of concern to policy makers and educators around the world. China is no exception. With its rapid expansion of economic growth, China has witnessed substantial differences by SES with respect to financial and other resources that parents and schools can provide children (Dong, 2011; Hannum et al., 2010). Given

Dynamic s of Mot ivat ion among Lear ners of English in China

171

English’s strong association with economic power in China, and the language’s influential role in the nation’s meritocracy-based, exam-driven education system, it makes sense that English education would be uniquely affected by SES in China. A phenomenon often described as an ‘English divide’ by SES appears to affect even young learners in China. Despite concerns about SES disparities among young learners of English, research on the topic is extremely limited. The Early Language Learning in Europe (ELLiE) study, a collaborative project involving seven European countries, found that parental educational levels significantly correlated with young learners’ foreign language comprehension, including English (Enever, 2011). When it comes to the role of SES in young learners’ motivation, empirical research is almost non-existent. In my earlier study based on a crosssectional data set in China (Butler, 2014, 2015), I found that SES differently influenced young learners’ motivations as well as their English performance (grade 4, 6 and 8 students). However, the study had limited capacity to examine developmental changes due to its cross-sectional design. In the present study, therefore, I use longitudinal data to further explore the role of SES in the development of motivation. Specifically, I address the following questions: • •

Research Question 1: How do young learners’ motivation and selfconcept differ across time and by the learners’ SES backgrounds? Research Question 2: What were major reasons for changes (or no changes) in young learners’ motivation and self-concept?

A Longitudinal Mixed Methods Study in China Context and participants This study is part of a large longitudinal project examining the role of contextual factors (e.g. parents, teachers, peers and others) in learners’ English and motivational development. As noted above, it was conducted in an eastern coastal city in China (referred to as C-city). The original study followed three cohorts of students for three years: a primary school cohort (followed from grade 4 to 6); a middle school cohort (grades 7 to 9); and a high school cohort (grades 10 to 11).1 This chapter focuses on students in the primary school cohort (n = 201), who started learning English in grade 3 (ages 8 to 9). With the help of the local education bureau, the participants were drawn from two primary schools with distinct SES backgrounds. Although the cohort was not based on a random sampling from the entire city, participants sufficiently represented the whole SES spectrum of the city’s school-age children.2 The students received English lessons with fixed classmates and were

172

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

taught by the same English teachers at school throughout the three years. Twenty-eight percent of participants were migrant children from outside of C-city, mostly from nearby low-SES rural areas. Out of the 201 participants, 32 focus group students were selected through a stratified random procedure while controlling the gender and general English proficiency levels as measured by a locally administered test and their English teachers’ judgments. Detailed interviews (described below) were performed each year with the 32 students in the focus group.

Procedures This study employed mixed methods procedures. Of the five major purposes of mixed research methods identified by Riazi and Candlin (2014), my primary motivations for adopting a mixed approach in this study were complementarity and triangulation. The study is complementary in the sense that it uses quantitative data to understand young learners’ motivation at the group level while using qualitative data to understand their motivation at the individual level. This approach allows us to analyse children’s motivation at multiple levels. According to Riazi and Candlin, triangulation allows us ‘to seek convergence and corroboration between the results obtained from different methods, thereby eliminating the bias inherent in the use of a single method’ (Riazi & Candlin, 2014: 144). In the current study, quantitative analyses based on the whole group were conducted in order to understand the general trajectories of motivation at the group level (Research Question 1) and qualitative analyses of the focus group data were performed for more in-depth understanding of the phenomenon (Research Question 2). In the original project, in addition to using multiple outcome measures, we administered extensive surveys to the students and their parents, interviewed teachers and focus group students, and conducted a series of classroom observations. For the quantitative analyses in the current study, we focused on four motivation-related variables in the student surveys (extrinsic, intrinsic, self-perceived competence and anxiety) and an SES variable from the parent surveys (mothers’ education level was used as a measure of SES since other SES measures, such as father’s education level and household income, were all highly correlated among themselves). The motivationrelated variables in the student survey were composed of five items with a five-level Likert scale for each. Data were collected at the end of each academic year (three times in total). Table 10.1 summarises the variables used, including internal reliabilities at each grade level. The qualitative analysis was based on interviews with students in the focus group. Interviews were conducted in Chinese in a semi-structured format and lasted for approximately 15–20 minutes for each student every year.3 See Butler (2014, 2015) for detailed information on the survey items and student interviews.

Dynamic s of Mot ivat ion among Lear ners of English in China

173

Table 10.1 Variables used for quantitative analyses in this study

Motivation – extrinsic

Motivation – intrinsic

Self-concept (self-confidence)

Anxiety

SES – mothers’ educational level

Content and coding

Internal reliabilities (α)

(1) I want to be fluent in English because speaking English is useful when I go abroad; (2) I think that English will help me to get a good job in the future; (3) I want to get good grades in English; (4) I study English hard because English is an important subject at school; and (5) I study English hard in order to get into a good school. (1) I like to use English whenever I have an opportunity to do so; (2) I want to talk to foreigners in English; (3) I like studying English; (4) I like English TV programmes, videos, and movies; and (5) I like hearing English spoken. (1) I am confident when speaking English; (2) I am doing well in English; (3) I think that I can speak English better than my classmates; (4) I think I can be fluent in English; and (5) I am good at speaking English. (1) I get nervous when the teacher asks me to respond to her/him in English; (2) I am nervous if somebody speaks English too quickly; (3) I am nervous when I hear new words or unfamiliar words in English; (4) I get nervous when I take tests in English; and (5) I worry when I do not understand what others are saying in English. The original survey had seven levels of educational attainment but it was converted into three levels for the analysis (low, middle and high education groups): (1) up to middle school education; (2) up to high school education; and (3) post-secondary education

0.66 (G4) 0.54 (G5) 0.68 (G6)

0.75 (G4) 0.76 (G5) 0.82 (G6)

0.79 (G4) 0.83 (G5) 0.89 (G6)

0.81 (G4) 0.83 (G5) 0.81 (G6)

n.a.

Note: G4, G5, and G6 refer to Grade 4, Grade 5 and Grade 6 respectively. n.a. – not available.

Results General trajectories of motivation The group means of the motivation-related variables are plotted in Figure 10.1. A series of analysis of variance (ANOVA) with repeated measures were performed in order to see if there were any differences in the students’ average responses over time and by SES. The results are summarised in Table 10.2.

174

Par t 3: Longitudinal Perspec t ives Using Mixed Methods 4.8 4.6 4.4 4.2

Average rangs

4 3.8 3.6 Low Edu

3.4

Middle Edu 3.2

High Edu

3 2.8 2.6 2.4 2.2 Grade Grade Grade 4 5 6 Extrinsic M

Grade Grade Grade 4 5 6

Grade Grade Grade 4 5 6

Intrinsic M

Self-confidence

Grade Grade Grade 4 5 6 Anxiety

Figure 10.1 Students’ mean responses by time (grade levels) and SES (mother’s educational level)

As Table 10.2 indicates, students’ motivation, self-confidence and anxiety all showed significant differences across time at the group mean level, although the effect sizes were relatively small in general. Different from the previous findings in European contexts, students’ extrinsic motivation in this study was already generally high even in grade 4 (the second year of their English learning) regardless of SES backgrounds and the average went up in grade 5. Table 10.2 ANOVA with repeated measures (shaded if significant at the 0.05 alpha level) Extrinsic motivation F(2, 360) = 13.15, p = 0.00, η2 = 0.07(a) Time × SES F(4, 360) = 0.15, p = 0.96, η2 = 0.002 SES F(2, 180 ) = 2.46, p = 0.09, η2 = 0.03

Time

Intrinsic motivation

Self-confidence

Anxiety

F(2, 344) = 5.34, p = 0.005, η2 = 0.03 F(4, 344 ) = 0.83, p = 0.50, η2 = 0.01 F(2, 172) = 1.88, p = 0.16, η2 = 0.02

F(2, 360) = 6.12, p = 0.002, η2 = 0.03 F(4, 360) = 2.71, p = 0.03, η2 = 0.03 F(2, 180) = 2.78, p = 0.06(b), η2 = 0.03

F(2, 360) = 5.56, p = 0.004, η2 = 0.03 F(4, 360) = 1.36, p = 0.25, η2 = 0.02 F(2, 180) = 3.03, p = 0.05, η2 = 0.03

Note: (a) Partial η squared; (b) p-value did not reach to the significant level (α = 0.05) but it was an approaching level (0.06).

Dynamic s of Mot ivat ion among Lear ners of English in China

175

Students’ intrinsic motivation also increased from grade 4 to grade 5. Although there was no significant interactional effect, intrinsic motivation indicated a divergent tendency by SES at grade 6. Indeed, our analysis of the secondary school cohort in the larger study indicated that disparities by SES in intrinsic motivation became apparent at the secondary school level (Butler & Liu, 2015). Such a tendency of divergence by SES was more apparent in the selfconfidence of the primary school students. While young learners with lower SES backgrounds tended to indicate lower confidence in their English learning over time, students with high SES backgrounds tended to indicate increased confidence over time. Students’ anxiety also dropped by grade 5, but the difference by SES was persistent across the time; lower SES (L-SES) students tended to have higher anxiety than their high-SES (H-SES) counterparts.

Reasons for the motivational changes in the given context The transcribed interview data were analysed in the following order based on SDT: (a) identifying major external factors at each grade level; (b) examining the extent to which young learners’ psychological needs specified by SDT (i.e. autonomy, competence and relatedness) were met; (c) identifying major factors for and against developing a higher degree of self-determined behaviours; and (d) identifying major differences in the preceding three elements by SES. Table 10.3 summarises the main findings. It was evident that the role of parents in the development of the young learners’ motivation was substantial but that the nature of parental influence changed over time. A number of factors associated with the children’s SES also influenced their motivational development. I next discuss major factors that contributed to the children’s motivation to learn English at each grade level.

Grade 4 Since the fourth graders were only in their second year of English study, they were still excited about learning something new. They were curious about differences between English and Chinese, and they were excited about the new orthographies, sounds and vocabulary. The fact that English can be expressed using only 26 letters made them feel that English was easy to manage. In addition, there already existed a prevailing notion among the children that ‘说中文也是太简陋了。[Speaking only Chinese is not cool enough.]’ (ZY,4 L-SES). Even though they were still relatively new to English learning, the fourth graders were already able to clearly articulate the utilitarian importance of English learning, a fact that supports the high extrinsic motivation in the survey data. It was apparent from the interview data that the parents had a substantial influence on the young learners’ mind sets. Virtually all parents of focus group students, regardless of their SES background, explicitly

• Autonomy: developing but still heavily influenced by parents. • Competence: Content of English lessons was becoming challenging but still manageable and this led feeling of being competent in general. • Relatedness: Continued to have a strong feeling of being connected with parents. • There were emergent signs of self-determined behaviours. • Parents continued to be controlling and some children started feeling pressure. • Some high-proficient children compared themselves to others

• Autonomy: Relatively less developed; appeared to be heavily influenced by adults (parents and teachers). • Competence: Had high confidence in general. • Relatedness: Strong feeling of being connected with their parents and teachers.

• Self-determined behaviours were less development. • There were strong parental controlling behaviours (e.g. imposing their goals, comparing with others) in general, but the children did not perceive it as controlling.

Learners’ needs (based on SDT)

Major factors influencing self-determined behaviours

• Increasing pressure to score high in English in order to get into a good secondary school. • The content of English was increasingly more difficult. • Increased access to wider information (such as internet and newspapers, etc.). • Autonomy: Some indication for preference for more autonomous behaviours. • Competence: Some divergence in competence (some children increased and others decreased). • Relatedness: Continued to have a strong feeling of being connected with parents; increasing feeling of having a bond with peers with similar traits. • There were more evident sign of developing self-determined behaviours. • There was a growing pressure from increasingly exam-driven English lessons. • The content of English was becoming challenging (grammar teaching began) but still manageable. • High expectation from parents. • Increasing popularity of English pop culture.

• Relatively new to English learning (initial excitement). • High expectation from parents. • Strong effort-based notion of success in the society.

Grade 6 (ages 11–12)

Grade 5 (ages 10–11)

Major external factors

Grade 4 (ages 9–10)

Table 10.3 Main findings from the interview data

176 Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Differences by SES

• While parents, regardless of their SES backgrounds, told the children the importance of learning English, higher SES parents could provide the children with more direct assistance to learn English.

• The children already adopted the effort-based notion of success. • To be able to understand English cartoons on TV and internet and to sing English songs were a major source of intrinsic motivation. • Children were curious about the differences between L1 and L2.

•

•

•

•

•

who were slightly higher than themselves in order to motivate themselves. Started envisioning the role of English as being associated with their career goals. Effort-based notion manifested as the English lessons became more challenging. As their proficiency increased, they could enjoy English pop cultures on TV and internet. Children continued to be interested in analysing differences between L1 and L2. Children with higher SES continued to enjoy more opportunities to use English (e.g. extra English lessons, traveling abroad, watching English movies with their parents, etc.). • Higher SES parents began to be less controlling, which was perceived positively by the children. • Lower SES parents started lowering their expectations for their children’s success of learning English. • Higher SES parents’ ability to give their children more opportunities to use English was increasingly advantageous as the English lessons began to be more exam-driven.

• More children started feeling the pressure of being compared with somebody else. • Some parents became less controlling. • The role of teachers (both at school and outside school) began to be more important to motivate children as English became more challenging.

Dynamic s of Mot ivat ion among Lear ners of English in China 177

178

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

stressed the importance of learning English for their future, as exemplified in this student’s response: 反正我妈妈就说一定要学好英语要不然的话，要不然的话就得像她一 样 …. 她，妈妈现在只会说几个单词，特别简单。… 有的时候就算妈妈加 班，早上都要说上五遍或者六遍。[Anyway my Mom tells me that I must learn English well; otherwise, I will be like her who knows only a few very simple English words. … Sometimes my Mom tells me five-to-six times a day that it is important to learn English, even in the morning when she worked overtime the day before.] (SK, L-SES) The extent to which the fourth graders internalised what the parents told them is unclear; it is likely that they simply repeated what they had been told. Also, it is important to note that high-SES parents tended to provide their children with more direct assistance for learning English, such as checking their homework, helping them pronounce and spell words, helping them prepare for tests, and in some cases, talking to them in English. The highSES parents, including those who could not speak English, often watched English TV programmes and movies with their children (with subtitles). In contrast, while lower SES parents in our sample stressed the importance of English-learning for their future, they could not afford to or did not have time to actively participate in their children’s English-learning. Regardless of SES, it was very common for the Chinese parents to compare their children’s achievement (including English) with their children’s classmates, neighbours, relatives and other acquaintances. In our data, more than half of the children named specific individuals with whom they were compared by their parents. Curiously, however, none of the children seemed to feel pressured by such comparisons, at least at the fourth-grade level. This may have been due to the fact that they still perceived English learning as easy. Another possible explanation is that they believed that academic success is largely attributable to effort rather than innate talent. This notion, which may be called the effort-centred notion, was found to be very popular in Chinese culture (Stevenson & Stigler, 1992) and was confirmed to be the case among our parents (Butler, 2014). The children already subscribed to this notion at the fourth-grade level, if not earlier, as the following statement attests: 我觉得英语只要是认真学就不难 … 多看书，多听磁带 … 还有就是反复得 复习。[I think English is not difficult as long as I study it hard …. (I can) read books and listen to tapes as many times as possible …. and review (English lessons) repeatedly.] (FX, L-SES)

Grade 5 The content of English lessons became more challenging in grade 5 with the introduction of grammar instruction. English classes were still manageable

Dynamic s of Mot ivat ion among Lear ners of English in China

179

for most children in the focus group, however. As they picked up more English, the children began to enjoy cartoons and other English programmes on TV and the internet, and a desire to understand and enjoy English-medium pop culture appeared to be a major source of their intrinsic motivation to learn English. Some children were aware of and interested in regional differences in accents in English, while associating them with their own regional accents in Chinese. This indicates their growing metalinguistic awareness. Many fifth graders, regardless of their SES backgrounds, expressed a desire to go abroad, and this wish was a strong source of motivation to learn English. But going abroad for them was about more than wanting to know about different people and culture (i.e. more intrinsic-oriented); it also represented the ability to seek opportunities that they would not have in China (i.e. more extrinsic-oriented), as revealed by one child’s remark that ‘虽然也 是有机会，但是没有国外那么多，那么好。[Although there exists some opportunities, the chances in China are not as many and as good as those abroad.]’ (HH, H-SES). By grade 5, children heavily subscribed to the effort-centred notion. For example, SK (L-SES, whose mother kept telling her the importance of English in the example above), started acknowledging that she was not doing well reciting English in class, but she attributed her poor performance to laziness and told us that she was very good at reciting in Chinese classes. There was some sign of emergence of self-determined behaviours toward English learning among the children when they entered grade 5; in other words, they started reflecting on their own performance and acting upon those reflections. While the parents still largely imposed pressure on their children, some children started internalising their parents’ intentions, as we can see below: 因为要找份工作的时候人家都会问你英语几级几级，而不是问你中文几级。 所以我在想 … 因为爸爸妈妈干的工作比较苦，我爸爸妈妈就是说不希望我 想他们那样苦，找份好工作。[Because when you want to find a job, people will ask you about your English level rather than your Chinese level. So I’m thinking… Because my parents both have tough jobs, they don’t want me to have such a laborious job as theirs and they hope that I can find a good job.] (LL, L-SES) Some high-proficient students voluntarily started comparing themselves with others who were achieving at slightly higher levels than them in order to motivate themselves. They believed that a certain degree of pressure was a good motivational source. Other high-proficient children, however, criticised such comparisons, by saying, for example, ‘人比人，气死人。[Comparing with others only makes you upset (comparisons are odious).]’ (YH, H-SES). In another example of the emergence of self-determined behaviours and beliefs, more children associated their English-learning with their career

180

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

goals. By grade 5, more than half of the children could articulate their career goals. Even children whose career goals did not seem directly related to English, such as police officers and Chinese novelists, had reasons for wanting to study English. For example, LL (L-SES), who wanted to be a policeman, explained that he would be able to help foreign visitors on the street or join international investigations if he could speak English well. Although these ideas might have come largely from their parents, a couple of children indicated that they had obtained career-related information from the internet and newspapers. Children started to seek information related to English language and their careers from wider sources in grade 6.

Grade 6 English lessons at grade 6 were increasingly challenging and exam driven. Because students need to do well in English in order to get into a good secondary school, more children started feeling anxious about English learning in general. In our focus group, by sixth grade, 12 out of the 32 children clearly indicated that English was difficult for them. Importantly, however, perceiving the subject as difficult did not necessary mean that they disliked their English lessons. In fact, the majority of them still liked English class. It appears that English teachers, both teachers at school and in private programmes, played critical roles in maintaining the children’s motivation when they began to feel English was challenging. The children had positive attitudes toward English teachers who, unlike many teachers in other subjects, were more open-minded, introduced activities such as games and movies, and were good at using technology in class. For example, HX described her English classes as follows: 我觉得英语课，老师还蛮有趣的那种，她上课的话不是特别严那种，比较放 松的气氛。她也会适当地给我们看一些电影啊，玩玩游戏之类的。[I think my English teacher is interesting. She is not very serious and strict in class. The classroom atmosphere is relaxing. She often shows us some movie clips and lets us play some games]. (HX, H-SES) Some children started communicating with their peers through the internet and mobile devices, using English in some cases, in grade 6. They also tended to associate and work with peers with similar traits, such as similar school achievement levels, and peers appeared to become increasingly influential. While parents who had some command of English still tried to engage in their children’s English learning, the children often realised their parents’ limitations in helping them (e.g. evaluating a mother’s English pronunciation as not being good; pretending to work on exercise books their parents bought for them) and wished to be more independent. Meanwhile, some parents, notably H-SES parents, became much less outwardly controlling and encouraged their children’s autonomy, as in ST’s

Dynamic s of Mot ivat ion among Lear ners of English in China

181

remark ‘我爸爸妈妈说只要尽自己力就好) [My parents said that if I did my best it would be fine.]’ Similarly, SB (H-SES) also appreciated that his parents respected his autonomy and ‘表面上说不关注 [pretended not to care]’ too much about his test scores at school, while acknowledging that his parents actually did care in great deal. HH (H-SES) actively collected information from the internet and from his peers and realised that being able to read English would give him access to a much broader range of news and information than if he relied only on Chinese sources. Although H-SES parents started having less direct control over their children’s English learning, their ability to give their children opportunities to use English outside of school (e.g. travelling abroad, sending children to private English conversational lessons, etc.) gave their children an advantage over their lower SES counterparts, particularly as English lessons at school became more exam-driven and less communicative, and therefore provided fewer opportunities for students to actually use English at school. Finally, while a majority of H-SES parents continued to have high expectations for their children’s abilities to acquire English, it seems that lowerSES parents started lowering their expectations at the sixth-grade level. This drop in parental expectations came at the same time that lower-SES students began to lose confidence in their English learning abilities.

Discussion This longitudinal study investigated the development of English-learning motivations, self-concept (self-confidence) and anxiety during the primary school years in China, paying specific attention to parental SES factors. By combining interview data with survey data, the study aimed to understand the general trajectories (i.e. group-level tendencies), as well as more specific mechanisms of motivational change, by identifying major underlying factors based on SDT. The study found that young learners’ motivations, self-concept and anxiety changed over time. Unlike the European cases reported previously (e.g. Heinzmann, 2013; Nikolov, 1999), the Chinese children in this study subscribed to utilitarian reasons for learning English at an early stage. Parents had a substantial role in this process. In general, the Chinese parents in this study, regardless of their SES background, exhibited strong controlling behaviours (e.g. setting a goal by comparing their child with others), as reported by the children. At least at the fourth-grade level, the children appeared to handle such controlling behaviours well; they did not seem to fully internalise parental pressures to succeed. As the children became older, however, they started showing greater preference for autonomy, for collecting information from wider sources including the internet, and for flocking with peers with similar traits (e.g. having similar academic achievement levels).

182

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Towards the end of the children’s primary school years, some higher SES parents started adjusting to their children’s changing needs by asserting less control and this appeared to foster their children’s autonomy and selfdetermined behaviours towards English learning. While the Chinese parents, regardless of their SES background, believed that English was important for their children’s future, higher SES parents had more resources to provide their children with opportunities to use English outside of the classroom, such as travelling abroad. The high SES parents’ greater resources became more advantageous for their children to maintain higher intrinsic motivation as English lessons at school became increasingly more exam-driven and less communicative; it appeared to be increasingly more difficult for children to maintain their motivation to use English. Teachers also played an important role in maintaining students’ intrinsic motivation when the students started experiencing difficulties. It is important to note, however, that intrinsic and extrinsic motivations are not necessarily separable, as our ‘going-abroad’ examples above indicate. Learners might want to go abroad for intrinsic reasons, such as to learn about a foreign language and culture, as well as for extrinsic, utilitarian reasons, such as to seek better opportunities. The complexity revealed by the qualitative data serve to reinforce a recent call for re-conceptualising motivational constructs, such as taking more contextualised, fluid and multi-layered approaches to motivations (e.g. Mihaljević Djigunović, 2016). Notably, by the end of the primary school years, lower SES parents appeared to lower their expectations for their children’s success for learning English. Such differences in parental behaviours by SES, at least in part, explain the growing gaps in young learners’ self-confidence by SES. It seems that lower SES parents and students fell into a vicious circle; parents lowered their expectation and children lowered their self-confidence. Perhaps the children’s drops in self-confidence led to the parents’ lowering their expectation. The two mind sets appeared to be mutually influential, however. What triggered such drops in students’ self-confidence and parents’ expectation was not totally clear. The drops might have been related to, at least in part, secondary school placement results, which came out in grade 6. (The secondary schools in China are overtly stratified by students’ academic achievement as well as SES, which are correlated.) My final point relates to the research methodology employed in this study. I used a mixed methods approach primarily for complementarity and triangulation purposes. In future studies, it would be beneficial to explore the other purposes of mixed research methods that Riazi and Candlin (2014) discussed, including development, initiation and expansion. When it comes to development, the current study did not fully integrate what was obtained in the previous year into planning the following year’s data collection. For example, I could have used the results of interview data to inform item development for the following year’s surveys, or used the results of quantitative

Dynamic s of Mot ivat ion among Lear ners of English in China

183

analysis for identifying interview questions in the following year. Mixed methods can also be used for initiation purpose, in order to ‘uncover contradiction and paradox, with consequent recasting of questions or results from one method in the light of those obtained from another’ (Riazi & Candlin, 2014: 144). In the current study, for example, I could have obtained qualitative data from the focus students’ parents. These data might have provided insights that would not have been fully captured by the survey data. Or it might have contradicted survey data results or offered different perspectives on the children’s interview data. Lastly, the expansion purpose of mixed methods research means ‘to extend the breadth and depth of inquiry by using different methods to study different components of a programme, such as its outcome and process’ (Riazi & Candlin, 2014: 145). Like previous research on this topic, the present study used the results of quantitative analyses as the outcome of students’ motivation and used qualitative data to understand the process of the motivation formation. In the future, we might flip the roles of quantitative and qualitative methods or integrate both methods throughout the study.

Conclusion This study was one of the first longitudinal investigations of the development of motivation among young learners of English in China. The study focused on parental SES, an influential variable over young learners’ motivation and learning that has been largely unexplored. At the group level, the Chinese children’s changes in motivation, self-confidence and anxiety were relatively minor during the primary school years (judging from the relatively small effect sizes), but we could also see signs of divergence between highand low-SES students by the end of the primary school years. Qualitative analysis of interview data uncovered some of the underlying mechanisms of the children’s motivational development, as well as contextual complexities observed at the individual level. In future studies, researchers might benefit from embracing the other purposes identified by Riazi and Candlin (2014) for conducting mixed methods research. Doing so might provide a more robust understanding of the dynamic trajectories of young learners’ motivation development.

Notes (1) The high school cohort was followed for two years. (2) Their parents’ educational levels ranged from ‘no schooling’ to ‘graduate school degree’ and their household annual income ranged from ‘less than RMB 30,000’ to ‘more than RMB 220,000.’ (3) Interview questions varied according to the students’ responses. Example questions included: ‘Do you like English?’, ‘Why do you like (or not like) English?’, ‘How

184

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

important is it for you to learn English?’, ‘Why is it important (or not important) for you to learn English?’, ‘Do your parents help you with learning English?’, ‘What do they say to you about your English learning?’, ‘Do you feel pressure from your parents (with respect to English learning)?’, ‘Is there anybody whom you admire?’, ‘How do you feel when you are compared with somebody?’, ‘Do you like English classes?’, ‘Are there any English teachers that you admire?’ and ‘Do you think that you are good at learning English?’ The interview also contained several questions concerning their friendships and social networks in English learning and motivation, which were not covered in this study. (4) The students’ pseudonym initials are used.

References Butler, Y.G. (2014) Parental factors and early English education as a foreign language: A case study in Mainland China. Research Papers in Education 29 (4), 410–437. Butler, Y.G. (2015) Parental factors in ‘children’s motivation for learning English: A case in China. Research Papers in Education 30 (2), 164–191. Butler, Y.G. and Liu, Y. (2015) The role of peers in the development of motivation for studying English. Paper presented at the 2015 American Association for Applied Linguistics (AAAL) Conference, 21–24 March 2015, Toronto, Canada. Carreira, J.M. (2006) Motivation for learning English as a foreign language in Japanese elementary schools. JALT Journal 28 (2), 135–158. Deci, E.L. and Ryan, R.M. (1985) Intrinsic Motivation and Self-determination in Human Behavior. New York: Plenum. Dong, J. (2011) Discourse, Identity, and China’s Internal Migration: The Long March to the City. Bristol: Multilingual Matters. Dörnyei, Z. (2003) Attitudes, orientations, and motivations in language learning: Advances in theory, research, and applications. Language Learning 53 (1), 3–32. Enever, J. (ed.) (2011) ELLiE: Early Language Learning in Europe. London: British Council. Gardner, R.C. (1985) Social Psychology and Second Language Learning: The Role of Attitude and Motivation. London: Edward Arnold. Hannum, E., Park, H. and Butler, Y.G. (eds) (2010) Globalization, Changing Demographics, and Educational Challenges in East Asia. London: Emerald. Heinzmann, S. (2013) Young Language Learners’ Motivation and Attitudes: Longitudinal, Comparative and Explanatory Perspectives. London: Bloomsbury. Jin, L., Liang, X., Jiang, C., Zhang, J., Yuan, Y. and Xie, Q. (2014) Studying the motivations of Chinese young EFL learners through metaphor analysis. ELT Journal 68 (3), 286–298. Kim, T.-Y. (2011) Korean elementary school students’ English learning demotivation: A comparative survey study. Asia Pacific Education Review 12 (1), 1–11. Mihaljević Djigunović, J. (2015) Individual differences among young EFL learners: Age- or proficiency-related? A look from the affective learners factors perspective. In J. Mihaljević Djigunović and M.M. Krajnović (eds) Early Learning and Teaching of English: New Dynamics of Primary English (pp. 10–36). Bristol: Multilingual Matters. Mihaljević Djigunović, J. (2016) Individual learner differences and young learners’’ performance on L2 speaking tests. In M. Nikolov (ed.) Assessing Young Learners of English: Global and Local Perspectives (pp. 243–261). New York: Springer. Nakata, Y. (2009) Intrinsic motivation in the EFL school context: A retrospective study of English learning experiences in Japanese elementary schools. The Journal of Asia TEFL 6 (4), 263–291. Nikolov, M. (1999) Why do you learn English? Because the teacher is short: A study of Hungarian children’s foreign language learning motivation. Language Learning Research 3 (1), 33–65.

Dynamic s of Mot ivat ion among Lear ners of English in China

185

Riazi, A.M. and Candlin, C.N. (2014) Mixed-methods research in language teaching and learning: Opportunities, issues and challenges. Language Teaching 47 (2), 135–173. Ryan, R.M. and Deci, E.L. (2000) Intrinsic and extrinsic motivations: Classic definitions and new directions. Contemporary Educational Psychology 25, 54–67. Stevenson, H.W. and Stigler, J.W. (1992) The Learning Gap: Why Our Schools Are Failing and What We Can Learn from Japanese and Chinese Education. New York: Touchstone. Yashima, T. (2002) Willingness to communicate in a second language: The Japanese EFL context. The Modern Language Journal 86 (1), 54–66.

11 Young Italian Learners’ Foreign Language Development: A Longitudinal Perspective Lucilla Lopriore

Longitudinal Studies in Early Language Learning Several research studies and reports have investigated the implementation of foreign language (FL) learning at primary level in different contexts within and outside Europe (Butler & Lee, 2006; Drew & Hasselgreen, 2008; Enever, 2011; Field, 2013; Garton et al., 2011; Hasselgreen, 2000; Johnstone, 2000; Lopriore, 2014; Lopriore & Mihaljevic Djigunovic, 2011a, 2011b; Mihaljevic Djigunovic & Lopriore, 2012; Moon & Nikolov, 2000; Nikolov, 2009a, 2009b; Nikolov & Mihaljevic Djigunovic, 2011; Nikolov et al., 2007; Rixon, 1999, 2013). The rate and order of children’s second language acquisition and the effects of early language learning on children’s cognitive development were also investigated (Benvenuto & Lopriore, 2000; Cenoz, 2003; Lopriore, 2001). Fewer projects have monitored young learners’ FL achievement at specific transition levels or in time through longitudinal studies in order to monitor the development of specific FL skills first developed at an early age (Enever, 2011; Hill & Launder, 2010; Johnstone, 2000; Roessingh & Elgie, 2009). Because of the very nature of early language learning, research studies investigating that field require appropriate methodologies and accurate measurement tools. Researching early language learning means taking into consideration different developmental factors affecting the learning process, such as learners’ age, children’s fast physical development, as well as cognitive and affective aspects. Thus, longitudinal studies appear the most appropriate form of research for monitoring children’s language learning development while also taking into account different physical, cognitive and affective variables (Enever & Lopriore, 2014). Longitudinal studies are widely 186

Young It alian Lear ners’ Foreign L anguage Development

187

acknowledged as extremely valuable and informative in educational research. Recent reports on several longitudinal studies (Butler, 2017; Enever & Lopriore, 2014; Ortega & Iberri-Shea, 2005; Rixon, 2013) have highlighted their contribution to second language acquisition (SLA) research and to the monitoring of FL learners’ pace and pattern of learning and literacy development since its first onset. Longitudinal studies are needed because they help monitor changes that ‘unfold gradually (but not linearly) over time and that are long-lasting’ (Ortega & Iberri-Shea, 2005: 9). Given the pivotal function of issues such as time and learning, such studies may help to clarify the effects of early language learning as children’s cognitive and emotive development stages unfold almost parallel to their second language achievement. This perspective brings new and different foci to the research area. However, one difficulty with longitudinal studies can be the loss of some learners in the follow-up studies. Learners, particularly primary school children, move on to more senior school levels, for example: from primary to middle schools, often located in different areas. Alternatively, they may move to other schools or to other locations. Nonetheless, on occasion it may be possible to trace the development in time of part of the original sample in terms of the development of their achievement, of their language skills, and of their attitudes and motivation.

A Longitudinal Perspective: An Insight from Two Studies A longitudinal research study on the development of English as a foreign language skills of a group of young Italian learners was carried out at two different points in their school career: during the first four years of the primary cycle (aged 6–10 years old) and a few years later, during the last year of the middle school (aged 13 years old). The study reported in this chapter is based upon the Early Language Learning in Europe Project (ELLiE, Enever, 2011) and on the Italian ELLiE follow up study carried out with a sample of the ELLiE Italian learners (Lopriore, 2012, 2014, 2015). In the first study, the tasks used had been specifically developed and administered at the end of each school year in order to measure learners’ FL comprehension and production. In the follow-up study, the tests used were standardised tests measuring the students’ listening, reading and writing skills. Post-listening questionnaires were used in both studies in order to identify perceived difficulties and strategies used by learners. Classroom practice and assessment procedures were the object of specific interviews with the teachers involved in the study. The follow up study aimed to investigate the changes in FL competence and motivation development of this group of young Italian learners aged from 6 to 13 years old, acknowledging that during this period learners

188

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

undergo many relevant cognitive changes, often resulting in increased awareness of their use of language and developing proficiency in FL use. It is important thus to try to understand what happens to FL learners as they grow older and move from one level to the next (Lopriore, 2015). The following section includes a brief background on the ELLiE study, providing an introduction to the Italian follow-up study reported here.

The ELLiE project The ELLiE study (2006–2010) is a longitudinal transnational project that involved over 1500 learners in seven European countries. The project was innovative in nature and scope because of its longitudinal and mixed method approach. It aimed to investigate the development of young FL learners, as well as to gain insights into the factors that influence both learners’ perceptions of language learning and their responses in school contexts, where a limited amount of time for FL study is available (Enever, 2011). In Italy the ELLiE study was carried out between 2006 and 2010 in one region, Latium, located in central Italy. Eight primary classes participated in the study, providing a cohort of 186 Italian young learners (aged 6–10 years old). A subsample of 49 learners drawn from all the classes also participated in additional, more detailed language tasks and interviews. Among the tasks used to monitor children’s language progress and achievement in the ELLiE study (aged 6–10 years old), specific listening, speaking and vocabulary tasks were created and administered during the four school years. Observation grids were used to observe classroom activities, attitude questionnaires (Smileys) were administered every year to the whole class and focal learners were interviewed about their motivation and their self-perception. Specific interviews with teachers were carried out once or twice a year in order to identify their attitudes and beliefs about FL teaching and learning and to discuss their views about FL learning and FL materials. Parents’ questionnaires were administered twice in the study. Teachers were observed during lessons at least three times during the school year (Lopriore, 2012).

The Italian ELLiE follow-up study The follow-up study was carried out between 2012 and 2013 with a subsample of the original ELLiE Italian group, in order to investigate ELLiE Italian learners’ FL learning development eight years after they had started learning English and four years after the end of the ELLiE study. The idea for the investigation was triggered by several factors. In June 2013, the original ELLiE students would have completed their middle school and would have been moving into high school in the following year. Transitions are usually considered influential factors in learners’ development.

Young It alian Lear ners’ Foreign L anguage Development

189

Table 11.1 Longitudinal study structure Italian cohort

Sample

Learners’ age (years)

Tools

ELLiE study (2006–2010)

186 learners in eight classes 48 focal learners (six in each class)

6 to 10 years

Follow-up (2012–2013)

76 learners in four classes

12 to 13 years

• Listening tasks (four years) and post-listening questionnaire (two years). • Reading tasks and post-reading questionnaire • Speaking tasks (three years) • Vocabulary tasks (three years) • Attitude (Smileys) four years • Motivation (interviews) four years • Teachers’ interviews (three years) • Parents’ questionnaire (2 years) • All language tests from a KET level A2 • Listening tests and post-listening questionnaires • Reading and writing tests + posttest questionnaire • Attitudes (questionnaires and interviews) • Teachers’ interviews • Parents’ questionnaire

In Italy, the transition from primary to middle school usually influences learners in terms of their FL study for two main reasons: FL in the middle school syllabus is a separate subject officially inserted in the curriculum and another foreign language is officially introduced, parallel to English, as part of the school curriculum. The follow-up study was therefore important, since it could provide relevant information about learners’ FL development and about changes occurring in their achievement, attitudes and motivation. Table 11.1 illustrates the structure of the sequential studies and the tools used. The teaching and learning of all skills were evaluated in both studies. In this chapter, the development of listening, reading and writing are explored, in relation to students’ attitudinal development and use of English both in and out of school over time.

Aims In this chapter, we investigate the development of the young learners after eight years of studying English. In order to provide a broad understanding of their FL development we relate their achievements in listening, reading

190

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

and writing to their perceptions of themselves as language learners, their perceived difficulties, and their exposure to English outside the classroom.

Research design The Italian follow-up study (2012–2013), with a view to tracking the development of the original Italian ELLiE sample, retraced 76 of the original Italian ELLiE participants in four different middle school classes. All 76 belonged to the larger Italian cohort (186 students) that had been observed during the first four years of the primary cycle (aged 6–10 years old). They thus became the main cohort of the follow-up study during the last year of the middle school cycle (aged 12–13 years old). In this new study the research design was mainly organised based on the type of tools and the main procedures used in the original ELLiE study, with modifications and adaptations reflecting the learners’ age and the new school context. For example, the Smileys were not used and longer interviews were carried out instead; the language tests used were similar in form to those the learners were exposed to at school. The follow-up study was conducted through: • • • •

•

Classroom observation, similar to the ELLiE project, in order to identify such factors as: activity types, learners’ participation, materials used, amount of exposure to L2, input source and the teacher’s role. Individual interviews to analyse learners’ attitudes to L2 and their selfperception as language learners (similar interviews were also used in the ELLiE project). Direct measures of reading, writing and listening at A2 level using the Key English Test (KET) published by Cambridge ESOL. In the ELLiE project only listening and some reading skills had been measured. Post-listening, post-reading and writing questionnaires administered immediately after the test, aimed at investigating learners’ response to the tasks. In the ELLiE project, post-listening questionnaires were administered only in a few cases, as in the Italian study. Interviews with the class teachers as in the original ELLiE project.

Language tests The language test used in the 2013 follow-up study was designed to widen the original construct used in the ELLiE study, including the comprehension of longer aural and written texts, as well as the ability to understand and use vocabulary in context. The original ELLiE study tests were mainly geared at finding out young and very young learners’ ability to comprehend and use vocabulary, to understand short aural and written texts, as well as to evaluate their own FL learning. The 2013 test on the contrary, was based upon a more traditional test format for two main reasons: first, the learners’ familiarity with that test type and second, the age factor – specifically their

Young It alian Lear ners’ Foreign L anguage Development

191

cognitive development and their accumulated foreign language experience. Standardised tests such as the one used in 2013, geared at a specific Common European Framework of Reference (CEFR) level, are often based upon a variety of skill-based task types that allow learners to refer to different skills and learning areas usually present in course-books traditionally used in middle schools.

Learners’ motivation and self-perception Interviews and questionnaires were employed to collect data on learners’ motivation and self-perception. The main themes reflected those of the original ELLiE study, asking about their: • • • •

motivation to study English; self-perception as FL learners and as compared to other classmates perception of difficulties in test performance; out-of-school exposure to English.

In both studies, the learners’ out-of-school exposure to English, the FL classroom practices and the assessment procedures traditionally used, were also the focus of interviews with the learners and with the teachers involved. Learners’ parents filled in a questionnaire on their perception of their children’s achievement and the exposure to English at home or in the nearby environment (Lopriore, 2014, 2015).

Preliminary Findings If compared to the ELLiE project final findings in 2010, the follow-up study in 2013, showed some interesting, partly predictable changes, because of the learners’ age and of their extended learning experience. These changes clearly emerged in learners’ motivation, in their preferences for FL classroom based activities, in learners’ increased exposure to English outside the classroom walls, as well as in their ability to measure their own FL achievement.

Learners’ awareness of their learning experience In 2013, learners showed clearer awareness of their competence in the FL and during the interviews they were able to refer to their primary school experience and discuss differences in the middle school system that either hindered or promoted success in the FL. During the interviews, learners’ memory of their learning experience – particularly after the age of eight – was quite vivid and enhanced their reflections upon their own learning. They mentioned, for example, how difficult it was for them to limit their use of English to the official timetable (three hours per week). They complained about the teaching approach used at middle school, reporting that it did not

192

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

provide enough opportunities for all class students to use English as, on the contrary, had occurred in the primary school. The notion of success in English was always associated in the interviews with the need for daily exposure to English and regular use of it. This opinion is most probably linked to their exposure to contexts where understanding and speaking are central for communicating in a FL, as experienced during school exchanges or when they go abroad with their families.

Learners’ perception of the FL When asked how they perceived the foreign language in terms of difficulties they encountered and whether they now found some skills easier than others, the students’ answers were quite similar and they all referred to the classroom context. As shown in Figure 11.1, most students found all four skills and using words neither easy nor difficult, while speaking was regarded as a difficult skill by over 40% of the students, whereas listening is apparently considered as one of the least difficult skills.

Learners’ self-perception The ability of young learners to identify their own level of competence was amazingly precise during the primary cycle. Their responses usually matched their test results as well as those of their classmates. This high degree of accuracy in the ELLiE study was most probably due to their full exposure to learning English since they were fully immersed, almost physically, in this new experience. They were capable of observing and capturing all the tiniest

Figure 11.1 Learners’ perceptions of the FL difficulties at age 13

Young It alian Lear ners’ Foreign L anguage Development

193

details. Since they were usually involved in activities as a group, they were also able to observe their own and their classmates’ interactions and performances. In contrast to this earlier high degree of perceptiveness, when asked to evaluate their own competence in the follow-up study, most middle school students were less precise and used ‘average’ as an answer and they were quite surprised by their results in the tests. When asked to compare their competencies to those of their classmates, the middle school learners were more prone to generally define themselves as ‘worse’ than their friends. It was interesting to discover that when asked how they knew that they were ‘worse’ or ‘better’ than their friends, they chose to use the official FL grades provided by their teachers as indicators of their competence (Figure 11.2). These responses suggest that possibly the middle school system does not provide any encouragement to learners to speak out about their abilities and may not offer opportunities for self-assessment.

How do you know? My homework I use English also out of school I'm good because I can speak… I hardly understand English… I am not good Difference in speaking I use English at home What I know I study less than my friends Out-of-school courses They speak more English than me I don't know My mates do not study enough We know the same things My grades 0

5

10 15 20 25 Number of students

30

Figure 11.2 Learners’ explanations for their comparison to their friends at age 13

Out-of-school exposure In order to better identify changes and improvement in the learners’ speaking and aural comprehension, it was important to investigate how frequently they were exposed to English out of the school context as previously in the ELLiE study. It was expected that as teenagers they would have had many more opportunities than when they were young children. The findings confirm that these teenagers were exposed much more to English

194

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Figure 11.3 Exposure to and use of English at home (age 13)

outside the classroom and in the home (Figure 11.3), particularly listening to English songs; also that they used videogames in English (Figure 11.4) very frequently and the internet almost for one hour per day (Figure 11.5). The internet is indicated by the students as the third main source of exposure to English at home after songs and videogames. They had also started watching films in English (Figure 11.6). The data tell us how pervasive the exposure to English is now at home, much more than when they were younger. It would be interesting to find out whether their results are actually being affected by all of this out-of-school exposure. Several students had been exposed to English outside the home, for example, when travelling abroad. Part of the questionnaire was designed to learn, as in the earlier ELLiE study, whether they had been abroad and how they had

Figure 11.4 Exposure to and use of English at home: Opportunities (age 13)

Young It alian Lear ners’ Foreign L anguage Development

195

Figure 11.5 Surfing the internet at home (age 13)

performed in English during their trip. The questionnaire responses were quite interesting – especially when compared to the original ELLiE study, when very few learners had had the opportunity to use English outside the classroom. Almost everybody in the follow-up study was able to say that they had used English very well, although some of them were able to identify that they had experienced difficulties in oral interactions (Figure 11.7).

Learners’ language test results The language tests (Cambridge ESOL, KET) were administered during the second week in March 2013, almost two months before the end of the school year (June 2013). The Key (KET) for Schools qualification is a commercially available English language test designed for use with school age learners (Cambridge ESOL, 2016). The tests were intended to measure whether the students had reached a full A2 level. Post-listening and post-reading questionnaires were administered immediately after the listening and the reading sections of the

Figure 11.6 Exposure to and use of English at home: Watching films in English

196

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Figure 11.7 Exposure to and use of English out of home: Travelling abroad

language test. Prior to these tasks, learners were instructed on the format and purpose of the post-listening and post-reading questionnaires. In the listening tests only 40% of the students reached the A2 level and 60% were just below it (Figure 11.8). In the reading and writing tests the results were slightly lower (Figure 11.9). In the post-listening questionnaire, 14% of the students reported that they found it difficult and 2% very difficult, 35% neither easy nor difficult, 27% easy and 22% very easy. Overall, the listening test was not considered particularly difficult, but the learners found the last part more difficult and

Figure 11.8 Listening test results (age 13)

Figure 11.9 Reading and writing test results (age 13)

Young It alian Lear ners’ Foreign L anguage Development

197

as a matter of fact most mistakes were in this section. This section consisted of a conversation between two young people with questions focused on the spelling of a name and on the recognition of figures. Apparently, this type of focused attention on letters and figures caused most confusion among students. They identified unknown words and the speed of the recording as sources of difficulty. As for the reading test, 41% of the students found it neither easy nor difficult, 20% easy, 16% very easy and 17% difficult, while 6% found it very difficult, particularly the part on text completion. Even in the reading test, unknown words were identified as the cause of difficulties in understanding the text. Although the percentages of answers show a mismatch in students’ overall perception of the test difficulties, their results in both tests have definitely improved since primary school. Their overall level is slightly below the expected level at the end of middle school; that is, a full A2. As for their perception of the listening tests and the slightly better results than in the reading and writing tests, most probably the reasons lie in the increased exposure of the learners to spoken English, particularly outside school, as indicated from the data collected and from the parents’ responses to the questionnaires that further confirm this.

Teachers’ opinions on listening in the classroom Even if listening is almost always combined with speaking, in school learning contexts, listening is often assessed on its own with specific listening tasks. When first interviewed about the type of listening activities students were exposed to, the Italian teachers declared that they mainly used the listening activities provided by the course-books and, in the absence of recorded materials, they read the texts aloud themselves. They admitted, though, that they seldom made use of specific listening tests to measure children’s development of aural comprehension, while they more frequently used tasks to measure the children’s lexical competence and their reading and writing skills. One of the teachers used the international standardised tests for the students’ practice and sometimes also for testing them, but she admitted that the listening part was often skipped. If learners perceive listening as a minor skill, they will inevitably devote less attention to it.

Conclusions One of the most interesting findings in the Italian ELLiE follow-up study is the minor role played by listening (and speaking) in the FL classroom on the one hand, and on the other hand, the increased role of out-of-school exposure to English that plays a relevant role in learners’ aural comprehension development. These two aspects deserve greater attention in both

198

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

classroom activities and in teacher education courses. Some of the positive findings of the original ELLiE study were challenged by the learners’ transition to a different school system where most of what had been encouraged at primary level was undermined, particularly in terms of exposure to English in the classroom. Overall, the study reveals that positive results in the listening tasks seem to be in most cases related to and affected by learners’ positive attitudes to the study of English, by more awareness of spoken language as an interactive skill where speaking and listening sustain learners’ performance, but also by being more independent in their out-of-school exposure to the foreign language. The findings from this study have shown that after eight years of ‘studying’ English, learners’ awareness of listening difficulties is heightened, they can better identify indicators of proficiency, even if they still relate their success to the grades they get. They are much more exposed to English in out-of-school situations. They clearly identify ‘speaking’ and ‘knowing words’ as the key to success in aural comprehension and in their overall proficiency. From a pedagogical perspective, this study highlights the importance of a varied classroom, in which the out-of-school context may also play an important role. The children in the study actually provided some useful information as to the teaching and learning of English. When asked in the questionnaire about what they thought would have helped them improve their performance in English, almost all of them suggested several ways, implying further exposure to spoken English and revealing a profound awareness of the FL learning process. They also indicated the importance of fostering more group activities together with the involvement of all class members and the use of ICT. Monitoring a group of FL learners from when they were very young children to their adolescence may be a powerful and very revealing way of learning how language acquisition and learners’ self-perception develops over time during that age period. A longitudinal study offers the opportunity to provide an incredible amount of information useful also for teachers and material developers. If accurately planned and implemented, these types of studies may become highly significant sources of information and unveil aspects of teaching and learning that are seldom explored (Cenoz, 2003; Enever, 2011; Enever & Lopriore, 2014; Lopriore, 2014, 2015; Johnstone, 2000).

References Benvenuto, G. and Lopriore, L. (2000) La valutazione delle competenze in lingua straniera nella scuola elementare. Rapporto di ricerca. [Evaluation of foreign language competencies in primary schools. Research report]. Rome: Italian Ministry of Education. Butler, Y.G. (2017) The dynamics of motivation development among young learners of English in China. In J. Enever and E. Lindgren (eds) Early Language Learning: Complexity and Mixed Methods. Bristol: Multilingual Matters.

Young It alian Lear ners’ Foreign L anguage Development

199

Butler, Y. and Lee, J. (2006) On-task versus off-task self-assessments among Korean elementary school students studying English. The Modern Language Journal 90, 506–518. Cambridge ESOL (2016) Cambridge English: Key (KET) for Schools. See http://www.cam bridgeenglish.org/exams/key-for-schools/ (accessed 29 October 2016). Cenoz, J. (2003) The influence of age on the acquisition of English: General proficiency, attitudes and code mixing. In M.P. Garcia Mayo and M.L. Garcia Lecumberri (eds) Age and the Acquisition of English as a Foreign Language: Theoretical Issues and Fieldwork (pp. 77–93). Clevedon: Multilingual Matters. Drew, J. and Hasselgreen, A. (2008) Young language learner (YLL) research: An overview of some international and national approaches. Acta Didacta Norge 2 (1), 1–18. Enever, J. (ed.) (2011) ELLiE. Early Language Learning in Europe. London: The British Council. Enever, J. and Lopriore, L. (2014) Language learning across borders and across time: A critical appraisal of a transnational, longitudinal model for research investigation. System 45, 187–197. Field, J. (2013) Cognitive validity. In A. Geranpayeh and L. Taylor (eds) Examining Listening (pp. 77–151). Cambridge: Cambridge University Press. Garton, S., Copland, F. and Burns, A. (2011) Investigating global practices in teaching English to young learners. ELT Research Papers 11–01. The British Council & Aston University. Hasselgreen, A. (2000) The assessment of the English ability of young learners in Norwegian schools: An innovative approach. Language Testing 17 (2), 261–277. Hill, S. and Launder, N. (2010) Oral language and beginning to read. Australian Journal of Language and Literacy 33 (3), 240–254. Johnstone, R. (2000) Context-sensitive assessment of foreign language in primary (elementary) and early secondary education: Scotland and the European experience. Language Testing 17 (2), 123–143. Lopriore, L. (2001) La valutazione degli apprendimenti delle lingue straniere nella scuola elementare [Foreign language learning assessment at primary level]. In F. Gattullo (ed.) La Valutazione degli Apprendimenti Linguistici (pp. 215–244). Scandicci (FI): La Nuova Italia. Lopriore, L. (2012) Early language learning: Investigating young learners’ achievement in a longitudinal perspective. Rassegna Italiana di Linguistica Applicata 44, 147–166. Lopriore, L. (2014) Research into early foreign language learning in Italy: Looking back, looking forward. In J. Horváth and P. Medgyes (eds) Studies in Honour of Marianne Nikolov (pp. 100–118). Pécs: Lingua Franca Csoport. Lopriore, L. (2015) Investigating learners’ development in time. In S. Letica Krevelj and J. Mihaljevic Djigunovic (eds) UZRT 2014. Empirical Studies in Applied Linguistics (pp. 141–158). Zagreb: FF Press. Lopriore, L. and Mihaljevic Djigunovic, J. (2011a) Role of language exposure in early foreign language learning. In G. Szabó, J. Horváth and M. Nikolov (eds) UPRT 2010. Empirical Studies in English Applied Linguistics (pp. 1–18). Pécs: Lingua Franca Csopot, Lopriore, L. and Mihaljević Djigunović, J. (2011b) Aural comprehension and oral production of young EFL learners. In J. Horváth (ed.) UPRT 2011: Empirical Studies in English Applied Linguistics (pp. 83–103). Pécs: Lingua Franca Csoport. Mihaljevic Djigunovic, J. and Lopriore, L. (2012) Aural comprehension in young learners’ foreign language development. In L. Pon, V. Karabalic and S. Cimer (eds) Applied Linguistics Today: Research and Perspectives (pp. 51–65). Frankfurt: Peter Lang. Moon, J. and Nikolov, M. (eds) (2000) Research into Teaching English to Young Learners: International Perspectives. Pécs: University Press Pécs. Nikolov, M. (ed.) (2009a) Early Learning of Modern Foreign Languages: Processes and Outcomes. Bristol: Multilingual Matters. Nikolov, M. (ed.) (2009b) The Age Factor and Early Language Learning. Berlin: Mouton de Gruyter.

200

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Nikolov, M. and Mihaljević Djigunović, J. (2011) All shades of every color: An overview of early teaching and learning of foreign languages. Annual Review of Applied Linguistics 31, 95–119. Nikolov, M., Mihaljevic Djigunovic, J., Mattheoudakis, M., Lundberg, G. and Flanagan, T. (eds) (2007) Teaching Modern Languages to Young Learners: Teachers, Curricula and Materials. Strasbourg: Council of Europe. Ortega L. and Iberri-Shea, I. (2005) Longitudinal research in SLA: Recent trends and future directions. Annual Review of Applied Linguistics 25, 26–45. Rixon, S. (ed.) (1999) Young Learners of English: Some Research Perspectives. London: Longman. Rixon, S. (ed.) (2013) British Council Survey of Policy and Practice in Primary English Language Teaching Worldwide. London: The British Council. Roessingh, H. and Elgie, S. (2009) Early language and literacy development among young ELL: Preliminary insights from a longitudinal study. TESL Canada Journal 26 (2), 24–45.

12 Employing Mixed Methods for the Construction of Thick Descriptions of Early Language Learning Eva Lindgren and Janet Enever

Introduction In this chapter we use a mixed methods approach for the longitudinal study of three young learners who were developing their attitudes towards learning a foreign language and towards themselves as language learners. In the light of evidence from this six year study we review their achievements across a number of dimensions. Both the individual and social nature of young children’s learning have long been priorities for research, since the influential work of Piaget (1936, 1959), Inhelder and Piaget (1958) and Vygotsky (1978, 1986). Piaget’s work on how the child develops as an individual precipitated a shift in approaches to teaching and learning in European primary school classrooms, resulting in a stronger focus on the needs of each individual learner. This brought movement away from whole class, lockstep teaching, with individual workbooks and work cards becoming popular in some contexts. Vygotsky’s work on children’s cognitive development identified the importance of a social environment, arguing that development occurs through interaction within a social context. This focus on the importance of interaction significantly influenced the reorganisation of classroom layouts at primary school level, facilitating pair and group work, the appearance of classroom nature corners and book corners, together with areas set aside for role play activities, reflecting the increasing recognition that classroom talk contributed valuably to the learning process (Pollard & Tann, 1987; Wood et al., 1976). These influential theories of cognitive development have led us to reflect on the processes through which young children begin to construct their confidence and fluency as learners of second and foreign languages (L2/FL). 201

202

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

In an effort to shed some further light on the very individual and complex nature of language learning for young children, we report here on findings from a longitudinal case study of three learners, conducted over the first six years of compulsory schooling. Adopting a case study approach to this research has allowed us to look at each individual learner’s response to learning in far greater depth than would be possible in a larger scale study, while the longitudinal perspective also offers an invaluable opportunity to consider how the process of learning fluctuates and progresses over time. Longitudinal studies have long been recognised as providing evidence of how learners develop and change over time (Wells, 1986). Nonetheless, the investment of time and funding, to say nothing of the commitment of schools, individual learners and researchers to take part over a number of years can be difficult to maintain. With regard to second and foreign language development in the first few years of schooling, few studies have yet been able to take a longitudinal view. Among these, Butler (2017), Cenoz (2003), Lopriore (2017), Mihaljević Djigunović (2017), Nikolov (1999) and Orosz (2009), offer some useful insights. In this study, the opportunity to take a longitudinal perspective is available to us as a result of the data bank established during the four years of the ELLiE study (Enever, 2011). In this new study we build on the initial research, contributing further evidence collected from two subsequent years of data collection to provide an illustration of what has been achieved by learners in Sweden during six years of learning English (grades 1–6).

Research Background Longitudinal studies of children’s L2/FL development have often tended to focus on language development with rather less documentation of the child’s attitudinal development over time. With this in mind, in this section we briefly summarise findings related to oracy and reading skills development, including also research exploring aspects of attitudinal development. Where possible we have drawn from longitudinal studies. First, in a small-scale study using interactive tasks (spot-the-differences) with 10 year olds, Pinter (2007) demonstrated how repetition of an oral task can contribute to children engaging in more complex and creative language use. Through repetition, Pinter showed how children became more confident and more productive in their oral interaction. From this we can suggest that routine repetition of oral task types with young children may be effective in progressively helping to build oral fluency. Pinter notes particularly that the children ‘enjoyed the experience of speaking English in a spontaneous manner with each other’ (Pinter, 2007: 201), emphasising the significance of their attitudinal response to tasks. Mihaljević Djigunović and colleagues have looked specifically at children’s attitudes during FL development

Thick Descr ipt ions of Early L anguage Lear ning

203

(e.g. see Lopriore & Mihaljević Djigunović, 2011; Mihaljević Djigunović, 2014; Mihaljević Djigunović & Lopriore, 2011). Reporting on the longitudinal findings from the ELLiE study, Mihaljević Djigunović and Krevelj (2009) reviewed data from the Croatian sample, looking at the development of attitudes to learning English over a three-year period (age 7–10 years old). Their findings emphasised the dynamic and very individual nature of attitudinal development. In another recent study, Mihaljević Djigunović (2017) has extended her work to record the further development of attitudes, motivation, self-concept and achievement as students progress through grades 5–8 (age 10–14 years old) of the Croatian school system. Investigating the complex interaction of these factors, she identifies that all three affective factors fluctuate over time, noting that the trajectories of the fluctuations for these older learners have followed somewhat different pathways than those of learners who learnt English from grade 1 (age 6–7 years old) (e.g. Mihaljević Djigunović, 2012). This factor may prove to be of significance for the study under discussion in this chapter. A large-scale questionnaire study by Heining-Boynton and Haitema (2007) of children’s attitudes to learning French or Spanish in North Carolina (n = 7038) resulted in similar findings to those of Mihaljević Djigunović (2012) regarding dynamic and fluctuating attitudinal patterns. However, they found a notable decline in enthusiasm by the end of grade 5, with boys becoming quite negative while girls remained fairly positive overall. A subsequent follow-up telephone interview study of 13 students involved in the original research was conducted when they were aged 16–18 years old. All had continued to learn a FL and were positive about their early experiences of learning a FL with 12 of the 13 students reportedly using the FL outside school. One further study of significance here is a longitudinal study of motivation for 201 learners beginning English from grade 3 in China (Butler, 2017). In this, our attention is directed to parental influence which Butler relates to the family’s socioeconomic status (SES). In a context where high stakes testing occurs in grade 6 and where English is closely associated with economic power, SES is shown to be a significant factor in children’s motivation and the development of self-concept over time. Turning to the question of development of literacy skills in English as a foreign language (EFL), young learners make use of a number of resources to create meaning. Lindgren and Stevenson (2013) found that even though 11-year-olds wrote significantly shorter texts in EFL as compared with L1, they interacted with the reader and expressed themselves through a variety of strategies, such as using smileys, capitalisation, orthographic transcription (invented spelling) and conceptual simplifications. As in the development of literacy in L1, phonological awareness (Goswami & Bryant, 2004) plays a role when children, for example, develop principles for invented spelling in EFL (He & Wang, 2009). However, drawing on data from three cross-sectional studies comparing monolingual and bilingual children aged

204

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

4–7 years old, Bialystok et al. (2003) found no conclusive evidence that bilingual abilities enhanced ability in phonological awareness tasks. Their inconclusive results lead them to recommend further research and to suggest the need for ‘individualized literacy instruction depending on linguistic background’ (Bialystok et al., 2003: 42). Nikolov and Csapó (2010) examined the complexities of L1 and L2 reading performance for reading comprehension tasks and other related variables. This large-scale crosssectional study of Hungarian learners of English found significant correlation between L1 and L2 reading skills for grade 6 learners of English (aged 12 years old). For learners in later years of the study, the links were less strong, indicating that ‘L2 proficiency becomes more independent from L1 as proficiency increases over the years’ (Nikolov & Csapó, 2010: 326). The study discussed in this chapter aims to build on these earlier findings to provide some further insights of young children’s L2 oracy and literacy development through a discussion of three longitudinal case studies, which look particularly at the dynamic nature of language development. Significantly, we include the additional measure of a national testing tool bringing an external standardised measure to the equation – a feature that the original ELLiE data were not able to include.

Data Sources The six-year longitudinal data reported on here are collated from the following sources: the three-year ELLiE project (2007–2010), a one year preceding scoping study (2006–2007), and from two years of continuing data collection directly following on from the ELLiE project (2010–2012). In this chapter we look at linguistic achievements, including reception and production in speaking, listening, reading and writing. Further, we look at extralinguistic achievements with regard to attitudes towards learning a foreign language (FL) and to being a learner of a FL. We also discuss the sociocultural environment of the home in terms of parental support and exposure to English.

Instruments and procedure Two sets of instruments were developed and used to trace the progress in children’s learning of English as a second/foreign language: one set was directed towards linguistic achievements and another towards extralinguistic achievements and the home context. All the instruments were designed to be as age appropriate as possible, resulting in the administration of five different production tasks and four different receptive tasks over the six-year period (see Table 12.1 for an overview). The instruments were slightly adjusted over the period to conform to the learners’ age and

Thick Descr ipt ions of Early L anguage Lear ning

205

Table 12.1 Overview of instruments used over the six-year period to collect data of production and reception achievements, attitudes and context Production

Reception

Attitudes and context

Year 1

Oral: Vocabulary retrieval

Listening task I

Year 2

Listening task II

Year 4

Oral: Vocabulary retrieval, Restaurant task Oral: Vocabulary retrieval, Restaurant task Oral: Guess Who task

Year 5

Written: letter to a pen friend

Smiley questionnaire Interview Smiley questionnaire Interview Smiley questionnaire Interview Smiley questionnaire Interview Smiley questionnaire Interview

Year 6

Written: descriptive narrative (national test)

Year 3

Listening task III Listening task IV Reading task Listening task (national test) Reading task (national test) Listening task (national test) Reading task (national test)

language experiences. Descriptions of the oral tasks, the listening tasks for years 1–4, the year 4 reading tasks and the extralinguistic and the contextual instruments can be found in the ELLiE book (Enever, 2011) and will only be described briefly here. The other instruments will be described in more detail or with reference to other publications. All tasks were administered once a year towards the end of the school year (May/June). Listening, reading and writing tasks, and the smiley questionnaires were administered in the classroom with the whole class, while the oral tasks and interviews exploring attitudinal development were conducted individually with students.

Receptive tasks During each of the first four years a similar listening task was used. In order to adjust to children’s language development, a number of items were changed every year. In year 5, the results from the listening part of the national test were used. (In the present curriculum system a national test is administered in year 6 and in the previous it was administered in year 5. Coincidentally, the change between systems occurred in time for our classes to have a national test both in year 5, according to the previous system, and year 6, the new system.) Reading achievements were studied through a cartoon task in year 4 and via the national tests in years 5 and 6. In the national tests children are asked to read both shorter and longer texts and reply to multiple choice questions as well as giving open written answers.

206

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Production tasks During the first year, a vocabulary task was used where children were prompted to name as many items as they could from one picture of a zoo and one of a home environment. In the second and third years, children were asked to produce as many English words and phrases as possible, prompted when needed by domains such as colours, animals, school and the home. Another production task was also used during years 2 and 3, the restaurant task, in which children had the opportunity to interact with the researcher in a fictional restaurant role play. During year 4, children were asked to interact with the researcher in a ‘Guess Who’ game, prompted by a detailed picture of a classroom illustrating children involved in different activities. In years 5 and 6, written production was targeted through a personal letter to a pen friend in year 5 (Lindgren & Stevenson, 2013) and via the national test in year 6, in which students were given a 45 minute period to write an imaginary descriptive narrative about the future. The instructions read: Let’s get into a time machine and travel to the year 2112 – one hundred years from now! Write about what things are like in the year 2112. What is the same? What is different? What do you like about 2112? What don’t you like? These were followed by another eight prompts suggesting things to write about (Nationella prov i främmande språk, 2015). The letter task in year 5 required command of the present tense only, while the task used in the national test in year 6 required knowledge of the future tense.

Attitudes and context Following Mihaljević Djigunović (2017), we have focused on early English language learners’ attitudinal development over time. We looked at whether learners liked learning English in general, their attitudes towards speaking and listening, singing and learning new words in English. Our aim in the design of these instruments is to provide an account of how three learners’ attitudes towards learning English changed over time, revealing the dynamic and individualised nature of attitudinal development. We do not attempt to link this evidence to causal effect; however, we provide some contextual features that may be considered as contributory. Two instruments were used to follow students’ attitudes towards learning English and to being a learner of English, a smiley questionnaire and an interview. The smiley questionnaire included statements such as ‘How do you feel about learning English this year?’ with three or five smileys with different expressions to choose from (Enever, 2013: 23). Questions about vocabulary learning, playing, listening, singing in English, etc. were included.

Thick Descr ipt ions of Early L anguage Lear ning

207

The interview included questions about their favourite subjects, English lessons, homework, help at home, exposure to English etc., also giving students room to expand on their answers. The two instruments were used from year one through year five, in year six we could not use them due to time constraints in the classrooms (see Table 12.1 for full list of instruments and time of administration).

Participants This study includes a deep analysis of three learners: Tomas, David and Mikael (pseudonyms), whom we followed through the six years of primary school in Sweden. They were selected from a larger sample based on the results of the first year vocabulary retrieval task as one with a high score, one average and one low result. All three learners were native speakers of Swedish. It was a coincidence that all three learners were boys. David and Tomas were in the same urban primary class for the first three years of schooling with one teacher during the first year and another in years 2 and 3. The school is a lower primary school only, including school years 1 to 3, with around 100 students. In year 4, they moved to a larger primary school where they continued in different classes, with different teachers. They kept the same teacher during years 4 to 6. Mikael attended a rural primary school with around 150 students from years 1 to 6. As is common in the Swedish system, he changed teachers and classrooms after the first three years of primary school. All teachers were qualified generalist primary teachers, with English language levels approximately B1–B2 and some training in teaching English to young learners. English is mainly taught as a separate lesson, with a focus on oracy in years 1–3, including activities such as communicative games and occasional stories read by the teacher. In years 4–6 a coursebook is introduced and teaching is organised around it.

Analysis The data have been analysed both according to quantitative aspects such as scores on listening or reading tasks or the number of vocabulary items produced, and according to qualitative aspects such as the type of vocabulary and complexity of phrases. Listening and reading scores have been calculated as percentages of the maximum scores, while vocabulary was evaluated based on the totals students produced in the vocabulary retrieval tasks, together with the calculation of Giraud’s index to understand vocabulary diversity in the restaurant task, the Guess Who task and the written tasks. Noun phrase complexity was also analysed in these tasks. The data collected with the attitude and context instruments were analysed qualitatively for

208

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

content information about students’ attitudes towards learning English and being a learner of English, exposure to English and type and frequency of exposure and parental support in the home.

Findings Students’ home environments, attitudinal development and achievements are described below. Their achievements are also illustrated graphically in Figures 12.1 to 12.3 (see Appendix 12.1 for the underlying data). The linear representation in each graph should not be compared, but rather followed as an indication of how achievements developed over the six years. The dotted lines indicate points where a major change of instrument occurred. This occurred in the two different vocabulary retrieval tasks between years 1 and 2, the change from oral to written production in year 5 and the change from ELLiE instruments to national tests in year 5.

Tomas The home Tomas reports that his parents are happy that he is learning English. They are happy to help him with English but he says that he rarely needs any help. There are English books in the home, but it is not until year 4 that he reports reading any English books at home when he reported reading Harry Potter. Throughout primary school, he reports that he hears English at home every day via TV or the internet-based video provider, YouTube. From year 4 and onwards, he also uses English on a daily basis at home while playing computer games, for example, World of Warcraft, often in combination with speaking or chatting with co-players via Skype or MSN. In year 5 he says he uses English ‘almost all the time’. He watches English films without subtitles and plays computer games for about five hours a day. He spends less than one hour a week doing his English homework and he does not particularly enjoy it. He has met people who do not speak Swedish and communicated with them in English. This felt ‘normal’ to him and he thought it was ‘fun’.

Attitudes towards learning English At the beginning of primary schooling, in year 1 and year 2, Tomas was neutral towards learning English. Classroom observations from year 2 confirm that his interest and involvement varied a lot depending on activity. From year 3 onwards, however, his opinion changed and he reported that he liked it more than previously. In particular, his attitudes towards learning new words, speaking and playing games became more positive over time. In

Thick Descr ipt ions of Early L anguage Lear ning

209

year 3 he did not report that he disliked any activities they did in the English classes, but one year later he found listening tasks boring. Towards the end of primary, in year 5, speaking was his preferred skill and he thought it was more difficult to write, mainly because he did not know what to write about. Throughout primary school, English has not been Tomas’ favourite subject; instead he reports reading and physical education (PE) to be particularly enjoyable. His views towards English as a school subject are neutral; it is ‘like any other subject’.

Attitudes towards being a learner of English Commenting on difficulty, Tomas did not consider learning English to be particularly difficult or easy. In year 3 he felt it was easy, and in years 4 and 5 he considered English to be very easy. Tomas feels he learns English at about the same speed as other pupils. He says that he already knows a lot. Generally he finds learning English is easy because he already knows most of what is done during the lessons. He says that the lessons changed in year 3 because now they have the English box and a smart board in the classroom. He feels he knows everything already, he finds English easy and in Year 5 he says that he would like to do ‘more difficult things’.

Achievements Tomas was the strongest of the three learners according to the results of the vocabulary retrieval task in year 1. His production was advanced already from the start of primary education, indicating that he already had receptive experience of English, possibly in his pre-school years or the home environment. In the restaurant task in year 2 his response to the prompt to ask for the toilet, was ‘He must go to toilet … she …’ and in year 3 he formed an elaborate question to the same prompt: ‘Can you show me where the toilet are, where the toilet is eller [or] bathroom’. In response to the prompt to order food for him and his family he produced ‘ice-cream’ in year 2 and ‘I want a hamburger and my brother wants a pizza and my dad wants a pizza’. During this year, he developed his knowledge not only of the noun phrase to more consistently include determiners, but also about the verb phrase, as he was able to conjugate the verb in the third person present. In year 3, he also used a larger variety of determiners, such as ‘a’, ‘the’, ‘my’ and ‘some’. In the guessing game in year 4 he demonstrated that he was able to construct questions such as ‘Does he wear glasses?’ It was also evident that he was developing knowledge about tense, for example, when he mixed the present simple with the present continuous ‘Does he sitting around the table?’ Knowledge of the verb phrase, however, did not surface in the writing task in year 5, where he consistently left out the third person ‘s’. However, he communicated with the reader by a closing phrase ‘sincerely’ and by asking questions designed to prompt the reader to write back ‘So what is your school called?

210

Par t 3: Longitudinal Perspec t ives Using Mixed Methods Tomas Listening

Reading Production, Girauds Production (Det/N)

Vocabulary retrieval

Yea 1

Yea 2

Yea 3

Yea 4

Yea 5

Yea 6

Figure 12.1 Illustration of Tomas’ achievements over six years

and how is it to go in school in England? I would be really happy for answer.’ In the year 6 writing task, he demonstrated how over the six years he developed knowledge about the noun phrase to include two determiners ‘the whole planet’ as well as no determiner ‘People never’, the verb phrase ‘and it floats’, sub-clauses ‘which allows us to see’, and idiomatic expressions, such as ‘I actually hate’ and ‘do the work for me’. His receptive skills were high throughout primary school with scores from 89 to 100% on the listening and reading tasks. At the end of primary school there seemed to be a ceiling effect for him in the national tests both for listening and reading (with scores of 100 and 94% respectively). Figure 12.1 provides a visual representation of how the skills of listening, vocabulary development, production and reading developed dynamically over time. The underlying data can be found in Appendix 12.1. In summary, Tomas is a confident learner whose exposure to English outside school is high, probably long before he starts primary education, given his high achievements in listening and vocabulary. He feels he gets help learning English from his parents, though he rarely needs it. He uses English to a considerable degree outside school both for play and communication, in particular from year 4 and onwards. His attitudes towards English are neutral and he finds English easy. Tomas came to school with a high level of both receptive and productive knowledge of English. He maintained this high level throughout primary school and gradually developed his production from single word utterances into an idiomatic and complex use of English.

David The home David’s family often help him with homework by asking him to repeat the words he has been learning. He believes they are happy that he is learning English and that they think it is a language that everyone in the world should know. He also mentions that his mother asks him about English

Thick Descr ipt ions of Early L anguage Lear ning

211

classes on a regular basis. He does not like English homework but always does it. For example, in year 5 he spends about 30 minutes per day on his homework. This mainly consists of learning vocabulary and writing short texts. From year 3 onwards, David reports using English on a daily basis, including: watching TV, listening to music, using the internet (e.g. YouTube) and playing computer games. There are no English books in his home. In year 5 he also reports that he talks to his friends in English about ‘things we would have talked about in Swedish’. By this he meant that they are not particularly restricted by the FL, but can chat about the same things as they would have in their L1 Swedish.

Attitudes towards learning English In the earliest years of primary education, David’s attitudes towards English are neutral. He claims that he neither likes nor dislikes English. Later, in years 5 and 6, his views have changed and he is more positive. Throughout primary education, he is more positive towards learning new words and listening than towards speaking. English is not his favourite subject in school and his preferences vary across the years. He mentioned music and making newspapers as preferred ‘subjects’. He likes most activities during English classes; however, in year 5 he reflects back on his previous experiences of learning English and says that ‘In year 3 we did not learn so much, we only watched the smart board, it was fun, we could watch TV but they did not pick up on what we did, we only watched the film and then nothing more, I think you should watch the film then rewind and pause and say ‘“what does that mean?” “I learnt most English in year 4 and 5.”’ In year 5, his replies to the smiley questionnaire were also more positive towards learning English than those he had previously given.

Attitudes towards being a learner of English During year 2, David believes he learns English slower than his classmates. This opinion changes during the later years of primary education, in years 4 and 5, when he thinks that he learns at about the same speed as others in the class. His replies are somewhat contradictory as he also states, throughout primary education, that he finds English easy, with the exception of year 3 when he finds it difficult. In the interview that year he says that he always found English difficult and that ‘I don’t understand anything’. At the same time, he likes to play in English and explains that he believes that ‘you learn a lot of English when you have fun’. In year 4, he finds it more difficult than the previous year and says that this is mainly because of the new vocabulary. In year 5, he finds it similar to the previous year and that the most difficult part is to write, both in English and in Swedish.

212

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Achievements David was selected as a learner who produced an average score on the vocabulary retrieval task in year 1. His listening skills in year 1 were average, with 12 out of 19 points on the listening task. David’s production of English seems to take off in year 3 where vocabulary size, diversity and noun phrase complexity increased substantially. In year 2, he could not respond to prompts to ask for the toilet, to order food for him and his family or to ask for bread and water, while in year 3 he produced ‘Can you … can my mum’, ‘Chicken nuggets’ and ‘Wait, can you give me and mama and dad water and bread?’ He did not include any determiners except numbers, but demonstrated knowledge of question formation and auxiliaries. In the guessing game in year 4, he produced the articles ‘a’, ‘the’ and the pronoun determiner ‘his’, and question formations including ‘what’ as in, ‘What colour is his hair?’ He struggled with some utterances ‘en boy, eller är det … boy eller girl? [a boy or is it … boy or girl?], but used prosodic strategies to indicate that the utterance was a question. In the written task in year 5, he used some additional pronoun determiners ‘my’ and ‘her’. He also shows interaction with the reader by including opening and closing phrases ‘Hey’ and ‘it was all about me bye’ and an intermediate discourse marker signalling change of topic ‘now I gonna tell more about me’. In the year 6 written task, mostly he used main clauses, but showed development of syntactic complexity with the sub-clause ‘im very hungry so i go and buy a pizza’. He had also not yet mastered the subject/verb agreement as reflected in the third person ‘he want’ and created the future tense consistently in the text by the use of a strategic device, stating that the story was set in the future and then continuing in the present tense, or by using the progressive ‘Im gonna meet him’. He does not use punctuation consistently as in ‘And now I go to him when i am there …’, where the first word phrase ‘And now I go to him’ should have been a main clause in the previous sentence followed by a full stop and a new sentence. This indicates uncertainty of writing per se, which could have affected the retrieval of English words as well as structures and contents. His listening skills improved from 63% in year one steadily during years three and four (93%) and dropped slightly in the national tests to 81% in year six. Reading skills also show a stable curve from year four through to year six. Figure 12.2 provides a visual representation of how the skills of listening, vocabulary development, production and reading developed dynamically over time. The underlying data can be found in Appendix 12.1. To sum up, David started primary education knowing some vocabulary and with relatively good receptive skills in English. This indicates that he had been exposed to English before he started school in year 1, and from year 3 he clearly stated that he used English at home. His attitudes towards himself as an FL learner fluctuates over the years, but his attitude towards learning English becomes more positive towards the end of primary education.

Thick Descr ipt ions of Early L anguage Lear ning

213

David Listening Reading

Production, Girauds Vocabulary retrieval Production (Det/N)

Yea 1

Yea 2

Yea 3

Yea 4

Yea 5

Yea 6

Figure 12.2 Illustration of David’s achievements over six years

Throughout, he gets support from his parents. He gradually developed his ability in production and demonstrated how he built vocabulary and the use of noun phrases over the six years, but it was also evident that his use of verb phrases and sentence level syntax did not develop to the same level.

Mikael The home Throughout primary education, Mikael reports that his parents are happy that he is learning English and that he receives a lot of support from them. They give him the names of things at home and help him with his homework by, for example, testing him on vocabulary. He spends 1–3 hours a week doing his English homework and he does it carefully. Together with his parents he also watches movies and listens to music in English. By the later years of primary education, he reports using the internet in English for about an hour a day, mainly to listen to music.

Attitudes towards learning English From classroom observations, Mikael showed a high level of interest and involvement during English classes. His favourite subjects in school, however, were PE and cross-curricular project work. Over the years he never reported English as a favourite subject, but nor did he identify anything that he particularly disliked about English or the classroom activities. In direct response to questions about learning English, by the end of his second year at school Mikael reported that he felt positive. Two years later, in year 4, his view had changed and he reported neutral feelings in response to the same question in the smiley questionnaire. In year 5, at the age of 11 years, Mikael had changed his perception of learning English yet again, now viewing lessons quite positively. His feelings towards different activities fluctuated across primary education. During year 1, he reported that he enjoyed

214

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

speaking and learning new words most. In year 3, he was neutral towards these activities and instead more positive towards listening. In year 4, his feelings towards listening had also become neutral. In year 5 he was, again, positive towards speaking and learning new words.

Attitudes towards being a learner of English Throughout primary education, Mikael consistently held the view that learning English was difficult. During year 2, he felt that he learnt English at about the same speed as others, a perception that changed in year 3 and onwards, where he thought he was slower than his classmates. He found listening to a recorded voice difficult as he had to listen and write at the same time and when translation was introduced in year 4, he perceived of that as particularly difficult. In year 4, he also reported that he found speaking and reading difficult and explained ‘if I don’t understand enough I can’t reply’. He felt lessons were different this year as compared with previous years. There were new words and sentences to learn. Even through Mikael often said that he was a slower learner of English than other children in the class, he always came through as a confident person with high self-esteem. When he talked about his achievements, he seemed realistic rather than negatively affected emotionally.

Achievements Mikael was the weakest of the three learners according to the results of the vocabulary retrieval task in year 1. Figure 12.3 provides a visual representation of how the skills of listening, vocabulary development, production and reading developed dynamically over time. The underlying data can be found in Appendix 12.1. The production lines showing vocabulary retrieval and vocabulary diversity indicate that Mikael’s vocabulary seemed to develop relatively consistently over the six years, even though there was a drop in year 6. For example, in the restaurant task in year 2 he did not respond in English to a prompt where he was asked to help his family order food. In year

Mikael Listening Production (Det/N)

Vocabulary retrieval Production, Girauds

Reading

Serie9 Yea 1

Yea 2

Yea 3

Yea 4

Yea 5

Figure 12.3 Illustration of Mikael’s achievements over six years

Yea 6

Thick Descr ipt ions of Early L anguage Lear ning

215

3, however, he said ‘Pappy pancakes, mummy chicken, jag tar en [I will have a] hamburger’. In response to a prompt to ask where the toilet was he did not respond in year 2, but showed use of both English and strategies in year 3, ‘Mummy it’s [makes a noise to suggest using the toilet]’. In the guessing game in year 4 it was difficult for him to form questions and again he used strategies in the form of prosody (rising end tone) in combination with an English pronunciation of the Swedish verb ‘är’ [is/are] to indicate question ‘är short?’ [is short?]. He struggled at times to find the right vocabulary, as in this sequence when he wanted to ask whether the person the interviewer was thinking about was a boy or a girl (English production in italics): ‘ehm … ska vi se … ska vi se … är det en kille eller tjej … eh, vad är är? Eller is … är … is … det? Is it a girl and, eller girld eller … eh … tro, nej … det är … det är två bokstäver … ja … ear boy?’ [ehm … let’s see … let’s see … is it a boy or a girl … eh, what is is? Or is … is … is … it? Is it a girl and, or girld or … eh … believe, no … it is … it is two letters … yes.. ear boy?]. The only determiners he used prior to year 4 were numerical ones. In year 4, however, he used the indefinite article ‘a’. In the written tasks in years 5 and 6 he also uses the definite article ‘the’, pronouns ‘my’ and ‘your’, and adjective ‘many’. It seems the writing task in year 5, which was to write a letter to a pen friend, supported the retrieval of a richer vocabulary and more complex noun phrases than the descriptive narrative of the national test in year 6. In the letter (year 5), Mikael shows that he can communicate about himself in the present tense. He interacts with the reader in opening and closing phrases ‘Hellow’ and ‘Bouy, bouy from Mikael from Sweden’, and by asking a question ‘How are your school and dear you live?’ In the year 6 task he does not master the future tense, but consistently uses a translation of the Swedish auxiliary ‘kommer’ (come) to create the future and uses ‘come’ instead of ‘will’ in several places, for example, in ‘I come like at’, [I will like it/Jag kommer gilla det] and ‘come it was’ [will it be/kommer det vara]. His text was shorter in year 6 than year 5 (74 as compared with 82 words), even though the time allowed for the task was longer in year 6. It seems that, since the task was set in the future with more general contents and vocabulary, these requirements proved to be far more challenging for Mikael than a letter about himself to a pen friend, which resulted in a shorter and less complex text. His listening scores fluctuated over the period and peaked in year 4. In the national tests, his listening scores dropped again, which indicate that he found those tests more difficult than the ELLiE instruments that were used in the first four years. Reading comprehension showed a different pattern with an increase from year 4 to year 5 and then a slight decline in year 6. The broad picture shows a learner who is hard-working and supported both in school and in the home, but whose opinion about himself as a learner is low and whose emotions about English fluctuate. His receptive skills in English were higher than his production skills from the onset of schooling. His production skills increased and seemed to be more in line with his

216

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

receptive skills by the end of primary education. His linguistic resources were limited throughout the six years of primary school, but he demonstrated a clear willingness to communicate and active use of strategies.

Discussion The case studies of Tomas, David and Mikael’s language learning provide thick descriptions reflecting the somewhat erratic nature of learning over time. From these it is evident that while their classroom contexts for learning are either identical (Tomas and David were in the same class) or quite similar (in the case of Mikael), their home environments varied considerably in terms of opportunities to engage with English. All three children report on a supportive home environment, with parents encouraging homework in English, viewing films and TV together in English and other similar activities. However, it is notable that Tomas experienced substantially more access to English, even before commencing his primary school education, as evidenced by the vocabulary retrieval task in year 1. Building on this positive start, he later sought out opportunities to engage in online gaming activities, conducted internationally, in English. By year 5 he claimed to be ‘using English almost all the time’ – a quite astonishing claim for a Swedish child aged just 11 years old. From the reports of David and Mikael’s home environments we understand that they also had contact with English at home, but to a substantially lesser extent. For example, Mikael mentioned listening to music in English for about one hour per day by year 5. From the data, it appears that Tomas’ early interest in English lay a foundation which led to later engagement in English for social purposes as he matured. His use of FL as a mediating tool to participate in online international social networks at this young age suggests a keen desire to reach beyond the community of his immediate school and home environments. From a Vygotskian perspective, this finding traces how his desire to participate in particular social worlds (by becoming part of an online social network) has been mediated through the use of culturally and historically formed artefacts – in this case, language (Vygotsky, 1978). Understanding exactly how and why each of the children in this study have positioned the tool of an L2 as more or less significant to their social worlds is beyond the scope of this study and would require further research into their home environments. However, the data presented here give a clear indication of the emergence of distinctive linguistic identities already, revealing the social practices that Toohey (2003: 71) suggests, ‘structure their appropriation of voice’. Turning to the application of multiple methods in this research study, the various instruments used offer a range of insights related to pathways in language learning. The quantitative data perspective indicates that Swedish children have generally high achievement levels in English (Szpotowicz &

Thick Descr ipt ions of Early L anguage Lear ning

217

Lindgren, 2011: 133), while qualitatively the more individualised nature of evidence is revealing. The downside of such qualitative evidence is the difficulty in drawing any kind of general conclusions. For example, in this study the detailed records of attitudes and achievements could be related for some learners at some points, but did not appear to be so strongly related to every stage of the learning process. The findings on attitudes illustrate the dynamic nature of individualised pathways. In the case of Mikael, he seems to think that he is poor at English and finds it difficult, although he does not regard it as boring. Overall, though, we can conclude that he seems to have a positive self-concept and is able to make an objective statement about his rather low ability level in English. Observations and interviews over time have confirmed the nature of his attitudes towards learning English showing that in year 2 he felt more positive; in year 3 his production had increased a little, but his listening declined. He was less positive in year 4 (this may be related to a change of teaching approach). In year 5, again he became more positive, yet his achievement level dropped. Overall, we found that, for Mikael, achievement seemed not to be very connected with his attitude, suggesting that possibly the declining achievement may be more connected with increased levels of challenge he faced in successive years. For David (assessed by his teacher in year 1 as an average learner), findings related to attitudinal development have been quite positive throughout, although his replies were somewhat contradictory on occasion. He mentions that in year 4 he found lessons more difficult than the previous year. In year 5 he explains this in more detail by reflecting on the change in the style of teaching (as a result of a change of teachers when moving to the middle school). He also explained that in year 4 he found it more difficult than in year 3, yet his achievements in year 4 indicated that he had progressed. His attitudes towards English were also more positive in year 5 than previous years. With the shift in approach, moving from a focus on meaning making towards a stronger focus on form in language learning, David recognised that he found this new challenge difficult – yet he was ultimately successful at it. From the findings for Mikael, David and Tomas it is difficult to fully interpret the evidence. A weaker learner has remained positive (Mikael), yet a stronger learner has become bored (Tomas). Explanations for each of these outcomes can be imagined, but mainly these findings emphasise the extremely complex nature of progress over time and the risks involved in categorising learners too soon.

Conclusion The combination of a mixed-method approach with a longitudinal perspective has shed light on the nature of variation within a group of young language learners. The six years of data have allowed us to explore these learners’

218

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

development as they progress from the very initial stages of language learning to the pre-adolescent phase where their linguistic identities have become more fully formed. In making sense of these data we have illustrated how both progress in language learning and attitudes to language learning vary over time, yet seem to come together by the time these learners reach year 6. The illustrations of learner achievement show how receptive skills steadily develop in the early phases of learning, with production generally following later. In the case of Tomas, however, production skills were strong even in his first year of primary schooling, suggesting that he had previously experienced significant exposure to English either in a pre-school context or in his home environment. Interview and questionnaire data on attitudinal development has also proved particularly revealing on questions related to perceptions of language learning. It seems that finding learning difficult does not necessarily mean that learning does not occur. Similarly, finding learning boring may not mean that the learner will be unsuccessful at it. Such evidence serves to emphasise the demanding nature of the teacher’s task in understanding and supporting each learner’s pathway towards becoming a successful language learner. What has become clear in this study is that extensive exposure to English in the home environment can have an extreme effect on the degree of variation in language competency between learners of the same age. The variation may be more extreme in countries such as Sweden, where exposure to English can be very high. In this study we have reviewed the progress of three language learners, one proceeding at a moderate pace, with limited exposure to English in the home environment (Mikael); the second progressing quite well, with regular engagement in digital technologies in English at home (David) and the third (Tomas) progressing substantially in advance of his peer group, reaching ceiling effects in both listening and reading. However, the national tests and current curriculum are outcomes-based, relying on the standards-based framework of the Common European Framework of Reference (CEFR) (Council of Europe, 2001). Such systems offer little to the learner who is able to go beyond the required target outcomes and thus pay a disservice to learners such as Tomas. Here, we problematise the indication that the target may be set too low. Evidence from the outcome of Surveylang (European Commission, 2012), a language test administered to 15 year old learners of foreign languages across 16 European countries, indicated that by the age of 15 many Swedish students may perform at a B2 level rather than the required B1 level. Such evidence raises questions about the need for further reform of the national curriculum to fully reflect the potential for these achievements. For teachers, the findings of this study suggest that it might be informative to regularly survey their class to learn more about the learners, in addition to the ongoing feedback from more commonly used testing procedures. The findings from this study strongly indicate the need for differentiated learning tasks to provide appropriate levels of challenge for each learner and

Thick Descr ipt ions of Early L anguage Lear ning

219

suggest that in some cases individual learning plans might even be beneficial. With increased home access to online networking this study also confirms the need for more research into young children’s developing social identities through the medium of global languages such as English.

Appendix 12.1 Data underlying Figures 12.1–12.3. Year 1

Year 2

Year 3

Year 4

Year 5

Year 6

Vocabulary retrieval, number of words Tomas

32

76

78

David

17

18

65

Mikael

7

28

28

Production, Giraud’s index, Y 2–4 oral, Y 5–6 written Tomas

4.8

5.1

4.9

6.0

9.2

David

1.0

4.1

4.4

5.2

6.9

Mikael

1.8

2.8

3.8

5.2

4.2

Production complexity, Det/N, Y 2–4 oral, Y 5–6 written Tomas

7.5

7.5

8.8

6.1

6.3

David

0.0

2.5

5.8

6.0

6.4

Mikael

1.0

3.3

6.4

6.2

3.2

Listening, Max 19/18/23/32/19/37 Tomas

17

17

22

31

16

37

David

12

12

20

26

16

30

Mikael

13

14

13

28

10

14

Tomas

7

28

46

David

5

23

49

Mikael

1

13

16

Reading, Max 7/31/49

References Bialystok, E., Majumder, S. and Martin, M. (2003) Developing phonological awareness: Is there a bilingual advantage? Applied Psycholinguistics 24, 27–44. Butler, Y.G. (2017) The dynamics of motivation development among young learners of English in China. In J. Enever and E. Lindgren (eds) Early Language Learning: Complexity and Mixed Methods (pp. 167–185). Bristol: Multilingual Matters.

220

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Cenoz, J. (2003) The influence of age on the acquisition of English: General proficiency, attitudes and code mixing. In M.P. Garcia Mayo and M.L.G. Lecumberri (eds) Age and the Acquisition of English as a Foreign Language (pp. 77–93). Clevedon: Multilingual Matters. Council of Europe (2001) Common European Framework of Reference (CEFR) Documentation. Cambridge: Cambridge University Press. Enever, J. (ed.) (2011) ELLiE. Early Language Learning in Europe. London: British Council. Enever, J. (2013) The ELLiE study: Capturing evidence transnationally and longitudinally. ELT Research 28, 22–26. European Commission (2012) SurveyLang. European Survey on Language Competencies. Brussels: European Commission. Goswami, U. and Bryant, P. (2004) Phonological Skills and Learning to Read. Hove & New York: Psychology Press. He, T. and Wang, W. (2009) Invented spelling of EFL young beginning writers and its relation with phonological awareness and grapheme-phoneme principles. Journal of Second Language Writing 18, 44–56. Heining-Boynton, A. and Haitema, T. (2007) A ten-year chronicle of student attitudes toward foreign language in the elementary school. The Modern Language Journal 91, 149–168. Inhelder, B. and Piaget, J. (1958) The Growth of Logical Thinking from Childhood to Adolescence. London: Routledge & Kegan Paul. Lindgren, E. and Stevenson, M. (2013) Interactional resources in the letters of young writers in Swedish and English. Journal of Second Language Writing 22 (4), 390–405. Lopriore, L. (2017) Young Italian learners’ foreign language development: A longitudinal perspective. In J. Enever and E. Lindgren (eds) Early Language Learning: Complexity and Mixed Methods (pp. 186–200). Bristol: Multilingual Matters. Lopriore, L. and Mihaljević Djigunović, J. (2011) Aural comprehension and oral production of young EFL learners. In J. Horváth (ed.) UPRT 2011, Empirical Studies in English Applied Linguistics (pp. 83–103). Pécs: Lingua Franca Csoport. Mihaljević Djigunović, J. (2012) Early EFL Learning in Context – Evidence from a Country Case Study. London: The British Council. Mihaljević Djigunović, J. (2014) L2 learner age from a contextualised perspective. Studies in Second Language Learning and Teaching 4 (3), 419–441. Mihaljević Djigunović, J. (2017) Developmental aspects of early EFL learning. In J. Enever and E. Lindgren (eds) Early Language Learning: Complexity and Mixed Methods (pp. 222–246). Bristol: Multilingual Matters. Mihaljević Djigunović, J. and Letica Krevelj, S. (2009) Instructed early SLA: Development of attitudes. Studia Romanica et Anglica Zagrabiensian (SRZA) 54, 127–146. Mihaljević Djigunović, J. and Lopriore, L. (2011) The learner: Do individual differences matter? In J. Enever (ed.) ELLiE: Early Language Learning in Europe (pp. 29–45). London: The British Council. Nationella prov i främmande språk (NAFS) (2015) The Future (Elevblad). Projektet Nationella Prov i Främmande Språk. Göteborgs Universitet. See http://nafs.gu.se/digitalAs sets/1539/1539189_the_future.pdf (accessed 8 February 2016). Nikolov, M. (1999) ‘Why do you learn English?’ ‘Because the teacher is short’. A study of Hungarian children’s foreign language learning motivation. Language Teaching Research 3 (1), 33–56. Nikolov, M. and Csapó, B. (2010) The relationship between reading skills in early English as a foreign language and Hungarian as a first language. International Journal of Bilingualism 14 (3), 315–329. Orosz, A. (2009) The growth of young learners’ English vocabulary size. In M. Nikolov (ed.) Early Learning of Modern Foreign Languages (pp. 181–194). Bristol: Multilingual Matters.

Thick Descr ipt ions of Early L anguage Lear ning

221

Piaget, J. (1936) Origins of Intelligence in the Child. London: Routledge & Kegan Paul. Piaget, J. (1959) The Language and Thought of the Child (Vol.5). New York: Psychology Press. Pinter, A. (2007) Some benefits of peer-peer interaction: 10-year-old children practising with a communication task. Language Teaching Research 11 (2), 189–207. Pollard, A. and Tann, S. (1987) Reflective Teaching in the Primary School. London: Cassell. Szpotowicz, M. and Lindgren, E. (2011) Language achievements: A longitudinal perspective. In J. Enever (ed.) ELLiE. Early Language Learning in Europe. London: British Council. Toohey, K. (2003) Learning English at School. Identity, Social Relations and Classroom Practice. Clevedon: Multilingual Matters. Vygotsky, L.S. (1978) Mind in Society: The Development of Higher Psychological Processes. Cambridge, MA: Harvard University Press. Vygotsky, L.S. (1986) Thought and Language (Revised edition). Cambridge, MA: MIT Press. Wells, G. (1986) The Meaning Makers: Children Learning Language and Using Language to Learn. Portsmouth, NH: Heinemann Educational Books. Wood, D., Bruner, J.S. and Ross, G. (1976) The role of tutoring in problem solving. Journal of Child Psychology & Psychiatry 17, 89–100.

13 Developmental Aspects of Early EFL Learning Jelena Mihaljević Djigunović

Introduction In many contexts early foreign language (FL) curricula stress that the key goals of early learning of FLs do not only include language development, but also developing positive attitudes to language learning and to oneself as a language learner, and raising and enhancing motivation for language learning. Based on these goals, in this chapter we look into developmental aspects that refer to both affective and language development. Acknowledging the potentially important role of the context of learning, we also take contextual factors into account.

Affective development in early FL learning Until recently, researchers did not pay much attention to young learner individual differences, because children were considered not to differ from one another to the extent that it would impact on their language learning development. It was thanks to longitudinal studies of early FL learning that young learners’ individual differences emerged as a potentially highly revealing factor which could increase our understanding of the language learning processes at a young age. The first individual learner characteristics that were looked into were attitudes and motivation. Most early studies on attitudes and motivation focused on their relationship with language achievement. Initially, findings pointed to significant positive relationships of attitudes and motivation and language learning outcomes (e.g. Bernaus et al., 1994; Harris & Conway, 2002). However, a more complex picture emerged when the types of measures of language achievement were taken into consideration. Thus, motivation turned out to be more significantly correlated with language learning outcomes measured with particular types of measures (e.g. Masgoret et al., 2001; Muñoz & Tragant, 2001). 222

Development al A spec t s of Early EFL Lear ning

223

Developmental aspects of attitudes and motivation became the focus of researchers’ interests quite recently. Investigations of such aspects are rather complex because they require a longitudinal research approach. The main early findings of attitudinal research suggested that young language learners’ attitudes initially reflect their parents’ attitudes or those of their significant others, but that they soon form their own views based on their experience of FL learning. Nikolov (1999) and Vilke (1993) found that the key factors influencing young learners’ attitudes in initial stages of learning are classroom processes and the teacher. Marschollek (2002) offered evidence of the importance of direct contact with native speakers for the formation of positive attitudes of young language learners. Kennedy et al. (2000) stressed that early FL learning has a positive impact on young learners’ attitudes to subsequent language learning. Findings of some studies (e.g. Kim, 2009; Nikolov, 1999) showed that attitudes and motivation change over time, and offer evidence of young learners becoming less motivated and developing less positive attitudes to language learning. There are, however, studies that offer evidence of the opposite; that high motivation of young learners can remain stable over extended periods of time if FL learning conditions are favourable (e.g. Cenoz, 2003; Mihaljević Djigunović, 1995). Findings of a Swiss study (Heinzmann, 2013), on the other hand, suggest that little variability occurs in motivation of young language learners and that they more or less establish their attitudes by age nine and then tend to stick to them. Few studies (e.g. Donato et al., 2000) document increasingly positive attitudes of young learners as they progress with their FL learning. Self-concept is quite a new topic in the early FL learning field. Arnold (2007) says that self-concept is an image that we form of ourselves as language learners. She also observes that ‘(l)earners must both be competent and feel competent’ (Arnold, 2007: 18). Mercer (2011: 14) defines self-concept as ‘an individual’s self-description of competence and evaluative feelings about themselves as a Foreign Language (FL) learner’. Recent developments in self-concept research (e.g. Mercer, 2011; Yeung & Wong, 2004) suggest that it is a complex and multifaceted phenomenon. Also, the initially broad conceptualisations of language learners’ self-concept have been somewhat refined. Lau et al. (1999) talk about an English as a foreign language (EFL) self-concept as a distinct type of selfperception and Mercer (2011) claims that it is possible to form separate selfconcepts for different aspects of learning English. Self-concept has been found to be a context-sensitive and unstable variable (Mercer, 2011; Mihaljević Djigunović, 2015; Mihaljević Djigunović & Lopriore, 2011).

Foreign language development Research-based insights indicate that early FL learning is a highly complex process. Its complexities become apparent when we consider the many

224

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

factors that impact on the language learning process, among which particularly salient ones are contextual and individual learner factors. With longitudinal approaches to researching early FL learning, as well as with the introduction of the qualitative research paradigm, its dynamic nature has surfaced too. In the present study we looked into early FL development by following oral production of our participants. Primary FL curricula insist on accuracy and fluency as overall aims in teaching oral skills. These two concepts are also considered as important characteristics of language learning and assessment tasks (Skehan, 2007).

Contextual factors The importance of context in which young learners are acquiring a FL is well acknowledged. Researchers often refer to contextual factors as belonging to the macro- or micro-level. The socioeducational aspects, FL policy and curricula, as well as out-of-school exposure to the FL, are usually considered to operate at the macro-level. The immediate learning environment (school, classroom processes, teacher and teaching materials), the learner’s socioeconomic status and home support belong to the micro-level. The contextual factors that operate at the macro-level are usually more stable than the micro-factors; they help contextualise early FL learning and are of special significance in comparative studies (Enever, 2009). The microfactors are important for understanding how early language learning proceeds and are vital for interpreting young learners’ development.

A Study of Affective and Linguistic Development of Young EFL Learners The study described in this chapter was carried out with Croatian young learners of EFL. Early FL learning in Croatia has been looked into in several projects in the past (see Vilke & Vrhovac, 1993, 1995), with the sevencountry study Early Language Learning in Europe (ELLiE) (Enever, 2011) being the most recent one. The present study applied a mixed methods approach and applied several instruments from the ELLiE study.

Croatian socioeducational context FL knowledge has always been highly valued in Croatia, as is the case with many tourism-oriented countries. FL has been a compulsory primary school subject for more than seven decades (Vilke, 2007). Primary school education in Croatia starts at age 6/7 and includes eight grades (grades 1–4: lower primary; grades 5–8: upper primary). Thus, students leave primary

Development al A spec t s of Early EFL Lear ning

225

education at age 13/14 years old. For a long time, FL was introduced in grade 4 (age 9/10 years old), but in 2003, FL became compulsory for all children from grade 1. As in many other contexts, English is the most popular FL, with over 85% of learners starting with English from grade 1 (Medved Krajnović & Letica Krevelj, 2009). German is the second most popular language, followed by Italian and French. The special status of English is secured by a regulation according to which learners who start with a language other than English have to take English from grade 4 as their second FL; thus no one can finish primary education without having learned English. Exposure to English is quite extensive and learners have a lot of access to English through TV (foreign programmes are subtitled), through listening to music, using the computer and so on. For many there are also abundant opportunities to use English with foreigners who visit Croatia, for example, as tourists or for business purposes. The Croatian National Educational Standards (2006) and the National Framework Curriculum (2011) are the key documents reflecting the national FL learning policy. In early FL learning the stress in these documents is put on oracy.

Aims of the study In this study, we wanted to look into development of young EFL learners’ attitudes to language learning and teaching, their motivation and selfconcept, as well as their EFL achievement. Our aim was to observe how these developed during the four upper primary years, while taking into account a number of contextual factors as well.

Sample A total of 81 Croatian learners of EFL participated in the study. They were 10/11 years old at the beginning of the study and were attending grade 5. English was their compulsory school subject from grade 4, thus it was their second year of learning English. We observed their development during grades 5–8. At the end of the study they were 13/14 years old. Table 13.1 shows the participants’ age at each of the four grades. Table 13.1 Participants’ ages in grades 5–8 Grade

Age

Grade 5 Grade 6 Grade 7 Grade 8

10/11 years 11/12 years 12/13 years 13/14 years

226

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

The sample was made up of three classes drawn from two town schools. In the school where two classes were included, each class was taught by a different teacher. A subsample of 18 pupils participated in the parts of the study requiring more in-depth investigations that could only be carried out on an individual basis.

Instruments Most of the instruments used in this study were taken over from the ELLiE study. Some were constructed by adapting the ELLiE instruments to suit the purpose of the present study. Still others were designed specifically for the purpose of the present study. In order to obtain relevant information about the school context in which the participants were learning English we interviewed the school principals and English teachers in the two schools. For this purpose we selected those items from the ELLiE interviews which elicited data on the status of English in the schools, attitudes of school staff to FL teaching, qualifications of the English teachers and their attitudes to teaching English to young learners. Information about the participants’ socioeconomic status, language learning behaviour outside school and out-of-school exposure was gathered using selected items from the ELLiE parents’ questionnaire. Socioeconomic status was determined on the basis of the parents’ level of education (primary, secondary, graduate level and post-graduate level). Insights into the participants’ out-of-school language learning behaviour were based on items that elicited information on whether they discussed English lessons at home, showed their classwork to members of the family, asked for help with homework, whether they were keen to use English, whether they felt insecure using it in real life situations, whether they liked speaking it, took pride in their knowledge of English and whether they complained that English was hard to learn. Data on exposure to English outside school were gathered by means of items referring to opportunities to use English during holidays, during visits by family friends and relatives or while watching TV, listening to music, reading magazines or books, and using the computer. Data concerning the participants’ attitudes to teaching and learning English were collected by means of an interview composed of three items. The participants were asked to name classroom activities they liked and disliked most and to explain why, and to select one out of three drawings depicting different classroom arrangements to indicate in which one they thought they would learn English best and to explain why. Motivation was measured by a smiley questionnaire, which was adapted from the ELLiE study. There were six items eliciting how much the participants enjoyed learning English in general, learning English words and using each of the four language skills. The participants assessed their

Development al A spec t s of Early EFL Lear ning

227

motivation by selecting one of the three smiley faces (unhappy, indifferent or happy). These were entered into the datasheet as 1, 2 or 3, respectively. The averages for each item and participant were then calculated and used in later quantitative analyses. The participants’ self-concept was measured in two ways. First, the participants were asked in the interview to compare themselves to their classmates in terms of whether they were better, the same or worse at learning English than their peers and to explain why. The three levels of comparisons were quantified on a 1–3 scale and analysed quantitatively, while the reported explanations were analysed qualitatively. Second, their parents were asked in the parents’ questionnaire how their child felt about learning and using English. EFL achievement was conceptualised in this study as the level of achieved accuracy and fluency in oral production. Annual personalised interviews were carried out with individual participants in which they could show how well they could use the vocabulary and structures they learned in their English classes and how fluently they could express themselves. The interviews focused on two topics: the home and the park the participants went to in their spare time. A picture of a house and a picture of a park were used as a starter; as the participants described each picture, they were lead to discuss their own home and the park they frequented. Thus, they were asked to describe the place they lived in, what their favourite room was and why, whether they could cook and what they liked to eat at home. Then they were asked to describe the park they went to and what they did there. These topics were selected because they were shown through a pilot study to be motivating enough for EFL learners in grades 5–8, and were also found to elicit enough oral production that could give insight into EFL development. Two independent raters assessed the participants’ oral accuracy and fluency on five-point scales. The accuracy scale contained descriptions of five levels of appropriate choice and use of grammatical structures, ranging from ‘speech is incomprehensible because of mistakes’ to ‘very good choice of grammatical structures used mostly correctly and appropriately’. The fluency scale reflected five levels of fluency/hesitancy, intonation and pronunciation and comprehensibility of oral production. The scale ranged from ‘speech is mostly incomprehensible’ to ‘fluent speech; good intonation and pronunciation; speech is easy to comprehend’. The inter-rater reliability for accuracy was 0.95 and for fluency, 0.91. Samples of learner language at different levels of the scales can be found in the Appendix.

Procedure School principal and English teacher interviews were carried out during the first and fourth year of the study. A parents’ questionnaire was administered at the beginning of the fourth year. Learner questionnaires were

228

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Table 13.2 Administration of instruments

Instruments administered

Year 1 Learner questionnaire (whole sample). Individual learner interview (subsample). Oral task. School principal interview. English teacher interview.

Year 2

Year 3

Year 4

Learner questionnaire (whole sample). Individual learner interview (subsample).

Learner questionnaire (whole sample). Individual learner interview (subsample).

Learner questionnaire (whole sample). Individual learner interview (subsample).

Oral task.

Oral task.

Oral task. School principal interview. English teacher interview. Parents’ questionnaire.

administered at the end of each year and the participants filled them out in class. Interviews with the selected participants were carried out at the end of each year, too. The oral production task was repeated with the subsample comprising 18 participants during the last week of each grade. Table 13.2 summarises the administration of instruments by year.

Results and Discussion Findings concerning contextual factors The two school contexts from which the sample was selected differed in two ways. One school was a typical Croatian primary school, which means that it followed the national curriculum for English and was averagely equipped (TV, CD-players, English books and dictionaries in the school library). The other school was considered an elite school because it offered international curricula for foreign learners as well, and was better equipped due to foreign investment (e.g. computer labs and interactive whiteboards in many classrooms). Also, learners in this school had opportunities to meet foreign learners during breaks and communicate with them in English. However, as far as classroom teaching of English is concerned, all the participants followed the same syllabus and teaching materials, and had three 45-minute lessons of English per week. In both school contexts, the principals welcomed early teaching of FLs, and English in particular, and the FL was considered a highly important school subject. As far as the English teachers’ qualifications were concerned,

Development al A spec t s of Early EFL Lear ning

229

all the three teachers were specialist English teachers having a university degree in English language and literature (teaching stream). All entertained positive attitudes to early FL learning and liked teaching English. Their teaching was, generally, communicatively oriented and only slight differences could be noticed in the amount of L1 they used during their classes; English was used between 60 and 70% of class time. It can be concluded that the participants learned English in, more or less, typical school and classroom conditions. The education levels of parents is considered to be a good indicator of one’s socioeconomic status. The results show that the majority of the participants’ parents had secondary education (60.7% and 65.6% of mothers and fathers, respectively). Interestingly, more mothers (27.9%) than fathers (19.7%) completed higher education, and the same was true for the postgraduate levels (4.9% and 1.6% for mothers and fathers, respectively). Our findings suggest that, overall, the participants’ socioeconomic status was average. Out-of-school contact with FL can be an important factor in early language learning. Thus, we collected data on: (1) how much time the participants were typically exposed to English through watching programmes in English, playing computer games, listening to music, reading and speaking; (2) whether they had access to the internet and what activities they performed on the computer; and (3) what type of contact, if any, they had with Englishspeaking people. From the obtained data we could see that the participants got most of their contact with English through watching TV (films, cartoons, soap operas and the like) and listening to music. This happened, typically, between two to four hours per week. The least amount of exposure was materialised through reading. About 90% of the participants had access to the internet at home. Less than half watched something in English on the internet or wrote things in English; the majority used it for listening or playing purposes, or to read about something. Half of the participants had had contact with English-speaking people. Most of these contacts were with foreigners in Croatia, usually tourists. Some participants used English on visits to foreign countries, either for holidays or to visit relatives or friends, and few used English in the home. The results indicate that English was present in the participants’ outside school lives to a significant extent. Experiencing English out of class can make its learning more meaningful and useful in terms of proficiency development. As evidence from other studies (e.g. Lindgren & Muñoz, 2013; Mihaljević Djigunović et al., 2008; Szpotowicz & Lindgren, 2011) suggests, exposure to the FL one is learning correlates with language achievement. Home support was measured by looking into the ways the parents were involved in the participants’ EFL learning. The parents’ involvement in the

230

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

participants’ EFL learning was reflected most in discussing what happened in their EFL classes, which was the case with over 60% of the parents. Only about a third of the participants reported showing their notebooks or workbooks to their parents, or asking them for help with English. The importance of the parents’ role in early FL learning has often been stressed (e.g. Enever, 2007) and research points to its interactions with language achievement (Lindgren & Muñoz, 2013; Lopriore & Mihaljević Djigunović, 2011).

Development of individual learner characteristics As can be seen from Figure 13.1, motivation proved not to be a stable variable. It was highest in grade 5, then dropped in grade 6 and kept increasing over the last two grades. Significant differences were found between grades 5 and 6 ( p < 0.001) as well as between grades 6 and 7 ( p < 0.001), while motivation in grade 7 did not differ significantly from motivation in grade 8 ( p < 0.092). As mentioned in the first part of this chapter, an increasing number of studies into early FL learning report that young learners’ initially high motivation decreases from year to year. Some authors (e.g. Carreira, 2011) attribute this to the observed negative trends in motivation for studying in general. Based on the findings of the present study we would like to suggest that conclusions about decreasing motivation is an oversimplification. It may be more useful to talk about fluctuations of motivation. As Figure 13.1 below shows, motivation among this sample of young EFL learners did not change linearly, but the levels of motivation went both up and down during the four years. As Butler (2015) suggests, due to its complex nature, changes in primary school learners’ motivation can hardly be expected to be consistently linear. We also agree with the authors (e.g. Nikolov, 2002) who claim that

Figure 13.1 Development of the participants’ motivation (three-point scale)

Development al A spec t s of Early EFL Lear ning

231

better understanding of motivational dynamics in early FL learning may emerge if contextualised approaches are used. Having this in mind, we believe that our participants’ motivation dropped significantly in grade 6 due to the introduction of new subjects and, at the same time, to changes in English teaching methodology from grade 6, which relies less on fun activities and more on formal and cognitively-oriented processes. By the beginning of grade 7, most young learners get used to the new workload and language teaching approach. A look at the participants’ reports on what they particularly enjoyed doing in their English classes showed that they varied from year to year, too. In grade 5, their accounts included playing games in English (e.g. ‘it’s great when we play different games’, ‘I enjoy guessing games’), speaking activities (e.g. ‘I like it when we talk in English’, ‘answering questions is my favourite activity’), listening activities (e.g. ‘I love it when we listen to songs in class’, ‘listening to stories is my favourite thing’), as well as more explicit learning activities (e.g. ‘I like best learning new words’, ‘doing workbook activities is the most interesting for me’). Many of the same activities were reported as enjoyable in grade 6, too; however, some new ones were added to the list. These referred to reading (e.g. ‘it’s best when I get to read out loud what a character in the text is saying’, ‘I like it when we read stories in English’), writing (e.g. ‘my favourite activity is writing a composition about an interesting topic’, ‘I like doing dictations’) and grammar (e.g. ‘learning new grammar is the most interesting’). While in grade 5, no one could think of anything they disliked in their English classes, in grade 6 the first reports on dislikes emerged. These referred to either classroom management issues (e.g. ‘the boys often disrupt the class’), general things (e.g. ‘I don’t like anything we do in English classes this year’, ‘I don’t like when we have to study hard’) or to particular learning activities (e.g. ‘I dislike dictations and tests’, ‘I don’t like answering questions’). Some of the preferences in grade 7 were similar to those reported in the previous two years (e.g. ‘I like best listening to songs and stories’), but there were interesting new reports referring to how teaching activities were carried out (e.g. ‘we do a lot of group work now’, ‘I like it when we can express our own opinions in discussions’), to teaching content (e.g. ‘we learn about such interesting topics this year’), as well as to the feeling of success in learning English (e.g. ‘I like everything we do because I’m good at English’). What the participants disliked in grade 7 referred to grammar and writing activities, which some found too difficult (e.g. ‘tenses are very difficult’, ‘we write too much and I find it physically tiring’) or to uninteresting teaching materials (e.g. ‘the texts are very boring’, ‘grammar is boring’). An analysis of grade 8 reports on likes suggests that many of the participants felt they had reached a stage at which they could use English to express themselves quite well. Thus, they reported liking texts in their textbook (e.g. ‘I know more this year and can use the words I learned outside school to understand the texts’, ‘now we can express our own opinion about the topic because we know a lot of English’) and the video programmes that focused on historical or

232

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

cultural topics (e.g. ‘I love the programmes we sometimes watch during English classes, we learn a lot from them’). What the participants reported disliking were grammar activities (e.g. ‘grammar is difficult’), tests (e.g. ‘I hate tests’) and the fact that more L1 used by the teacher would be useful (e.g. ‘if the teacher used more Croatian when explaining difficult things, it would be easier for me to understand English’). Compared to reports of the Croatian sample of EFL learners observed from grade 1 to grade 4 (Mihaljević Djigunović, 2012), the choices of preferred classroom activities in the present study indicate that such preferences are age-dependent. As young learners mature cognitively, they also develop new preferences in line with their understanding of the FL learning process. In the present study, a possible age effect can be detected in an earlier emergence of disliking many teaching activities, or all of them, because learners found them boring. The participants’ attitudes to learning and teaching English, as reflected through their choices of the classroom arrangement in which they believed they could learn English best, seemed a bit more stable than their preferences for particular classroom activities. In each of the four grades this measure was administered, the majority opted for the group work arrangement (see Figure 13.2). As fifth graders, they liked it as much for being able to sit next to their friends as for the ‘better learning’ reasons. In grade 6, they reported liking this arrangement because everyone had a chance to speak more, they could help each other and they could work more independently, too. In grade 7, they claimed that the classroom atmosphere was better when working in groups and it was less boring than working in the traditional way. In grade 8, the stress was on being able to negotiate what to do and how to go about the tasks, and on the fact that the teacher was around to help if necessary.

Figure 13.2 The participants’ preferences for classroom arrangements (%)

Development al A spec t s of Early EFL Lear ning

233

The second most frequent choice was the traditional classroom arrangement. Fifth graders liked it because it looked similar to learning in other subjects and because they were used to it. As sixth graders, they said there was peace and quiet in the classroom and everyone was paying attention to the teacher, looking at the blackboard and could follow what was going on. In grade 7, the same things were mentioned and some participants also added that it was the most normal way to learn. In grade 8, the participants liked that everything was clearly shown on the blackboard, which they thought made learning easier. Consistent preferences for the group work arrangement and the reasons offered for this choice show that the majority of the participants were aware of what can contribute to more effective learning of EFL. This corroborates the findings which indicate that young learners can reflect on their language learning and can clearly express their reflections. The participants’ self-concept (Figure 13.3) kept changing during the four years. Although the between-grade differences were not statistically significant, it is still interesting to note the observed trends and consider the participants’ explanations. The participants’ self-concept was the most positive in grades 5 and 7, less positive in grade 6 and was least positive in grade 8. In grade 5, the participants explained that they knew they were better at English than their classmates by referring to the higher grades they were assigned by their English teacher. Some also claimed they could do the tasks faster than some of their peers. Those with a less positive self-perception reported that it took them longer to figure out things in English or that they could not concentrate for too long. In grade 6, most explanations as to why they were doing better ranged from their learning English outside school too, mostly in private language

Figure 13.3 Development of the participants’ self-concept (three-point scale)

234

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

courses, to being in touch with English through the media, to talking in English to people from different parts of the world. The explanations for being less good than others centred round the fact that they did not find English as easy as others did. In their comparisons with classmates the seventh graders mostly relied on the English grades and did not offer elaborate reports. In grade 8, the participants also referred to their English grades, but offered additional explanations as well. Those with a highly positive selfconcept claimed that their pronunciation was better, or that they could understand English better and faster than their classmates. Some stressed that they volunteered answers more frequently than others and that their classmates often asked them for help. The participants with a less positive self-perception said that they could not read as well as others and could not pronounce English as well as some of their peers. Data collected by the parents’ questionnaire offered valuable information on the participants’ out-of-school language behaviour, which can be considered as a reflection of their motivation and self-perception. As Figure 13.4 shows, according to the parents’ reports, the majority of the participants were keen to learn English, took pride in learning this language and liked to speak English outside school. Less than 20% of the participants found learning English hard and only about 10% felt anxious or insecure when using English. It is interesting to observe that the parents’ reports on their children’s language behaviour in the home and other non-school contexts paint a much more positive picture of the participants’ motivation and self-concept than the one we would get directly from the data offered by the participants themselves. One possible explanation is that the parents were too partial and

Figure 13.4 The participants’ outside school EFL behaviour as reported by their parents (%)

Development al A spec t s of Early EFL Lear ning

235

wished to present their children in a positive way. Another possibility is that young learners really do behave more confidently out of class, that is, in situations where chances of competition, assessment, error correction or language anxiety are lower. This result suggests a need to find out how revealing it may be to distinguish young learners’ motivation and self-concept in class from those emerging out of class.

EFL achievement Figure 13.5 presents the findings related to the development of accuracy and fluency in oral production of our participants during the four years of the study. A paired-samples t-test was conducted to compare differences in accuracy and fluency between the year groups. Although accuracy displayed an increase over the years, only the increase between grades 5 and 6 was found to be statistically significant (t(16) = –2.340, p = 0.033). Fluency, on the other hand, showed some fluctuation. After increasing in grade 6, when it was significantly higher than in grade 5 (t(16) = –2.562, p = 0.021), fluency decreased in the last two grades, especially in grade 7; however, these drops were not statistically significant. The notable increase in accuracy in grade 6 may be attributed to the fact that in grade 6, more attention is paid to formal aspects of English than in previous grades. At that time, many young learners are already entering the formal operations stage of cognitive development (Piaget, 1923), which may enhance this aspect of their language learning process. Also, with increasing vocabulary size, sixth graders can engage in a more varied and interesting practice of language structures. Such developments in accuracy do not

Figure 13.5 Development of the participants’ accuracy and fluency in oral production (five-point scale)

236

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

continue at the same rate. This is not surprising; second language acquisition (SLA) research (e.g. de Bot & Larsen-Freeman, 2011; Herdina & Jessner, 2002; Larsen-Freeman & Cameron, 2008) tells us that language learning does not proceed in a linear fashion. Besides knowledge of vocabulary and structures, fluency may depend on extralinguistic factors as well. One of these is language anxiety, which may emerge at different points during language learning. Grade 7, when fluency decreased in the present study, is a particularly sensitive time for primary school learners. Studies on language anxiety of Croatian language learners (Mihaljević Djigunović, 2002) have shown that seventh graders can feel very anxious when performing orally in the formal classroom context. Communication strategies, as a factor contributing to fluency, probably played a role, too; it was observed each year that they were more frequently used by more fluent participants.

Interactions of EFL achievement with individual and contextual factors We looked for interactions by means of t-test and correlation analyses. Significance of differences in accuracy and fluency with regard to language learning behaviour was tested by means of t-test, while correlational analyses were carried out to look into interactions of accuracy and fluency with motivation and self-concept. A summary of analysed interactions is presented in Tables 13.3 and 13.4. It was interesting to note that no significant relationships of accuracy and fluency were found with motivation or self-concept as reported by the participants themselves. However, the particpants who were perceived by their parents as keen to learn English, in comparison with those who were not, achieved higher accuracy (t(11) = 3.033, p = 0.015) and fluency (t(11) = 2.582, p = 0.032) in grade 5, but not in the other grades. Showing pride in knowing English made no difference in any of the grades. No significant interaction was found at any grade level between accuracy or fluency and the parents’ perception of whether the participants found English hard to learn. Liking to speak English was connected with significant differences in accuracy at all grade levels (grade 5: t(11) = 3.033, p = 0.015; grade 6: t(11) = 2.870, p = 0.015; grade 7: t(11) = 3.563, p = 0.004; grade 8: t(11) = 2.852, p = 0.016). Differences in fluency were significant only in grade 5: t(11) = 3.666, p = 0.008) and grade 7: t(11) = 3.738, p = 0.003). Feeling insecure proved to interact with accuracy in all grades except grade 5 (grade 6: t(11) = –2.839, p = 0.016; grade 7: t(11) = –3.819, p = 0.003; grade 8: t(11) = –2.760, p = 0.019). The differences were significant in case of fluency in all grades (grade 5: t(11) = –6.055, p < 0.001; grade 6: t(11) = –4.025,

Development al A spec t s of Early EFL Lear ning

237

Table 13.3 Summary of significant (+) and non-significant (−) interactions of motivation, self-concept (Pearson correlation, n = 18) and language learning behaviour (t-test, n = 81) Grade 5

Grade 6

Grade 7

Grade 8

Accuracy Fluency Accuracy Fluency Accuracy Fluency Accuracy Fluency

−

−

−

−

−

−

−

−

−

−

−

−

−

−

−

−

+ − + − − −

+ − + + − −

− − + + − −

− − − + − −

− − + + − −

− − + + − −

− − + + − −

− − − + − −

− −

− −

− −

− −

− −

− −

− −

− −

Language learning behaviour

Motivation Self-concept Keen to learn Show pride Like to speak Feel insecure English is hard Discuss English lessons Show classwork Ask for help

p = 0.003; grade 7: t(11) = –4.733, p = 0.001; grade 8: t(11) = –3.911, p = 0.008). In all cases the participants who did not feel insecure when using English performed better. No differences in accuracy or fluency were found at any grade level between those participants who discussed their English lessons at home, showed EFL classwork to parents, or asked them for help with English and those who did not. Table 13.4 Summary of significant (+) and non-significant (–) interactions of accuracy and fluency scores with out-of-school exposure to English in grade 8

Out-of-school exposure to English

Activity Listen Listen on computer Speak Speak on computer Read Read on computer Write Write on computer Watch programmes Watch programmes on computer Met English-speaking people

Accuracy

Fluency

+ + − − − − − + − − −

− + − − − − − − − − −

238

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

The amount of exposure to English through listening positively correlated with accuracy in all the grades (grade 5: r = 0.60, p = 0.024; grade 6: r = 0.61, p = 0.020; grade 7: r = 0.58, p = 0.031; grade 8: r = 0.66, p = 0.011), but the correlation with fluency was significant only in grade 5 (r = 0.60, p = 0.024). The participants who used the computer to listen to English achieved significantly higher fluency results in all grades (grade 5: t(12) = 6.633, p < 0.001; grade 6: t(12) = 3.952,p = 0.003; grade 7: t(12) = 4.684, p = 0.001; grade 8: t(12) = 3.612, p = 0.016). Accuracy was, however, higher only in grade 7 (t(12) = 4.395, p = 0.002) and grade 8 (t(12) = 2.911, p = 0.013). Using the computer for writing purposes contributed to higher accuracy in grade 5 (t(12) = 2.908, p = 0.019), grade 7 (t(12) = 2.375, p = 0.048) and grade 8 (t(12) = 2.324, p = 0.042) but not grade 6. In the case of fluency it proved a contributing factor only in grade 5 (t(12) = 2.731, p = 0.028). Using the computer for reading, watching programmes or for playing did not interact with either accuracy or fluency at any grade level. No differences were found in either accuracy or fluency between the participants who had met English-speaking people and those who had not. No significant differences were found among the two schools or the three groups the participants belonged to. What emerges from the interactions found are differences in the impact which contextual and individual learner factors had on the development of accuracy and fluency in different grades. The interactions of affective factors underscore our observation made earlier, concerning what emerged about motivation and EFL self-concept development from the participants’ reports, as opposed to their parents’ reports. It seems to us that it would be useful to consider the role of these affective characteristics separately in language learning contexts (in class) and from language use contexts (out of class). The lack of significant interactions with parental support suggests a decreasing impact of parents on their children’s schooling at this age. It seems that extensive out-of-school exposure to English took a key role as a contextual factor, which had an impact on both affective and language development of the young learners in this study. The role of listening emerged as particularly salient and can probably be attributed to the increasing use of new technology gadgets that young learners find appealing. Many of the established significant interactions in particular grades, or their lack, are difficult to interpret. We believe that this is so because, in contrast to popular belief, young learners differ from one another perhaps more than the more mature learners. It is likely that the differences may be caused by their cognitive, emotional and language development happening in parallel. On the other hand, a comparison of the results of this study with those that observed first graders’ development (e.g. Mihaljević Djigunović &

Development al A spec t s of Early EFL Lear ning

239

Lopriore, 2011) suggests that perhaps we should not consider young EFL learners a monolithic group. We would like to suggest that a useful distinction could be made in research between young learners who start EFL learning in grade 1, at the very beginning of their formal education, when they have not yet learned the ‘rules of the game’ and therefore approach EFL learning differently from young learners who start with EFL with a ‘baggage’ of general school experience.

Conclusions and Implications The study described above offers insights into how Croatian 10/11-yearold EFL learners’ affective and language development proceeded during four years (from grade 5 to grade 8). The obtained results show that their attitudes, motivation and self-concept varied from year to year. In contrast to many previous studies, our findings suggest that young EFL learners’ motivation and self-concept can be better described as fluctuating than decreasing over the years. Another interesting finding is the significant differences observed between the participants’ affective development when related to EFL classroom learning, as opposed to using English out of class. This suggests that it might be useful to distinguish, and investigate, these as two different phenomena. The trajectories that the affective development of the fifth graders in this study followed over four years differed from those found in studies with first graders (e.g. Mihaljević Djigunović, 2012; Mihaljević Djigunović & Lopriore, 2011). We, therefore, believe that it might be a good idea not to consider young learners as a compact or uniform group of 6 to 14 year olds, but break them into meaningful subgroups based on their general learning experience and the affective and cognitive stages they are at. Language development, in terms of accuracy and fluency of oral production, was also found to fluctuate from year to year in different ways. The interactions of accuracy and fluency with contextual and individual learner factors over the years proved to be too complex to lead to clear interpretations. What we see as the potentially best way of reaching an informed understanding of key developments in early EFL learning is to resort to case studies of young learners.

Appendix Examples of participants’ oral production (Grade 7, fourth year of learning English) Note: I = investigator; L = learner.

240

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Accuracy and fluency were assessed on a five-point scale (1 – lowest level; 5 – highest level)

Excerpt 1 (accuracy: 5; fluency: 5) I: L: I: L: I: L:

I: L: I: L:

I: L:

I: L:

I:

I: L:

Can you compare each room with your rooms in your house? Well…no, my mom and dad are obsessed with our house to be tidy so…no. I’m going to tell this to your mom No please. OK, I won’t. OK, generally speaking but what about each room? Can you compare, for example, the bathroom with your bathroom? Err…no, I don’t think sso…my bathroom is smaller, I don’t have a bath because we use a shower, we always take a shower. Upstairs there should be another bathroom but we haven’t….we built another but we don’t have stuff in it…err…my living room is much bigger, it doesn’t have bookshelves, I have a study for that…err…my dad usually reads his paper at work or I don’t know while eating lunch or something…and…I don’t know…the only thing I can compare here is my mom doing chores. Ok, would you like to live in this house? No. Why not? Well, I like to play PC a lot and I don’t see it here. I would have to sha…If I would live here I would have to share my room with my brother and in this case a sister I don’t have…well…I don’t know, it’s small… Is there anything that you like in this house? Err…well…yes…I like that it’s untidy. I don’t care if my room is tidy or so…but…my mom makes me clean it…so…I kind of even like it untidy…everything is out so I can find something easily…I don’t have to look in the drawers, closets… What is your favourite room in your house? My study…well there it’s my PC, I spend a lot of time playing it… and…there are two very comfortable chairs in it…so I even study there, not in my room but there. Well my mom and dad are usually in the living in the living room watching TV so don’t bother me. There is no kitchen in the picture, there is no dining room, so do you have a dining room in your house? Well, my kitchen is very large so it’s a kitchen and a dining room… together… And where do you eat? In the kitchen, you have…our kitchen is even larger than our living room and so…half of it is for cooking and the other half is for… eating…

Development al A spec t s of Early EFL Lear ning

I: L: I: L: I: L: I: L:

I:

241

Who cooks in your family? Well, usually my mom but sometimes my dad cooks Can you cook? No, I don’t think so, maybe tea or something… Just tea? Well, I can make….err…I can make pasta or something like that… but… Now the last question, is there anything I haven’t asked and you would like to tell me about you, your house, pictures, English lessons, English as a language? Well, I think English is very amusing and fun to learn language… err…it’s very useful if you know a lot of it…you can talk with people all over the world, you can get…you can be more social if you move outside of the country and so… Perfect, thank you very much.

Excerpt 2 (accuracy: 4; fluency: 3) I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L:

Can you compare your room with this one? Yes, my it’s messy like this… Uhm, really? Yes. Is your room that big? Yes. Can you tell me everything you see in the hall? In the hall it’s a door, mail on the floor, a jacket…some umbrella… abau…some books, telephones, pictures… And? Compare your hall with this one. I don’t have…err… What don’t you have? What, a staircase? I don’t have this for letters…a hole… Mailbox. Yes…I don’t have this clothes under the stairs and stairs I don’t have… Would you like to live in this house? Yes, it’s…big and… I like it. Why do you like it? Because this room is big and there some closet…many books, television, DVD…i tako [and so]. Is there anything you don’t like? Err…no…and I like this for pikado [darts]. What is your favourite room in your house or flat, where do you live? In your house? Yes, it’s my room. Why is it your favourite? Err, because I got a computer…TV…I got balls and I play, I like it.

242

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

I:

There is no kitchen in this picture and there is no dining room, do you have a dining room? Yes…and the kitchen… And a kitchen? But, but it’s together… So where do you eat actually? Da… [Yes] Wait, where do you eat? In the…dining room. So, who cooks in your family? My mom cooks. Can you cook? Yes. Wow, what can you make? Err…pancakes…egg..err…some meat taco. Ahm, great…Is there anything else you would like to tell me? …No… Good…

L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I:

Excerpt 3 (accuracy: 3; fluency: 3) I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L:

Can you compare your bathroom with this one? Err…yes…my have a shower…window in the bathroom…sink…and err…bathtub ne znam kako se to kaže [don’t know how to say this]… (unintelligible) …and door… Can you tell me everything you see in the living room? There is…one man is sitting and reading a newspaper… I see…one table, TV, window…one painting… Can you compare your living room with this one? Yes. How is it different…or similar? Err…my is living room is have a TV like this…painting, window… table… Is your living room bigger than this one? Yes… Much bigger or not so much? Yes…much bigger… Can you tell me everything you see in the room? Err…I see one girl is sleeping…err…I see a …door…a mirror… Is there any bed? Yes…a bed…posters…kak se to kaže [what’s the word for that]… Oh yeah posters, you’re right. Can you compare your room with this one? Yes…my rooms is bigger…my rooms is…don’t have a mirror…I have a computer in my room…I have bed…smaller bed…

Development al A spec t s of Early EFL Lear ning

I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I:

A smaller bed… Yes…and that’s all… Can you tell me everything you see in the hall? Err…one women is…walking down the stairs…I see the…door, the window…paintings… that womens…err…carrying a jacket? Yes… And… Err…I don’t have a stairs… in this house where I’m living. I don’t have a….I don’t…and. I don’t have a painting-paintings. Would you like to live in this house? …Err…no! Why not? Because…I…because I …house is bigger…and……..I ….want to live in my house because ….it’s bigger….and it’s….. It’s yours. Yeah, it’s mine. Is there anything that you like or don’t like in this house? I don’t like because it’s …err…messy in the room….and in the living room… What is your favorite room in your house? My room. Your?… Why? Because there is a computer and I always play in the my room. There is no kitchen in the picture and there is no dining room. Do you have a dining room in your house? No, we eating in living room. Who cooks in your family? My mom. Can you cook? Yes. Oh really, what can you make? Err…toast…eggs… Ahm, and pasta? Yes…err… What about tea or coffee? Yes I can make a tea, coffee a… Ok, is there anything you would like to tell me about your house or your room, or this house? No. Okay, thank you.

Excerpt 4 (accuracy: 1; fluency: 1) I: L: I:

243

Would you like to live in this house? Err… Would you like to live in this house?

244

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

L: I: L: I: L: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L: I: L:

Err…(unintelligible). What is your favourite room in your house? Err…my room Your room…why? Err… My room…because…err…(unintelligible) … Do you have a dining room in your house? …Yes Where do you eat lunch? …err… Who cooks in your family? …my mom. Can you cook? …nekad [sometimes] Sometimes. Sometimes. What can you make? … What can you make? Eggs, pancakes, what? Tea, coffee? …coffee…eggs Ide li to?[How are we doing?] Ne baš…[Not really good].

References Arnold, J. (2007) Self-concept as part of the affective domain in language learning. In F. Rubio (ed.) Self-esteem in Foreign Language Learning (pp. 13–29). Newcastle: Cambridge Scholars Publishing. Bernaus, M., Cenoz, J, Espí, M.J. and Lindsay, D. (1994) Evaluación del aprendizaje del inglés en ninños de cuatro años: Influencias de las actitudes de los padres, profesores y tutores. [Assessment of EFL learning in four-year-old children. Impact of teacher, parent and guardian attitudes]. APAC of News 20, 6–9. Butler, Y.G. (2015) Parental factors in children’s motivation for learning English. A case in China. Research Papers in Education 30 (2), 164–191. Carreira, J.M. (2011) Relationship between motivation for learning EFL and intrinsic motivation for learning in general among Japanese elementary school students. System 39, 90–102. Cenoz, J. (2003) The influence of age on the acquisition of English: General proficiency, attitudes and code mixing. In M. García Mayo and M. García Lecumberri (eds) Age and the Acquisition of English as a FL (pp. 77–93). Clevedon: Multilingual Matters. Croatian National Educational Standards (2006) Zagreb: Croatian Ministry of Science, Education and Sports. Croatian National Framework Curriculum (2011) Zagreb: Croatian Ministry of Science, Education and Sports. Donato,R., Tucker, G.R., Wudthayagorn, J. and Igaraschi, K. (2000) Converging evidence: Attitudes, achievements, and instruction in the later years of FLES. Foreign Language Annals 33 (4), 377–399. de Bot, K. and Larsen-Freeman, D. (2011) Researching second language development from a dynamic systems theory perspective. In M.H. Verspoor, K. de Bot and W. Lowie

Development al A spec t s of Early EFL Lear ning

245

(eds) A Dynamic Approach to Second Language Development (pp. 5–24). Amsterdam: John Benjamins. Enever, J. (2007) Yet another early start language policy in Europe: Poland this time! Current Issues in Language Planning 8 (2), 208–221. Enever, J. (2009) Can today’s early language learners in England become tomorrow’s plurilingual European citizens? In M. Nikolov (ed.) Early Learning of Modern Foreign Languages (pp. 15-29). Bristol: Multilingual Matters. Enever, J. (ed.) (2011) ELLiE: Early Language Learning in Europe. London: The British Council. Harris, J. and Conway, M. (2002) Modern Languages in Irish Primary Schools. An Evaluation of the National Pilot Projects. Dublin: Institiúid Teangeolaíochta Éireann. Heinzmann, S. (2013) Young Language Learners’ Motivation and Attitudes. London: Bloomsbury Academic. Herdina, P. and Jessner, U. (2002) A Dynamic Model of Multilingualism: Changing the Psycholinguistic Perspective. Clevedon: Multilingual Matters. Kennedy, T.J., Nelson, J.K., Odell, M.R.L. and Austin, L.K. (2000) The FLES attitudinal inventory. Foreign Language Annals 33 (3), 278–289. Kim, I.-O. (2009) An analysis of primary school students’ English learning motivation based on their grade levels and regions. English Language Teaching 21 (2), 259–282. Larsen-Freeman, D. and Cameron, L. (2008) Complex Systems and Applied Linguistics. Oxford: Oxford University Press. Lau, I.C., Yeung, A.S., Ying, P. and Low, R. (1999) Toward a hierarchical, multidimentional English self-concept. Journal of Educational Psychology 91 (4), 747–755. Lindgren, E. and Muñoz, C. (2013) The influence of exposure, parents, and linguistic distance on young European learners’ foreign language comprehension. International Journal of Multilingualism 10 (1), 105–129. Lopriore, L. and Mihaljević Djigunović, J. (2011) Aural comprehension and oral production of young EFL learners. In J. Horváth (ed.) UPRT 2011. Empirical Studies in English Applied Linguistics (pp. 83–103). Pécs: Lingua Franca Csoport. Marschollek, A. (2002) Kognitive und affektive Flexibilität durch fremde Sprachen: Eine empirische Untersuchung in der Primarstufe. [Cognitive and Affective Flexibility in Foreign Languages. Empirical Study of First Grade]. Münster: Lit. Masgoret, A., Bernaus, M. and Gardner, R.C. (2001) Examining the role of attitudes and motivation outside of the formal classroom: A test of the mini-AMTB for children. In Z. Dornyei and Schmidt (eds) Motivation and Second Language Acquisition (pp. 281– 295). Honolulu: University of Hawai’i Press. Medved Krajnović, M. and Letica Krevelj, S. (2009) Učenje stranih jezika u Hrvatskoj: Politika, znanost i javnost [Foreign language learning in Croatia: Policy, research and the public]. In J. Granić (ed.) Jezična politika i jezična stvarnost [Language Policy and Language Reality] (pp. 598–607). Zagreb: Croatian Association of Applied Linguistics. Mercer, S. (2011) Towards an Understanding of Language Learner Self-concept. New York: Springer. Mihaljević Djigunović, J. (1995) Attitudes of young foreign language learners: A followup study. In M. Vilke and Y. Vrhovac (eds) Children and Foreign Languages II (pp. 16–33). Zagreb: Faculty of Philosophy. Mihaljević Djigunović, J. (2002) Strah od stranoga jezika [Foreign Language Anxiety]. Zagreb: Naklada Ljevak. Mihaljević Djigunović, J. (2012) Early EFL Learning in Context – Evidence from a Country Case Study. London: The British Council. Mihaljević Djigunović, J. (2015) Individual differences among young EFL learners: Age- or proficiency-related? A look from the affective learner factors perspective. In J. Mihaljević Djigunović and M. Medved Krajnović (eds) Early Learning and Teaching of English: New Dynamics of Primary English (pp. 10–36). Bristol: Multilingual Matters.

246

Par t 3: Longitudinal Perspec t ives Using Mixed Methods

Mihaljević Djigunović, J. and Lopriore, L. (2011) The learner: Do individual differences matter? In J. Enever (ed.) ELLiE: Early Language Learning in Europe (pp. 29–45). London: The British Council. Mihaljević Djigunović, J., Nikolov, M. and Ottó, I. (2008) A comparative study of Croatian and Hungarian EFL students. Language Teaching Research 12 (3), 433–452. Muñoz, C. and Tragant, E. (2001) Motivation and attitudes towards L2: Some effects of age and instruction. In S. Foster-Cohen and A. Nizegorodcew (eds) Eurosla Yearbook (Vol. 1), (pp. 211–224). Amsterdam, The Netherlands: John Benjamins Publishing Company. Nikolov, M. (1999) ‘Why do you learn English?’ ‘Because the teacher is short.’ A study of Hungarian children’s foreign language learning motivation. Language Teaching Research 3 (1), 33–56. Nikolov, M. (2002) Issues in English Language Education. Bern: Peter Lang. Piaget, J. (1923) The Language and the Thought of the Child. New York: Harcourt Brace and World. Skehan, P. (2007) Language instruction through tasks. In J. Cummins and C. Davison (eds) International Handbook of English Language Teaching (pp. 288–300). New York: Springer. Szpotowicz, M. and Lindgren, E. (2011) Language achievements: A longitudinal perspective. In J. Enever (ed.) ELLiE: Early Language Learning in Europe (pp. 125–151). London: The British Council. Vilke, M. (1993) Early foreign language teaching in Croatian primary schools. In M. Vilke and Y. Vrhovac (eds) Children and Foreign Languages I (pp. 10–27). Zagreb: Faculty of Philosophy. Vilke, M. (2007) English in Croatia – A glimpse into past, present and future. Metodika 8, 17–24. Vilke, M. and Vrhovac, Y. (eds) (1993) Children and Foreign Languages I. Zagreb: Faculty of Philosophy. Vilke, M. and Vrhovac, Y. (eds) (1995) Children and Foreign Languages II. Zagreb: Faculty of Philosophy. Yeung, A.S. and Wong, E.K.P. (2004) Domain specificity of trilingual teachers’ verbal self-concepts. Journal of Educational Psychology 96 (2), 360–368.

Part 4 Evaluating Early Language Learning Programmes

14 Child EFL Interaction: Age, Instructional Setting and Development María del Pilar García Mayo and Ainara Imaz Agirre

Introduction Research on second language (L2) acquisition within the interactionist framework (Gass & Mackey, 2007; Long, 1996) has shown that oral interaction among learners facilitates the L2 acquisition process (see meta-analyses by Keck et al., 2006; Mackey & Goo, 2007). Negotiated interaction provides learners with comprehensible input, feedback and opportunities to produce modified output (García Mayo & Alcón Soler, 2013; Mackey, 2007). Most research to date within this framework has focused on adult English as a second language (ESL) or English as foreign language (EFL) populations, whereas interaction among child learners has been under-researched. Except for the pioneering work by Oliver (2009) with child ESL learners, little is known about whether interaction actually occurs in EFL settings, where exposure to the language is much less frequent than in ESL contexts, and even less about a new methodological approach that is becoming prevalent in Europe: content and language integrated learning (CLIL), where exposure to the target language is increased. It is well known that the early introduction of foreign languages in school contexts is on the increase worldwide (Enever, 2011; García Mayo & García Lecumberri, 2003; Muñoz, 2006, 2014; Nikolov & Mihaljević Djigunović, 2011). Gathering data about what actually occurs in EFL settings regarding the existence of child interaction in communicative contexts is of great importance in order to maximise the children’s language learning opportunities (Pinter, 2011). In order to lessen this research gap, the aim of this study is to assess whether or not EFL children use conversational interaction strategies and, if so, whether these vary depending on the age of the children, the context (mainstream EFL or CLIL) in which they are exposed 249

250

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

to the language, and whether they develop over time. A total of 27 dyads of children (age range 8–12 years old) were video-recorded while completing a communicative task and their use of conversational strategies was codified and analysed. For the group of children studied, the findings point to a clear impact of age (with more conversational adjustments and repetitions among the younger learners and more L1 use among the older) and of context (with the mainstream EFL children using more conversational adjustments and repetitions). However, changes over time differ depending on instructional context.

Review of Recent Research Negotiation of meaning among young learners Pica (1994: 494) stated that when learners experience difficulties in comprehending a message, they negotiate for meaning with their partners, native speakers (NS) or non-native speakers (NNS), modifying words and sentence structure until they achieve mutual understanding. Consider example (1): (1) NS: Do you like California? NNS: Huh? NS: Do you like Los Angeles? NNS: Uhm … NS: Do you like California? NNS: Yeah, I like it (Long, 1983: 180) As Gass et al. (2013: 351) explain, this example illustrates how, as a response to the NNS’s indication of misunderstanding (Huh?), the NS narrows down the topic (‘California’ to ‘Los Angeles’) and finally repeats the original question. The NNS understands the question because, thanks to his initial indication, the NS has used conversational strategies that facilitate the learner’s task. Long (1983) operationalised the conversational adjustments (CA) that interlocutors use during negotiation of meaning as confirmation checks, clarification requests and comprehension checks (see section on ‘Procedure for data collection and codification’ below for brief definitions and examples). As mentioned above, negotiated interaction has been claimed to have a facilitative role in language learning since it provides opportunities for comprehensible input, comprehensible output and opportunities for learners to receive feedback (Oliver & Grote, 2010). Oliver’s (1998, 2000, 2002, 2009) pioneering work on child interaction in ESL showed that children do negotiate for meaning with age-matched peers while working on communicative

Child EFL Interac t ion

251

tasks, and that they use conversational strategies similar to the ones adults use, but in a smaller proportion. However, probably due to their egocentric nature, the children very rarely used comprehension checks. Regarding possible age effects, Oliver (1998) did not find significant differences when analysing the interaction between 8–10- and 11–13-year-old learners in the use of negotiation of meaning strategies, but she identified a general trend indicating that negotiation moves drop when there is a better command of the language (Oliver, 2000). Little research has been carried out regarding negotiation of meaning strategies in EFL settings, in low-input level contexts where learners are exposed to the foreign language for about 3–4 hours per week. Mackey and Silver (2005) studied the oral interaction of 26 L1 Chinese EFL learners, aged 6–9 years old, in Singapore while they carried out different communicative tasks with native English speakers. The findings showed that their experimental group (n = 14), who had received interactional feedback in response to problems with question formation in English, showed more improvement than the control group in the target structure. Pinter (2007) analysed the benefits of learner-learner interaction in one pair of 10-year-old Hungarian EFL learners while completing and repeating a spot-the-difference task. She reported instances of peer assistance and of learners’ attention to each other’s utterances and suggested that task repetition can work effectively with children of this age and low proficiency levels. More recently, Butler and Zeng (2013) assessed developmental differences in interaction among 24 fourth grade students (aged 9–10 years old) and 24 sixth grade students (aged 11–12 years old) and their teachers in a task-based programme in public elementary school in China. One of the issues they considered was the general interaction patterns based on Storch’s model (2002) of dyadic interaction. Their findings showed that the most popular interactional pattern among the sixth grade learners was the collaborative one, whereas the fourth grade learners had a parallel-passive type of interaction, with little mutual engagement. García Mayo and Imaz Agirre (2016) reported similar findings regarding fourth grade learners in the Spanish EFL context. These researchers compared the oral production of 57 third grade students (aged 8–9 years old) with that of 66 fourth grade students (aged 9–10 years old) while completing a picture-differences task. As in Butler and Zeng (2013), the group of fourth grade learners featured mainly a parallelpassive type of interaction, whereas the younger learners were collaborative. Lázaro and Azpilicueta-Martínez (2015) identified the interactional strategies used by a very young group of 16 Spanish EFL learners (aged 7–8 years old) while completing a guessing game task. Couched with an interactionist framework, the study concludes that these very young children do negotiate for meaning, but much less than ESL adult and older child learners. An interactional strategy that was included in their inventory was the use of the children’s shared first language (L1), Spanish. The researchers concluded that

252

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

their study provided evidence for the use of communicative tasks to promote interaction in the EFL classroom even at low proficiency levels. If research on interaction within EFL contexts is scant, much less is known about interaction within a new methodological approach that is spreading throughout Europe, content and language integrated learning (CLIL) – but see García Mayo and Basterrechea (2017) – which we briefly present in the next section.

Content and language integrated learning (CLIL) Content and language integrated learning (CLIL) was adopted in the European Union as a methodology in which ‘[…] the foreign language is used as a tool in the learning of a non-language subject’ (Coyle, 2007: 545). In contrast to a mainstream foreign language setting (henceforth, mainstream setting (MS)), in a CLIL context both language and content should be equally focused. Thus, learners enrolled in these programmes are taught content subjects through the medium of the foreign language, which is English in most cases. CLIL methodologies have been claimed to develop risk-taking and problem solving skills in the language learner, to raise the learner’s linguistic competence and confidence, and to motivate and encourage student independence (see Dalton-Puffer, 2011 for a recent overview). CLIL learners are not only exposed to more hours of English, but also to a type of input that is qualitatively different from the input in regular EFL lessons (Coyle, 2007; Muñoz, 2007). Research to date has mainly reported overall benefits for CLIL learners in fluency and speaking confidence (Dalton-Puffer et al., 2009), a greater vocabulary than their MS partners and more morphosyntactic resources (Jexenflicker & Dalton-Puffer, 2010), although the findings related to specific morphosyntactic features once fine-grained analyses are carried out are hard to maintain (García Mayo & Villarreal Olaizola, 2011). To the best of our knowledge, so far there is only one study that has compared conversational strategies between CLIL and MS learners. García Mayo and Lázaro Ibarrola (2015) analysed the conversational interaction of 80 children, paired to form 40 age- and proficiency-matched dyads (20 CLIL dyads, 20 MS dyads), as they completed a picture-placement task. The study focused on CA, repetition (self and other-repetition) and L1 use. Their findings revealed that the CLIL groups doubled the number of CA compared to the MS groups in both age groups and they also resorted to their L1 to a lesser extent. Regarding the impact of age on CA when comparing learners within the same instructional approach, the researchers reported that the older group (fifth year, aged 10–11 years old) used fewer CA and repetitions than the younger one (third year, aged 8–9 years old), but fell back on their L1 more often, probably because the task was not challenging enough and, therefore, their level of motivation decreased. Azkarai and Imaz Agirre (2016) analysed the

Child EFL Interac t ion

253

negotiation of meaning strategies (CA, repetitions and L1 use) of 43 pairs of CLIL and MS EFL learners in fourth (9–10-year-old) and sixth (11–12-yearold) grade children. The main aim of their study was to assess whether three variables, namely, age, instructional setting and task type (picture placement, guessing game) had an impact on negotiation of meaning strategies. Regarding age, their conclusion was also that the younger children in both instructional settings used some of the negotiation of meaning strategies more frequently than their sixth grade peers. Regarding context, the researchers reported that, contrary to García Mayo and Lázaro Ibarrola (2015), children in MS programmes employed some negotiation of meaning strategies significantly more than their CLIL peers. As for task-related differences, their findings showed that the guessing game led to the use of some of the strategies more frequently than the picture placement.

The Present Study This study aims to contribute to research trying to identify the impact of different variables on child EFL oral interaction. Thus, its main goal is to assess whether the conversational strategies young EFL learners in MS and CLIL contexts use (operationalised as CA, repetitions and L1 use) would vary depending on the age of the learners, the context where they are learning the language, and most importantly, whether those conversational strategies would change over time. Following Oliver (1998), conversational strategies were operationalised as CA and repetitions. As in García Mayo and Lázaro Ibarrola (2015), we also consider the use of the learners’ shared L1. The following research questions were addressed and their corresponding hypotheses entertained on the basis of the few studies carried out so far: (1) Does age influence the use of conversational strategies? Oliver (1998) did not find significant age differences regarding the use of negotiation of meaning strategies in the age groups she studied (aged 8–10 and 11–13 years old), but identified a general trend indicating that there are fewer negotiation moves when there is a better command of the language (Oliver, 2000). García Mayo and Lázaro Ibarrola (2015) and Azkarai and Imaz Agirre (2016) also reported a more frequent use of negotiation of meaning strategies among the younger groups. Therefore, we would expect a decrease in the use of CA and repetition with increasing age. On the basis of García Mayo and Lázaro Ibarrola’s findings on L1 use, we would also expect older learners to use their L1 more frequently. (2) Does instructional context (MS versus CLIL) influence the use of conversational strategies?

254

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

The findings regarding the impact of instructional context are mixed. On the basis of the findings reported on by García Mayo and Lázaro Ibarrola (2015), we expect that CLIL learners will use more CA and more repetitions than their MS counterparts, and will use their L1 to a lesser extent. However, on the basis of Azkarai and Imaz Agirre (2016) we would expect just the opposite: children in MS programmes would use more CA and repetitions. (3) Do learners’ conversational strategies change over time? Given the lack of longitudinal studies considering conversational strategies among children of this age range, we follow Oliver (2002) to hypothesise that the lower the learners’ proficiency level, the more they will use CA. Although our participants are all beginner learners, we hypothesise that, after one more year of classroom exposure, the participants’ use of CA, repetitions – normally a sign of a less-developed interlanguage- and their shared L1 will decrease – especially in the CLIL group.

Participants A total of 44 dyads (88 learners) participated in this longitudinal study. Due to some participants’ temporary absence and the fact that some dyads were outliers at one of the times of data collection, the production of 17 pairs was discarded from the analysis. Hence, the conversational strategies of 27 dyads (54 learners) were analysed on two occasions (henceforth, T1 and T2). Participants were divided into four groups on the basis of their age, third grade (8–9 years old) and fifth grade (10–11 years old) at T1 of data collection and of their instructional setting, CLIL and MS. All learners had started learning English at the age of 6 years old, that is, in first grade of primary education but, whereas the MS groups received five hours of exposure to English per week, the CLIL groups received 12 hours. The children had a beginner proficiency level, as assessed by school-internal tests. Table 14.1 features the details of the learners that participated in the present study.

Materials Learners completed two oral tasks, a guessing game and a picture placement but, due to space constraints, this study will only report data on the latter. In the picture placement (PP) task, one poster with the drawing of a classroom on the upper part and of a playground on the lower part was given to each participant. Both learners had similar posters with six pictures of children (pictures a–f ). Learner A would have the poster with pictures a and b on it and the rest of the pictures outside the poster, whereas learner B

Child EFL Interac t ion

255

Table 14.1 Distribution of the participants in the study Grade and setting

TIME 1

Third CLIL Third MS Fifth CLIL Fifth MS

6 5 8 8

Number of Age Hours of dyads (years) exposure 8–9 8–9 10–11 10–11

12 5 12 5

Grade and setting

TIME 2

Fourth CLIL Fourth MS Sixth CLIL Sixth MS

6 5 8 8

Number of Age Hours of dyads (years) exposure 9–10 9–10 11–12 11–12

12 5 12 5

MS, mainstream; CLIL, content and language integrated learning.

would have pictures c and d and the rest outside the poster. The goal of the game was that the two participants should complete their posters so that both had the same photos in the same positions.

Procedure for data collection and codification Data collection was carried out at two points in time over a year. Learners completed the same task in the third grade (T1) and fourth grade (T2), as well as in fifth grade (T1) and sixth grade (T2) at the same time of the school year. Participants worked in pairs and their production was video-recorded. Both participants sat next to each other with a folding screen between them so that they could not see each other and could not fall back on non-verbal communication for task completion. A total of 9.26 hours of learner interaction were transcribed verbatim. The whole data set was tallied and analysed, excluding the production of the outliers. Following Oliver (1998), all the CAs and repetitions (self- and otherrepetitions) were independently codified by two raters and compared. Instances of L1 use were also tallied and analysed. Inter-rater reliability was 96%. Each conversational interaction strategy was analysed as a proportion of the total utterances taken by each dyad to complete each task. Utterances in which the researcher was involved or where learners addressed the researcher were excluded from the analysis. The following examples from our database illustrate the different conversational strategies: (2) Clarification request: utterances made by the listener to clarify what the speaker had said: 1 Student B: how many girls are in the picture? 2 Student A: eight 3 Student B: what? → clarification request (Dyad 1_CLIL_3rd, Time 1_PP)

256

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Student B has not understood what his partner has said and therefore uses a clarification request to make sure of the piece of formation necessary to go on with the conversation and move the task along. (3) Confirmation check: utterances made by the listener to establish that the preceding utterance has been heard and understood correctly: 1 Student A: playing football in this black and white in the middle is 2 Student B: in the middle? → confirmation check 3 Student A: Yes, in this black and white. (Dyad 10_CLIL_4th_Time 2_PP) Student B asks his partner about the location of a character in order to confirm his location. The character is wearing a black and white shirt and Student A points to it (this) and refers to its colour. (4) Comprehension check: utterances made by the speaker to check whether a preceding utterance has been correctly understood by the listener: 1 Student B: […] Is in… In your right of the blackboard. In the classroom between the-the table and the blackboard. 2 Student A: Hmm… 3 Student B: Ok? → comprehension check 4 Student A: Ok. Next photo. (Dyad 2_CLIL_4th_Time 2_PP) Student B is giving instructions to his partner about where to put an item in his picture. Student A seems to be thinking about what his partner has told him and Student B wants to make sure his partner has understood and asks him ‘ok?’ Student A reassures him in turn 4 and goes on with the task. (5) Self-repetition 1 Student A: it is in the park? 2 Student B: eh? 3 Student A: is in the park? (Dyad 4_EFL_6th_Time 2_PP) Student A asks about the location of the character in the picture. Student B asks for clarification and Student A repeats the question. (6) Other repetition 1 Student A: Very funny ears? 2 Student B: Funny ears? 3 Student A: Yes. (Dyad 4_EFL_6th_Time 2_PP)

Child EFL Interac t ion

257

Student A asks about the ears of the character and Student B repeats what he has just listened to. (7) L1 use 1 Student A: ‘el banco’, cómo se decía (how do you say bench)? 2 Student B: eee banco (bench). 3 Student A: al lado del (next to the) tobogán (slide)? 4 Student B: no, there isn’t. (Dyad 10_EFL_6th_Time 2_PP) The following section will report the findings of the study.

Results This section addresses the three research questions posited above. In order to do so, the conversational strategies used by CLIL and MS learners were tallied at each data collection time. A two-sample binominal test was conducted fixing the significance level at (α = 0.05). The first research question focused on the impact of age on the use of conversational strategies. Figures 14.1 and 14.2 illustrate our findings at T1 and T2, respectively. Figure 14.1, in which statistical significance is marked by a star, shows that the third grade learners made more frequent use of repetitions than fifth grade learners in both settings and the differences are significant (CLIL (z = 3.4560; p < 0.0001) and the MS (z = 2.5360; p = 0.011). Third year children also used more CA than fifth year children in CLIL (z = 2.6182; p = 0.008) and MS settings (although the latter difference does not reach significance). Regarding L1 use, the older learners fell back on it more frequently in the MS group (z = 2.8161; p = 0.004). These findings mirror the ones reported in García Mayo and Lázaro Ibarrola (2015), who also showed that the younger learners used more CA and repetitions, and, therefore, support previous results, although a more frequent L1 use by the fifth grade children is only found in the MS setting. One year later, however, the predictions are borne out for L1 use as in both contexts the older children use it more frequently in a significant way (MS, z = 3.7578; p < 0.0001; CLIL, z = 3.0736; p = 0.002). They are only partially confirmed for repetitions, as the only significant difference is found in the MS group at T1 (z = 3.4062; p < 0.0001). Our second research question focused on the impact of instructional context (CLIL versus MS) on the use of conversational strategies. Figures 14.3 and 14.4 illustrate the differences found.

258

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

25 20 15

*

*

3rd grade

*

10

5th grade

*

5 0 CA

REP

L1

CA

CLIL

REP

L1

Mainstream

Figure 14.1 Age group differences in the percentages of use of conversational strategies at Time 1 25 20 15

*

10

*

4th grade 6th grade

*

5 0 CA

REP CLIL

L1

CA

REP

L1

Mainstream

Figure 14.2 Age group differences in the percentages of use of conversational strategies at Time 2

Figure 14.3 shows that third grade learners in the MS setting use CA, repetitions and their L1 more frequently than their CLIL counterparts (CA: z = 3.1900; p < 0.0001; REP: z = 2.18 = 0.029; L1: z = 4.4072; p < 0.0001). In the fifth year, differences were found in L1 use (z = 8.2443; p < 0.0001). So, contrary to the predictions in García Mayo and Lázaro Ibarrola (2015), third grade children in the MS group use more CA and repetitions than their CLIL counterparts, but as predicted by these researchers, they also fall back on their L1 more frequently. Fifth grade children behave as predicted by these

Child EFL Interac t ion

259

25 20

* 15 10

*

*

CLIL Mainstream

*

5 0 CA

REP

L1

CA

3rd grade

REP

L1

5th grade

Figure 14.3 Instructional setting differences in the percentages of use of conversational strategies at Time 1

25 20 15

*

CLIL

*

10

Mainstream

*

5 0 CA

REP 4th grade

L1

CA

REP

L1

6th grade

Figure 14.4 Instructional setting differences in the percentages of use of conversational strategies at T2

researchers as they use more CA and repetitions than their MS counterparts, although not in a statistically significant way, but once more, it is the MS learners that make a more frequent use of their L1. At T2 the same findings can be observed. Thus, the fourth grade children in the MS group use more CA, more repetitions and also fall back on their L1 more frequently, although only the use of the last two strategies displays statistically significant differences (repetitions: z = 2.3209; p = 0.020; L1: z = 2.1396; p = 0.032). Among sixth grade children, we find the same pattern observed with the group one

260

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

year earlier (fifth year) and expected on the basis of García Mayo and Lazaro Ibarrola’s (2015) findings: more use of CA and repetitions by CLIL children (although not in a significant way) and more use of the L1 by MS children (z = 3.1755; p = 0.001). Our third research question examined whether the use of conversational strategies changed over time. It was hypothesised that learners would use fewer CAs, repetitions and L1 at T2 in both age groups and both instructional settings. Figures 14.5 and 14.6 display our findings.

25 20 15

* *

10

Time 1

*

Time 2

*

5 0 CA

REP

L1

CA

3rd-4th grade

REP

L1

5th-6th grade

Figure 14.5 Percentages of conversational strategies at T1 and T2 in the CLIL group

25 20

* 15 Time 1

*

10

Time 2

5 0 CA

REP 3rd-4th grade

L1

CA

REP

L1

5th-6th grade

Figure 14.6 Percentages of conversational strategies at T1 and T2 in the mainstream group

Child EFL Interac t ion

261

The findings point to a significant decrease in the use of CAs (z = 3.4933; p < 0.0001) and repetitions (z = 4.3249; p < 0.0001) at T2 among the CLIL children in the younger group and of CA among those in the older group (z = 3.5474; p < 0.0001). L1 use significantly increased (z = 3.2835, p = 0.001) in the older group. Figure 14.6 shows that the only significant differences found at T2 in both groups is the decrease in L1 use (younger: z = 2.8395; p = 0.004; older: z = 3.3735; p < 0.0001).

Discussion In this section we will consider the answers to the three research questions posited above and discuss whether or not they support previous research on the topic. Our first research question asked whether the age of the learners could have an impact on their use of conversational strategies, namely CAs, repetition and L1 use. Our findings showed that age had an impact at T1 when the younger learners (third and fifth grade) in both CLIL and MS groups were compared, that is, our results partially mirror those reported by García Mayo and Lázaro Ibarrola (2015). The younger learners used more CA and repetitions, although only the use of the latter reached significant differences. Contrary to the findings of the previous study, where both third and fifth year learners used the L1 more frequently in both contexts, in this study only the older fifth year MS learners fell back on their L1 more frequently. One year later, however, the impact of age cannot be clearly seen. When data from the same children completing the picture placement task are analysed, the only clear significant differences can be found in the more frequent use of repetition among the younger (fourth year) learners in the MS setting and, interestingly once more, the more frequent use of the L1 by the older (sixth year) learners in both contexts. Although Azkarai and Imaz Agirre (2016) found that the younger group in their study (fourth year) used some negotiation of meaning strategies significantly more, we cannot forget that these researchers considered child output from two tasks and they reported more detailed differences between the so-called three Cs (clarification requests, comprehension and confirmations checks), which is beyond the scope of this chapter. Considering the cross-sectional findings from the same learners at T1 and T2, the generalisations that we can arrive at are the following: (1) CA are used more frequently by the younger learners at T1 and only significantly so by the CLIL group; (2) repetitions are used more frequently by the younger learners in a significant way both at T1 and T2, except for the younger CLIL learners at T2; and (3) there is a clear significant tendency by the older learners in both settings and at both T1 and T2 to make a more frequent use of the L1 (the only exception being the older CLIL learners at T1).

262

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Therefore, the overall tendency is for the younger learners to use more CA and repetitions, a result that would go along with the claim made by Oliver (2000) in an ESL setting, that is, even though they are all beginner learners, the older groups seem to fall back on CA and repetitions to a lesser extent and to use their shared L1 more. Of course, in this particular study, the influence of task repetition cannot be overlooked. Although one academic year had gone by from T1 to T2, one could speculate that, as the children could still remember what the task was about, it was easier for them to solve it and, therefore, they did not feel the need to use CA. The fact that L1 use was more frequent among the older learners would also need further research as, at this point, we could only speculate that the type of task used was not motivating enough for their age. Our second research question asked whether the instructional context, namely, CLIL versus MS, could have an impact on the children’s use of conversational strategies, namely CAs, repetition and L1 use. The hypothesis, based on the work by García Mayo and Lázaro Ibarrola (2015) was that the CLIL learners would use CA and repetitions more frequently than their MS counterparts and that they would use their L1 less; however, Azkarai and Imaz Agirre (2016) reported that the MS children in their study were the ones using more CA and repetitions. In the present study, the CLIL fifth grade learners at T1 and T2 (sixth grade) use more CA and repetitions than the MS group, but the differences were not significant. However, among the younger group at Time 1 (third year) and T2 (fourth year), it is the MS group that uses significantly more repetitions (at both recording times) and CA (at T1). These findings would then support those reported in Azkarai and Imaz Agirre (2016) and would lead us to claim that context does play a role in the sense that CLIL learners, more used to participating in classroom interaction, are probably less likely to display misunderstandings and, therefore, to fall back on CA and repetitions. A very consistent finding is that, as predicted, the MS group is the one which, in both grades and at both recording times, used the shared L1 more frequently and always in a statistically significant way. It is likely that, due to the methodology followed in CLIL classes, where a more active participation through the L2 is enhanced, children in those programmes make an effort to use the target language more frequently and not their L1. Our third research question dealt with whether conversational strategies would change over time. Our hypothesis was that, after one year, the participants’ use of CA, repetitions – normally a sign of a less-developed interlanguage – and their shared L1 would decrease, especially in the CLIL group. What the findings showed was that, in the CLIL programme, the younger children featured a significant decrease in their use of CA and repetitions at T2. The decrease occurred in the expected direction. There was a significant decrease in the use of CA among the older group, but their use of repetitions did not experience any change. In any case, the decrease in CA followed the

Child EFL Interac t ion

263

pattern expected, even though we were dealing with beginner learners. In the MS group, however, there was a decrease in the use of CA among both the younger and the older learners, but it was not significant, as is the case with repetitions. Regarding the expected decrease in L1 use in both contexts at T2, it is only found among the younger and older MS learners, but not among the CLIL learners. In fact, the older CLIL group increases L1 use at T2 in a significant way. At this point, we could only speculate that the older CLIL children could find the tasks less motivating at T2, but no motivation questionnaires were administered so it would be difficult to know if that was exactly the reason for the significant L1 increase. What we have presented so far is a quantitative approach to the data, which helps to identify trends in the effects of age, learning context and time on the use of conversational adjustments, repetitions and the children’s shared L1. However, the question of whether we can actually observe qualitative changes in the children’s oral production from T1 to T2 is also interesting. Although a detailed qualitative analysis is beyond the scope of this chapter, in what follows we will comment on some observations made on the basis of the study of the transcripts of the learners’ production at both recording times. The reader should bear in mind that it would be necessary to conduct a further empirical study to ascertain whether or not these observations could be found among children of this age range in other EFL contexts (MS or CLIL). One interesting observation regarding the MS group is that the children’s use of the shared L1 significantly decreases at T2, which is a welcome finding considering the very limited amount of input these children are exposed to. When it is used, we find it in lexical language-related episodes (LREs). An LRE is ‘[…] any part of a dialogue in which students talk about the language they are producing, question their language use, or other- or self-correct’ (Swain, 1998: 70). Consider example (8): (8) 1 Student A: […] cómo se dice banco (how do you say ‘bench’?) 2 Student B: bench 3 Student A: this girl is next to the bench, the tree and then more or less the bench (Dyad 8_EFL_6th_Time 2_PP) In example (8), Student A needs a particular lexical item, ‘bench’, to go on with the description of the part of the poster he is trying to complete. His partner provides the needed item and solves the problem so the child can go on with the task. In this group we also observe how the type of questions asked have evolved and are closer to the target language pattern. Example (9) below

264

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

features a question asked by the same dyad 1 at T1 (fifth year) and at T2 (sixth year). Whereas in T1 the strategy used for question formation is simply rising intonation added to the statement, at T2 the child is using the accepted question formation pattern: (9) Time 1: the girl have a ball? Time 2: have you got a girl eating a sandwich? However, it should also be mentioned that whereas at T1 some dyads used a question yes/no answer pattern, at T2 the same dyads move to what we could call a ‘descriptive strategy’, that is, respecting the conversational turns, each member of the dyad describes what they have in their poster and the other member simply confirms saying ‘yes’ or provides a simple ‘no’ or asks more questions. Consider examples (10) and (11): (10) 1 Student A: what color is jumper? 2 Student B: is the color red (Dyad 3_EFL_3rd_Time 1_PP) (11) 1 Student B: he is after to the bank and the girl has got one sandwich… 2 Student A: is a long hair and a fair hair? (Dyad 3_EFL_fourth_Time 2_PP) Student B, who at T1 has basically used ‘yes/no’ answers and the longest stretch of output was what we see in example (10), uses more elaborate sentences at T2. Regarding the CLIL group, two clear patterns can be observed at T2: (1) Those pairs that, even though one academic year has gone by, clearly remember what the task was about and, therefore, make ‘to the point’ questions and do not need as many conversational turns as they used at T1. Consider example (12), with pseudonyms to protect students’ anonymity: (12) 1 Student A: I think that Carlos have the… the boy of the jacket blue on the ..on the …on the playground. 2 Student B: eh .. yes .. I think Álvaro have the .. eh .. boy of the .. eh ..grey jacket on the classroom (Dyad 5_CLIL_fourth_Time 2_PP This pair displays this type of interaction in the low number of turns – just 14 turns – that they need to complete the task, mostly correctly, compared to the 42 turns they used at T1.

Child EFL Interac t ion

265

(2) Those pairs that modify their interactional routines by being more precise in the description of the elements they have to place on the game board. Consider (13): (13) 1 Student A: ok, and where is it? 2 Student B: is where the door a little bit more in the left 3 Student A: in the door? 4 Student B: in the left part of the door 5 Student A: left part of the door. Eh .. in the blue? In the .. 6 Student B: yes but not exactly on the door ..eh .. on the floor (Dyad 2_CLIL_6th_Time 2_PP) The members of this dyad try to provide as many details as possible so that they can both complete the task. This is reflected in the 71 turns they employ at T2 compared to the 42 they had employed at T1. It should be noted that, probably because of the way they are taught in their regular foreign language classes, children participating in this study have a tendency to ‘speak in written language’ and are not fully aware of how spoken language operates. As mentioned above, it is surprising to find that the older children in the CLIL group (sixth year primary) significantly increase the use of their L1 at T2. With the data at hand we can only speculate that the task was probably not engaging for them the second time and that is why they fell back on their L1 to just complete it. As is usually the case in these cohorts, there is acrossdyad variability and more controlled research should be done on this topic.

Conclusion The main goal of this chapter was to examine whether age, instructional context and time would have an impact on the conversational strategies used by young beginner learners in an EFL setting. Our findings have shown that the general trends regarding age point to a more frequent use of CA and repetitions by the younger groups and an increase in L1 use among the older groups. As for instructional context, in this study the younger MS learners used more CA and repetitions at T1 and more repetitions at T2, which could point to a moderate influence of context. The longitudinal analysis of the data shows that there is a general tendency for a decrease in the use of both CA and repetitions by both MS and CLIL learners. There is also a decrease in L1 use among the MS learners, but this tendency was not observed among the CLIL learners. However, this finding could be due to the nature of the task used, which may not have been sufficiently motivating for this group of

266

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

learners. This study has also provided a brief sketch of some qualitative differences observed in the interactive patterns of the groups examined. The study has considered oral interaction data from children completing a picture-placement task. Future studies should gather data from other tasks in other foreign language settings to assess whether the tendencies reported here can be observed in these new contexts. To provide further insights, it would be valuable for motivation questionnaires to be administered to gauge the learners’ opinions about the task and to analyse whether their attitude toward it could have an impact on the final performance.

Acknowledgements The two authors gratefully acknowledge the funding from research grants FFI2012-32212 (Spanish Ministry of Economy and Competitiveness), IT311-10 (Basque Government) and UFI11/06 (University of the Basque Country). Last, but not least, our sincere thanks to the schools that allowed us to gather the data reported in this chapter and to the teachers and children who participated in the study.

References Azkarai, A. and Imaz Agirre, A. (2016) Negotiation of meaning strategies in child EFL mainstream and CLIL settings. TESOL Quarterly 50, 844–870. Butler, Y.G. and Zeng, W. (2013) Young learners’ interactional development in task-based paired-assessment in their first and foreign languages: A case of English learners in China. Education 3–13: International Journal of Primary, Elementary and Early Years Education 43, 292–321. Coyle, D. (2007) Content and language integrated learning: Toward a connected research agenda for CLIL pedagogies. The International Journal of Bilingual Education and Bilingualism 10, 543–562. Dalton-Puffer, C. (2011) Content and language integrated learning: From practice to principles? Annual Review of Applied Linguistics 31, 182–204. Dalton-Puffer, C., Hüttner, J., Schindelegger, V. and Smit, U. (2009) Technology-geeks speak out: What students think about vocational CLIL. International CLIL Research Journal 1, 18–25. Enever, J. (2011) (ed.) ELLiE: Early Language Learning in Europe. London: British Council. García Mayo, M.P. and Alcón Soler, E. (2013) Negotiated input and output. Interaction. In J. Herschensohn and M. Young-Scholten (eds) The Handbook of Second Language Acquisition (pp. 209–229). Cambridge: Cambridge University Press. García Mayo, M.P. and Basterrechea, M. (in press) CLIL and SLA: Insights from an interactionist perspective. In A. Llinares and T. Morton (eds) Applied Linguistics Perspectives on CLIL. Amsterdam: John Benjamins. García Mayo, M.P. and García Lecumberri, M.L. (2003) (eds) Age and the Acquisition of English as a Foreign Language. Clevedon: Multilingual Matters. García Mayo, M.P. and Imaz Agirre, A. (2016) Task repetition and its impact on EFL children’s negotiation of meaning strategies and pair dynamics: An exploratory study. The Language Learning Journal 44, 451–466. García Mayo, M.P. and Lázaro Ibarrola, A. (2015) Do children negotiate for meaning in task-based interaction? Evidence from CLIL and EFL settings. System 54, 40–54.

Child EFL Interac t ion

267

García Mayo, M.P. and Villarreal Olaizola, I. (2011) The development of suppletive and affixal tense and agreement morphemes in the L3 English of Basque-Spanish bilinguals. Second Language Research 27 (1), 129–149. García Mayo, M.P. and Basterrechea, M. (2017) CLIL and SLA: Insights from an interactionist perspective. In A. Llinares and T. Morton (eds) Applied Linguistics Perspectives on CLIL (pp. 33–50). Amsterdam: John Benjamins. Gass, S.M. and Mackey, A. (2007) Input, interaction, and output in second language acquisition. In B. VanPatten and J. Williams (eds) Theories in Second Language Acquisition: An Introduction (pp. 175–199). Mahwah, NJ: Lawrence Erlbaum. Gass, S.M., Behney, J. and Plonsky, L. (2013) Second Language Acquisition: An Introductory Course. New York: Routledge. Jexenflicker, S. and Dalton-Puffer, C. (2010) The CLIL differential. In C. Dalton-Puffer, T. Nikula and U. Smit (eds) Language Use and Language Learning in CLIL Classrooms (pp. 169–190). Amsterdam: John Benjamins. Keck, C.M., Iberri-Shea, G., Tracy-Ventura, N. and Wa-Mbaleka, S. (2006) Investigating the empirical link between task-based interaction and acquisition: A quantitative meta-analysis. In J.M. Norris and L. Ortega (eds) Synthesizing Research on Language Learning and Teaching (pp. 91–131). Amsterdam: John Benjamins. Lázaro, A. and Azpilicueta-Martínez, R. (2015) Investigating negotiation of meaning in EFL children with very low levels of proficiency. International Journal of English Studies 15, 1–21. Long, M.H. (1983) Does second language instruction make a difference? TESOL Quarterly 17, 359–382. Long, M.H. (1996) The role of the linguistic environment in second language acquisition. In W.C. Ritchie and T.K Bhatia (eds) Handbook of Language Acquisition: Vol. 2. Second Language Acquisition (pp. 413– 468). New York: Academic Press. Mackey, A. (ed.) (2007) Conversational Interaction in Second Language Acquisition. Oxford: Oxford University Press. Mackey, A. and Goo, J.M. (2007) Interaction research in SLA: A meta-analysis and research synthesis. In A. Mackey (ed.) Input, Interaction and Corrective Feedback in L2 Learning (pp. 379–452). Oxford: Oxford University Press. Mackey, A. and Silver, R.E. (2005) Interactional tasks and English L2 learning by immigrant children in Singapore. System 33, 239–260. Muñoz, C. (2006) (ed.) Age and the Rate of Foreign Language Learning. Clevedon: Multilingual Matters. Muñoz, C. (2007) CLIL: Some thoughts on its psycholinguistic principles. Revista Española de Lingüística Aplicada (Resla) 20, 17–26. Muñoz, C. (2014) Exploring young learners’ foreign language learning awareness. Language Awareness 23, 24–40. Nikolov, N. and Mihaljević Djigunović, J. (2011) All shades of every color: An overview of early teaching and learning of foreign languages. Annual Review of Applied Linguistics 31, 95–119. Oliver, R. (1998) Negotiation of meaning in child interactions. The relationship between conversational interaction and second language acquisition. Modern Language Journal 82, 372–386. Oliver, R. (2000) Age differences in negotiation and feedback in classroom pairwork. Language Learning 50, 119–151. Oliver, R. (2002) The patterns of negotiation for meaning in child interactions. Modern Language Journal 86, 97–111. Oliver, R. (2009) How young is too young? Investigating negotiation of meaning and feedback in children aged five to seven years. In A. Mackey and C. Polio (eds) Multiple Perspectives on Interaction (pp. 135–156). New York: Routledge.

268

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Oliver, R. and Grote, E. (2010) The provision and uptake of different types of recasts in child and adult ESL learners: What is the role of age and context? Australian Review of Applied Linguistics 33, 1–22. Pica, T. (1994) Research on negotiation: What does it reveal about second language learning conditions, processes, and outcomes? Language Learning 44, 493–527. Pinter, A. (2007) Some benefits of peer-peer interaction: 10-year-old children practicing with a communication task. Language Teaching Research 11, 189–207. Pinter, A. (2011) Children Learning Second Languages. London: Palgrave Macmillan. Storch, N. (2002) Patterns of interaction in ESL pair work. Language Learning 52, 119–158. Swain, M. (1998) Focus on form through conscious reflection. In C. Doughty and J. Williams (eds) Focus on Form in Classroom Second Language Acquisition (pp. 64–81). Cambridge: Cambridge University Press.

15 Evaluating the Educational Outcomes of an Early Foreign Language Programme: The Design of an Impact Study for the Primary English Programme in Mexico Peter Sayer, Ruth Ban and Magdalena López de Anda

The Need to Evaluate Primary English Language Programmes This chapter describes the design of a large-scale evaluation of the pilot phase of an early English language programme in public elementary schools in Mexico. The researchers, at the behest of the Ministry of Education, were asked to document a range of outcomes to address the question of the broader effects of the programme on the students’ learning and educational experiences. Therefore, they conceptualised the project as an impact study using a mixed methods approach (Johnson et al., 2007). The development and implementation of the impact study of an early foreign language (FL) programme is described, including how the quantitative and qualitative components of the project were tailored to explore both the linguistic and non-linguistic outcomes. The design allowed the researchers to look quantitatively at the second language (L2) gains students made across a range of contexts and socioeconomic levels. The qualitative component examined the types of connections between studying English and students’ learning across other content areas in the curriculum, and the relation between the ways students used English inside and outside the classroom. The synergetic 269

270

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

result of utilising quantitative and qualitative data helped to inform the larger question of the value of early FL programmes on the educational and social experiences of Mexican children. Across the globe, there has been a steady trend towards more children studying English for longer periods of time during their schooling and starting from a younger age (Enever & Moon, 2010). This ‘English for everyone’ (Hamid, 2010) or ‘more and earlier’ approach (Sayer, 2015) has been largely compelled by ministries of education who have included English as part of their public primary curricula. In developing countries, the decision to initiate primary English language teaching (PELT) programmes is based largely on two premises. First, the notion that for individuals, English skills equate with greater economic opportunity and hence social mobility. Likewise, at the national level, it is commonly accepted that a country surely needs to have a large number of its citizens to be competent in English in order for the nation to be economically competitive in the global marketplace. Second, PELT programmes are based on a folk theory of second language acquisition (SLA), whereby children are considered ‘language sponges’, who can soak up a foreign language with ease and hence it makes sense to start them as early as possible. Although both of these premises for early English programmes are commonly accepted (Matear, 2008; Park & Wee, 2012), in fact there is a lack of empirical evidence about what types of programmes, pedagogies and materials are effective. Writing about the PELT programme in China, Knell and colleagues (2007) begin their paper by acknowledging: ‘Instruction in English as a foreign language at an early age is becoming more common world-wide even though the effects of this early instruction are not yet known’ (Knell et al., 2007: 395, emphasis added). Most work in SLA has traditionally focused on adult learners and the true nature of the ‘critical period hypothesis’ has long been debated (Birdsong, 1999). Much of the work in SLA with younger learners has focused either on bilingual immigrant children (e.g. Cummins, 2000; Curtain & Dahlberg, 2010) or on instructed contexts in private schools and in Europe (Muñoz, 2006). Most of the recently implemented PELT programmes in developing countries, however, are English as a foreign language (EFL) programmes in public schools. They typically provide students with minimal exposure, generally only two to three hours per week of instruction. Given the massive global investment in PELT, especially in resourcestrapped developing countries, researchers need to respond to the question: what are the effects of minimal exposure English programmes as part of the public primary school curriculum? In this chapter, we describe the design of an impact study of the PELT programme in Mexico. The impact study methodology was developed to document the many linguistic, educational and social learning outcomes that a PELT programme could have, to account for how the social context shapes the programme and to include the voices of all the stakeholders involved. We describe the characteristics of the programme,

Evaluat ing the Educat ional Outcomes of an Early FL Programme

271

the design of the research using mixed methods and briefly highlight some of the main findings from the project in Mexico.

The English Programme in Mexico The national programme in Mexico was launched in 2009 to introduce English as an additional language in public primary schools for all grades kindergarten through six (K-6). Historically, English instruction in the early grades had been limited to private schools and students in public schools did not study English until lower secondary (seventh grade, about age 12 years old). By the mid-2000s, several of the 32 states had launched state programmes for English and many others had small, local initiatives. In other schools, the principal worked with the parent associations to hire someone who knew English to teach extracurricular classes. Hence some students had access to English, but most did not and there were a wide variety of approaches and materials used of varying quality. The national programme aims to provide classes to all students, and unifies instruction under a single curriculum. This ambitious plan represents the largest expansion of English instruction in Mexico’s history, and the Ministry of Education estimates that when fully implemented it will require 98,300 teachers to instruct 14.7 million K-6 students across the country. Beginning in kindergarten, students are to receive 2½ hours of instruction per week, or about 100 hours each year. The curriculum projects the students’ progression in terms of the levels based on the Common European Framework of Reference (CEFR) scale. With 700 hours of instruction, students should reach an A2 level by the end of elementary school, as shown in Figure 15.1. The programme was introduced as part of a wider curriculum reform, which reoriented all subjects taught at the elementary and middle school

Figure 15.1 Progression in the Mexican programme Source: Adapted from Mexican English Curriculum (SEP, 2010).

272

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

levels towards a sociocultural approach. This means that contents are organised around ‘learning environments’ (the home/community, ludic/play and academic/literary) and for English, that the language is situated within social practices of language (Vargas & Ban, 2011). The overall vision of the English programme is expressed by the Ministry as: The articulation of the teaching of English in all three levels of Basic Education [pre-school, elementary, and middle] has the aim to guarantee that, by the time students complete their secondary education, they will have developed the necessary multilingual and multicultural competencies to face the communicative challenges of a globalized world successfully, to build a broader vision of the linguistic and cultural diversity of the world, and thus, to respect their own and other cultures. (Secretaria de Educación Pública, 2010: 21) While the programme aims to significantly expand access to English for all Mexican children, its implementation has not been without challenges. Although it has been 90–100% implemented in several states, in other areas it is only operating in a small percentage of schools and in some cases been suspended because of administrative problems. On average, as of 2016, perhaps 25% of Mexican elementary students receive some English instruction, mostly in the urban areas. Amongst the challenges are problems with production and distribution of textbooks and materials, the difficulty in expanding the programme to rural areas, particularly in indigenous communities, the creation of the administrative infrastructure to ensure that teachers receive timely payments and benefits, and perhaps most importantly, the availability of qualified teachers who have sufficient English proficiency (Ramírez Romero et al., 2014). The hurdles the Mexican programme is experiencing are shared by other countries, where similar programmes have been launched. Other authors have noted that implementing PELT programmes in developing countries require large investments of resources and the shortage of qualified teachers is recognised as the main obstacle to building effective programmes (Davies, 2009; Kaplan et al., 2011; Wu, 2012).

A Mixed-Methods Impact Study This chapter describes the design of a large-scale evaluation of the pilot phase of an early English language programme in public elementary schools in the central state of Aguascalientes in Mexico. The impact study methodology was developed to document the many linguistic, educational and social learning outcomes that a PELT programme could have, to account for how the social context shapes the programme and to include the voices of all the

Evaluat ing the Educat ional Outcomes of an Early FL Programme

273

stakeholders involved. In fact, the methodology was developed organically over several years, as we the researchers worked with the Ministry administrators and programme coordinators to define what a comprehensive evaluation of the programme should consist of.1 On the one hand, there was a stated need to assess the students’ progress in terms of the Common European Framework of Reference (CEFR) standards, both to provide an objective measurement of the overall linguistic gains, as well as to be able to identify and certify students who surpassed the required levels. On the other hand, the goal was to evaluate the programme in order to be able to carry out curriculum development and build teacher training courses, so we recognised that there were many other elements of the programme that ought to be evaluated in order to inform these activities. Richards (2001) explains that in language programme or curriculum evaluation there are three types of approaches: formative, illuminative and summative evaluations. He lists 13 aspects that can be evaluated, ranging from student learning outcomes to classroom processes, to materials and teacher training (Richards, 2001: 286– 287). The aim of the impact study was to balance the formative, illuminative and summative goals. Rather than presume a priori which aspects were most important, we included various stakeholders so that they could tell us which aspects to prioritise. We started the project by looking carefully at the curriculum document itself and talking to leaders in the Ministry. In the general objective cited above, the goals are stated in terms of ‘multilingual and multicultural competencies’, ‘communicative challenges’, ‘a broader vision of the linguistic and cultural diversity of the world’ and ‘respect [of] their own and other cultures’. Our overarching orienting question then became: how well is the programme fulfilling this vision? We realised that the answer to this question was going to have many aspects and to capture the complexity we would need to include both quantitative and qualitative components, and include as many different types of participants as possible: students, English teachers, classroom teachers, parents, principals, programme coordinators and administrators. We also studied other models and standards in language education. In particular, we found that the sociocultural approach as defined by the Mexican English programme fitted well with the standards of foreign language programmes defined by the American Council of Teachers of Foreign Languages (ACTFL). The ACTFL’s standards are called the ‘Five Cs’, for communication, communities, connections, comparisons and cultures (ACTFL, 1996). We added a sixth ‘C’, curriculum/programme, to include elements that related directly to the Mexican curriculum. Within this framework, we generated a more detailed set of research questions and aligned each question to the type of data we would collect. An example of one area is given in Table 15.1; note how the areas (cultures) and the ACTFL standards (2.1 and 2.2) align with its corresponding standard from the Mexican curricula, to the

274

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Table 15.1 Alignment of framework, competencies and research questions Framework and standards

Competencies defined by the Mexican curriculum

Research questions

Data sources

2. Cultures 2.1: Students demonstrate an understanding of the relationship between the practices and perspectives of the culture studied.

Students will recognise and respect differences between their own and other culture where English is spoken.

Q 2.1: How has the English programme contributed to students’ development of global vision?

Interviews with focal groups of students Interviews with English teachers Interviews with classroom teachers

2.2: Students demonstrate an understanding of the relationship between the products and perspectives of the culture studied.

research question(s) we posed for that standard, and to the sources of data that we collected to address the question. We then developed the instruments and protocols for each data source. For each project, we spent an intensive two weeks in the schools collecting data, followed by several months of transcribing, coding the qualitative data using NVivo and compiling the results of the quantitative measures. The main unit of analysis for the project was the school site. Therefore, a key aspect of the project was the selection of the schools that would constitute the sample. For each state, we chose 10 to 15 schools and the research team spent a full day at each site. The schools were selected to encompass as much diversity as possible, so as to allow us to compare across sites on the basis of urban/suburban/rural, socioeconomic level and other characteristics of Mexican schools (whether it was morning or afternoon shift and in a community identified as having ‘high migration’ to the United States). The research team generally consisted of five to six members. We were all outsiders; that is, we were not teachers or administrators in the programme we were studying, though we had worked in English language teaching in Mexico for many years. The first two authors served as the principal investigators (PIs), directed the research team in the field, and conducted most of the observations and some of the interviews. They were also responsible for the analysis of the qualitative data. The third author was the project manager. She was the direct contact with the administrators in the Ministry and worked with the PIs to conceptualise the research questions

Evaluat ing the Educat ional Outcomes of an Early FL Programme

275

and design the methodology, supervised the administration of the quantitative instrument and did the analysis of the quantitative elements. The other members of the research team conducted the interviews with the children and administered the assessments. There were additional support members, especially for transcribing interviews and organising the logistics of school visits (getting permission forms, transportation between sites and organisation of digital files). Due to the scope of the project, the difficulty in arranging dates to fit the public school schedule and the relatively short intensive nature of the data collection, it was necessary to coordinate carefully beforehand all the pieces of the project. Admittedly, this became easier as we did several projects, making many mistakes along the way, but refining our methodology and becoming better able to anticipate the problems we were likely to encounter in the field.

Quantitative measures The nature of the different types of data called for a mixed methods approach. Data about students’ proficiency levels were obtained by applying standardised language proficiency tests2 to a representative sample of sixth and eighth grade students, a quantitative measure. We also applied a survey to students; we wanted to find out how many of them studied English in private classes outside of school, whether they had ever lived in or travelled to the United States or other English-speaking countries, their level of satisfaction with their English classes and how much exposure they had to English outside the classroom in order to correlate their responses to the results on the test. Based on earlier phases of the project we had found that for a given school, the success of the programme depended largely on how much support the principal gave to the English teachers. Since primary English teachers in Mexico are usually itinerant (they may be assigned to two or more schools and usually ‘teach off the cart’ and do not have their own classrooms), they depend heavily on the principal and regular classroom teachers to provide the physical and temporal spaces they need to work effectively. The principals’ survey allowed us to gauge how much support the teachers and programme were receiving at the school level. Finally, a question had arisen from parents and teachers about the extent to which the kindergarten and first grade contents of the English class emphasised reading and writing. There was a concern expressed by some that because children did not yet know how to read in their native language, Spanish, that early exposure to English – where the letters represented a different set of sounds – might cause confusion. Others felt that early exposure to another writing system could actually help young readers learn phonemic awareness faster (the association of graphemes to phonemes, or sound-symbol correspondence). Again, we decided to test this quantitatively, by designing an instrument in Spanish.3 We gave the test to 60 first grade students from

276

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

four matched schools: two with the English programme and two where the students had not studied English. The results indicated that generally a student’s participation in the English programme (after one and a half years or about 300 hours of instruction from ages 5–6 years old) had a negligible or slightly positive affect on their development of literacy skills in their L1.

Qualitative components The main sources of qualitative data came from interviews, classroom observations and primary source documents. The qualitative aspects of the impact study responded to the need to document the processes. These helped explain the results of the quantitative measures, for example, why some schools or students had scored better on the proficiency test and it also allowed us to capture many of the social dimensions of the programmes that could not be measured quantitatively. The students talked about what they liked to do with English outside of school and how they saw it as part of their everyday lives. The English teachers described how they understood and tried to implement the new programme and what their relationship was like within the school with the principal and other classroom teachers. We felt that, particularly at the younger grades, it was important to include the views of parents, who are often not taken into account in research projects. The parents discussed what their aspirations were for their children and often the difficulties they had trying to help their children learn a language that they themselves did not know. We considered the interviews our primary source of data for the impact study, since most of the research questions were addressed by speaking directly with key stakeholders. In retrospect, placing interview data at the centre of our study probably reflected what Johnson et al. (2007: 123) call our ‘home’ as researchers; that is, we identify ourselves as qualitative researchers using a mixed methods design because it fitted how the impact study project evolved. Hesse-Biber (2010) refers to this approach as qualitative mixed methods. She asserts that as mixed methods approaches have emerged and the advantages of combining data types has been accepted, there is nevertheless a quantitative and positivist bent to most mixed methods work, with the qualitative data serving as the second fiddle that ‘often takes the form of sprinkling in some vignettes to provide narrative examples of the conclusions reached by means of quantitative methods’ (Hesse-Biber, 2010: 457). She addresses eight areas where qualitative research can benefit from incorporating mixed methods. In our study, while the test scores and survey results – the quantitative information – provided important baseline information to help us choose our sites and develop our interview protocols, we considered the qualitative components as the primary sources of data, since they more directly helped us answer the ‘how’ and ‘why’-type research questions the impact study was designed around (see research questions in Table 15.1).

Evaluat ing the Educat ional Outcomes of an Early FL Programme

277

At each school, we individually interviewed the English teachers, the principal, at least one classroom teacher and carried out focal group interviews with students and parents. The interviews were semi-structured, in order to strike a balance between asking a uniform set of questions tailored to the research objectives, and included an element of exploratory research by allowing for open-ended responses (King & Horrocks, 2010). One of the most important sources of data was the interview with parents and students. Kvale (2007: 68) states that ‘Interviews with children allow them to give voice to their own experiences and understanding of their world’. We interviewed at three points in the programme: kindergarten, second and sixth grades. We decided to interview the children in focal groups of three or four, to give each student as much time as possible to talk, while at the same time trying to create a comfortable atmosphere with their peers so they did not feel ‘put on the spot’ in having to answer. We chose a quiet area in the library or study area so they knew teachers or school administrators would not overhear. The oldest children we interviewed were sixth graders (about 11–12 years old) finishing primary school, many of whom were the first generation to have studied English in public schools since kindergarten. We asked them what they liked and did not like about their English classes, to grade their teachers and what type of activities helped them learn best. We also asked their connections with English in their everyday lives and many reported they had family living in the United States. In interviewing children, it can often be effective to embed questions within the context of some more concrete task, as Piaget suggested in the 1930s in his work on children’s developmental stages (Kvale, 2007; cf. Pinter, 2006). For the older students, we asked them to imagine that they had been asked to work as volunteers in an information booth at the regional festival and that some foreigners approached them speaking English. The scenario allowed them to conjecture about what the foreigners would ask, what they would do to try to communicate with them and what sort of cultural differences they would encounter when visiting Mexico. For the younger children, we used a set of puppets depicting people from various cultures. We asked young ones to pick out which puppet they thought could speak English, to name the puppet and explain where she was from. In Figure 15.2, Regina, an 8-year-old girl in second grade who lives in town in central Mexico, choose the blonde doll in a pink dress: Interviewer: Regina, which doll do you think speaks English? Regina: [Chooses the blond puppet]. This one. Interviewer: That one? Take it. Okay, why do you think that one speaks English? Regina: [Thinks for a moment]. Well, you can tell from her face that she’s speaking English. Interviewer: Yeah? And where do you think she’s from? Regina: From the United States.

278

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Figure 15.2 The puppet study

These tasks allowed us evaluate to what extent the students’ study of English was enabling them to make broader cultural connections and understanding as envisioned in the curriculum (see again above the Ministry’s objective statement and Table 15.1). We also compared the students’ responses to what we were observing in the classroom observations. Like our interview protocols, our observation sheets were semi-structured, including elements that the programme administrators focused on, but allowing us to pick up on whatever proved salient. For example, the kindergarteners we talked to were forming clear ideas about cultural differences between Mexicans and others, and had ideas about who spoke English. By eight years old, the second graders had even stronger notions about why people spoke English and how they were different. However, we observed very few instances where the English teacher served to help mediate children’s understandings of cultural elements and the cultural content of the classes was usually limited to superficial aspects prescribed by the book, such as a comparison of the American holiday Halloween with the Mexican celebration of Day of the Dead. The children’s notions therefore seemed to draw from popular culture – especially cartoons and movies for young children – and often conflated English with a general notion of ‘foreign-ness.’ For instance, many children said that their puppet spoke English because they were from China or Japan, although this did not necessarily correspond to their choice of a doll with Asian features. The primary source documents consisted of the textbooks and teachergenerated materials, the teachers’ lesson plans and examples of student work. We also took photographs to document if and how English had a presence on the walls of the classrooms and hallways. Brown (2012) refers to visual display of texts in schools as the schoolscape. We found that one feature of schools with a strong English programme was that the principal supported

Evaluat ing the Educat ional Outcomes of an Early FL Programme

279

the English teacher and promoted the programme with parents by allocating spaces in the school for English to be prominently displayed, such as studentcreated murals or signage in English visible from the entrance where parents dropped off their children.

Relation between quantitative and qualitative elements An impact study should integrate all the components of the project to build the ‘big picture’. Similarly, a mixed methods study is not simply a quantitative study plus a separate qualitative study, but rather the elements of mixed methods research should be articulated so that together they conform and respond to the particular questions to be asked. The mixed or blended design we settled on reflected what we considered the best method for addressing each research question, or what Creswell and Tashakkori (2007: 307) calls a ‘bottom-up’ approach to mixed methods. Patton (2015: 319–320) refers to this approach as a ‘mixed design strategy’ and explains that although there are many ways of combining qualitative and quantitative elements, the end result should be a design that reflects a ‘coherent strategy’. The elements of our project were articulated in several ways. Since we began by applying the language proficiency test prior to the site visits, when we began the qualitative phase, we had a clear map of what the scores were for students and schools. We could therefore know when we arrived if a school was ‘special’ in some way, for example, a school in a working-class neighbourhood with very high English scores, and ask, how had they achieved that? Was it something special the teacher was doing in her classroom? Was it because the principal had worked with the parents to create an extracurricular after-school programme? At each school, we also knew which students had scored the highest on the test, as well as additional information about them from the survey, such as whether they had ever lived in or travelled to an English-speaking country. We used this information when selecting students for the focal group interviews and organised at least one group of the high achieving students and one group who had about average scores. We chose at least one group of students at each school who had scored high on the test, but indicated they had never travelled outside of Mexico. We asked them what strategies they had used and found that at 10–12 years old they were invested substantially in their learning by routinely looking for authentic uses of English outside of school, such as studying song lyrics, saving money to buy books in English and engaging in online gaming communities (Sayer & Ban, 2014).

Some Findings of the Impact Study in Mexico Each project had somewhat different results and many of the findings were quite specific to issues related to the Mexican curriculum and structural components of the programme in the particular context we were studying. The purpose of the impact study was to provide administrators with a view

280

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

of the educational processes of the programme to inform curriculum development and teacher training. However, a secondary purpose for us as researchers was to try to understand how our findings might be generalised or transferred to other contexts; that is, how the projects in the Mexican programme might inform our understanding of how to more effectively organise PELT programmes. Our findings were of seven types: linguistic outcomes, learning/educational outcomes, social outcomes, implementation of the curriculum, level of satisfaction, organisational aspects and broader impacts. We briefly explain these findings below with some examples of each.

Linguistic outcomes The most common metric for measuring a language programme’s success is in terms of students’ L2 learning of vocabulary, grammar and language skills. Even though curricular objectives are often stated more broadly, as in the Mexican case which speaks of a ‘broader vision of a globalised world’, the students’ proficiency level seems to remain the main criterion of success. In the impact study, we used the results on the standardised assessments, scaled to the CEFR levels, as a referent for the qualitative elements of the programme. This allowed us to focus on, for example, how some students who had never travelled outside of Mexico or studied private English classes were able to achieve much higher than their classmates. For the sixth graders taking the proficiency, our premise was that very high-achieving students (level B1 or better) had probably lived in the United States, studied in private schools or language academies. However, we found that over half (54%) of these ‘top performers’ had only studied in the public school programme, a result which indicates the potential of even limited instruction EFL programmes (see Figure 15.3). By asking these students in the interviews what they did to learn, we identified the context-specific strategies that the Type of School/Program 200 180 160 140 120 100 80 60 40 20 0

185 87

Public

Private

25

20

16

5

Pubic and Private

Public and Foreign

Foreign (in US)

Private and Foreign

2

1

Combinaon No response and Foreign

Figure 15.3 Where did the highest achieving students learn English? Note: Results based on sixth-grade students (n = 341) who scored B1 or higher on proficiency test.

Evaluat ing the Educat ional Outcomes of an Early FL Programme

281

successful English learners in Mexican public schools use. Likewise, from the quantitative data we were able to identify which low socioeconomic status (SES) schools had more students scoring well on the test, and the site visits allowed us to examine which aspects of how the school organised their programme made them more successful. For the younger children, we measured linguistic outcomes using the puppets and a colourful poster in groups of three. This made the assessment more relaxed and interactive. We found that children often had a hard time transferring what they had seen in class to a slightly different, more communicatively-oriented task. For example, although we had observed children in many classrooms singing ‘Head, shoulders, knees and toes, knees and toes… Eyes and ears and mouth and nose…’ as a warm-up song while doing the corresponding body movements, they often could not identity these body parts on themselves or their puppets. Likewise, although they could recite most of the numbers and colours, only a few students aged 7–8 years old who had been in the programme for over two years could respond to questions such as ‘How old are you?’ or ‘What’s your favourite colour?’ This suggests that, although the curriculum calls for a communicative approach, the PELT programme in Mexico still tends to present linguistic elements in isolation. For younger children, this is not presented as explicit grammar instruction, but usually takes the form of thematic vocabulary such clothing, animals or transportation. Whereas songs, chants and rhymes are also common classroom warm-up and transition activities, they are rarely integrated to reinforce L2 learning, such as the ‘Head and Shoulders’ song example.

Learning/educational outcomes One encouraging finding was that English lessons often help students build connections to content learning in other subject areas. The main idea behind content-based instruction (CBI) or content-language integrated learning (CLIL) approaches, is that English can be effectively acquired through the learning of other content areas.4 Although the CLIL approach is not used in Mexican public schools, we found that in the EFL programme the reverse is often true: that students learn content through their English class. Most often, this took the form of reinforcing things learned first in Spanish. In the kindergarten, the children learned numbers, colours, shapes and the days of the week, which consolidated the knowledge of the concepts from Spanish. The results from the early literacy instrument reported above were another key area where learning in English helped the development of parallel abilities and knowledge in the students’ first language. For older children, we found intercurricular connections in language arts (e.g. narrative story elements), science (e.g. the steps of the scientific method), and social studies and geography (e.g. maps and flags of countries).

282

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Figure 15.4 A student’s notebook describing a science experiment

However, whereas regular classroom teachers teach all the subjects and can be more aware of where intercurricular connections exist, English teachers are usually more narrowly focused on their subject and may miss opportunities to help students make connections between learning across areas (Sayer & Ban, 2013). Figure 15.4 shows a student’s notebook from an experiment the children had conducted to describe the growth of plants. The teacher explained that it was to teach the children the sequence words: first, next, then, after, finally. However, when we asked her if the children had studied life cycles in their science class, she admitted she had taught the same lesson several times, but had never really thought of it as a science lesson or of asking students to connect it to what they knew about life cycles. Rather, she had only approached it as a language lesson in terms of its linguistic goals.

Social outcomes The interviews with parents and students illuminated the many social connections that exist in Mexico through English. In particular, almost all students reported to have family members, in some cases siblings or parents, but most often aunts, uncles and cousins, living in the United States. By 10–12 years old, many students were connected to their family members through social media; for example, Facebook, Snapchat, Instagram, Skype and WhatsApp. As with the intercurricular connects, we found that the social connections that students had with English were largely unrecognised or unexploited by teachers in the programme. Specifically, student responses were analysed quantitatively (see list below) to provide insight into their use of English outside the classroom;

Evaluat ing the Educat ional Outcomes of an Early FL Programme

283

conversely, qualitative teacher responses uniformly expressed that the only English input the students had was in the classroom. Clearly, this example of the how the application of both qualitative and quantitative data allowed for the researchers to appreciate this lack of understanding by the teachers. • • • • • • • • • • • • • • • •

Listening to songs/music in English (n = 18, from pop to thrash and reggaeton). Watching movies in English (n = 15, especially without subtitles). Video games (n = 15, many popular Xbox games come with English instructions). Internet (n = 8, to access English sites). Google Translate (n = 5, as a tool to translate words or texts in English). Television and cartoons (n = 5, using the options to put the soundtrack in English). Books (n = 4, in English, some sent by family members in the USA). Clothing tags (n = 3, reading tags on clothes from the USA). Communicating with family members in the USA (n = 2). Reading instruction manuals of products that are in English (n = 2). Watching videos on YouTube in English (n = 2). Writing messages on Facebook (n = 2). Watching iPod TV (n = 1). Singing karaoke songs in English (n = 1). Taking a Disney English course on television (n = 1). Writing emails in English with Mexican and foreign friends (n = 1).

This was somewhat surprising, since the curriculum itself called for a sociocultural approach based on social practices of language. However, the social practices came mostly from the textbook and not from the lived experiences of the students. So ‘writing an invitation to a party’ was an academic exercise for a fictional audience, rather than say an actual invitation students could write and send to cousins living in California.

Implementation of the curriculum The launch of the English programme in Mexico coincided with a major educational reform which reoriented teaching of all subjects within a ‘sociocultural approach’. The new English curriculum adopted the Ministry’s approach. The sociocultural approach, defined in the programme in terms of competencies, social practices and learning environments, was fairly openended and allowed a good degree of teacher autonomy in lesson planning. Therefore, the concern of the project was not to determine the degree of teacher fidelity to the model, which would have been impossible given the lack of specifics about the teaching methods teachers were supposed to use, but rather to look at what teachers understood by ‘sociocultural approach’, how they related that to students’ learning of English and how they were trying to put that into practice in the classroom.

284

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

We found that many teachers had not received any training about the programme and so they brought their own approach from their previous job – many had worked in private schools – or pre-service training. Those who had been trained in communicative language teaching had a good (around B2) level of English and had previous teaching experience often liked the curriculum, precisely because it was not prescriptive. Less experienced teachers reported that the programme was too vague and usually ended up relying almost exclusively on the textbook. For many teachers, one of the main challenges was not with the curriculum itself, but rather with classroom management. They had often been trained to work in secondary schools or with adults, since programmes to prepare English teachers to work specifically with children still do not exist. Hence, they felt ill-prepared to address behaviour issues, children with special needs and other challenges common to public primary schools in Mexico. We also found that the curriculum’s objectives for each grade level were not always consistent with the contents of the programme and less so with the textbooks. In particular, they tended to focus on the language skills and linguistic elements of the programme and mostly neglect the cultural and social aspects. Overall, this may reflect a conceptual problem with the programme. Early foreign language programmes can be organised as foreign language exploratory or experience (FLEX) or foreign language in elementary school (FLES) (Shin & Crandall, 2014). In a minimal exposure setting, supposedly two and half hours ,but sometimes less than two hours per week in the Mexican programme, it may make more sense to begin in K-3 as a FLEX programme and then adopt a FLES model for the higher grades.

Level of satisfaction The impact study also allows evaluators to gauge the vitality of a programme by documenting the various stakeholders’ level of satisfaction. In the Mexican programme, this was important because the programme was just finishing the pilot stage and administrators wanted to know how much support there was for continuing or expanding the programme. In particular, in order to have the programme run well at the school level, they depended on the ‘buy in’ from parents and principals. In our project, this was done qualitatively through the interviews. Although there was overall support for the programme, because English is seen as a necessary 21st century skill, participants were critical of many of the particulars of how the programme had been implemented.

Organisational aspects Although initially we had, as language teacher trainers ourselves, conceptualised the project more in terms of the academic and pedagogical aspects; we soon realised that the effectiveness of the programme in the classroom

Evaluat ing the Educat ional Outcomes of an Early FL Programme

285

depended on a range of organisational factors that had not yet been worked out. The distribution of textbooks, the allocation of physical spaces in the schools, very large class sizes and class scheduling all impacted how well the teachers could deliver their lessons. For example, in a school where there was little support for the programme from the principal, the regular teachers often felt it was okay to use the time allocated for English lessons to organise students for other extracurricular events, such as the president’s birthday. If there were no lines of communication between English teachers and their homeroom colleagues, the English teacher – rotating between eight or more classrooms each day and transporting her materials in wheeled crate – would often arrive breathless to her next class to find that the regular teacher had cancelled the English lesson. By contrast, in other schools the English teacher was integrated into the faculty and the school Christmas festival for parents would showcase children singing carols in English. Finally, since the programme was still in its pilot phase, most importantly among the organisational aspects was the fact that most of the teachers did not have regular contracts, which generated uncertainty and stress about whether they would receive a paycheck, job security, teaching assignments and ability to address problems with working conditions.

Broader impacts Many parents stated that they were pleased that their children had a chance to take English classes because they could not afford to send them to private lessons. This seemed to speak to the broadest impact of the programme; that is, greater equity in access to learning English. There are still issues with the quality of the programme in the public school and the programme certainly does not address problems in socioeconomic disparities between the public and private education systems in Mexico. Also, we should not immediately assume that a ‘more and earlier’ or ‘English for everyone’ is automatically the right approach for English education (Sayer, 2015). An impact study therefore can document the benefits and challenges of a programme that can serve as a referent to weigh against the costs and resources that need to be invested.

Conclusion In this chapter, we have described a language programme evaluation project conceptualised as an impact study. An impact study design can effectively incorporate quantitative and qualitative elements to provide a holistic picture of a primary English language teaching programme. It combines formative, illuminative and summative approaches (Richards, 2001) to language programme evaluation. The project we described in Mexico was

286

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

a larger-scale study sponsored by the Ministry of Education and involved a research team working across multiple school sites. The sites had been selected to represent the diversity of contexts across a whole state; however, there is no reason that an impact study cannot be done on a smaller scale. It could be organised as a case study of one school and could involve the participation of teachers-as-researchers. One key element of the impact study in elementary schools that should be consistent is the inclusion of the voices of multiple stakeholders, especially of students and parents. Our project used a mixed methods design, with quantitative and qualitative data sources addressing different research questions, although interviews were the primary source. The elements were integrated in several ways as well, such as the use of test results to select participants for interviews. We summarised our results in seven different areas, including language, learning and social outcomes, as well as aspects of the curriculum and programme itself. An impact study in a different context, with distinct research questions, could therefore integrate the mixed methods differently depending on what the researchers judged would produce a coherent design (Patton, 2015). One point of constructive feedback that we received was the question of whether the project was really looking just at impacts. For example, in documenting how children used English in their everyday lives (Sayer & Ban, 2014), was that actually an impact or result of the programme, or something they would do anyway, regardless of whether they had English classes? This is a fair critique, since the naturalistic design of the project does not have a way of isolating what are strictly outcomes of the instruction children received in the English classroom. What the study can do is document these types of engagements that students have with the language outside of school and ask: if this is the children’s authentic use of English, how can the programme support these practices? How can we align our teaching to these practices? What other practices or opportunities can we open up for students?

Notes (1) The impact study methodology described here was developed as part of a national study of the pilot program in 2009–2012, and refined during projects carried out in the states of Aguascalientes (2013–2015), Durango (2014–2015) and Estado de México (2015). (2) Since language proficiency tests are difficult and expensive to design and validate, we contracted with the Cambridge Examinations Syndicate of Mexico to design tests for Mexican public schools based on their Starters and Movers tests for young learners and normed to the CEFR scale. (3) The instrument combined Woodcock and Muñoz-Sandoval (1995) and following Páez and Rinaldi (2006). (4) Content-based instruction, such as the Sheltered Instruction Observation Protocol (SIOP) model (Echevarría et al., 2013) is commonly used in classrooms with ESL students in the United States. CLIL is a similar approach more widely used in Europe. In Mexico, the approach is used somewhat in bilingual private schools, but not at all in public schools, where methods are based a communicatively-oriented EFL model.

Evaluat ing the Educat ional Outcomes of an Early FL Programme

287

References American Council of Foreign Language Teachers (ACTFL) (1996) Standards for Foreign Language Learning: Preparing for the 21st Century (3rd edn). Alexandria, VA: Allen Press. Birdsong, D. (ed.) (1999) Second Language Acquisition and the Critical Period Hypothesis. Mahwah, NJ: Lawrence Erlbaum. Brown, K. (2012) The linguistic landscape of educational spaces. In H. Marten, D. Gorter and L. van Mensel (eds) Minority Languages in the Linguistic Landscape (pp. 281–298). New York: Palgrave McMillan. Creswell, J.W. and Tashakkori, A. (2007) Differing perspectives on mixed methods research. Journal of Mixed Methods Research 1 (4), 303–308. Cummins, J. (2000) Language, Power and Pedagogy: Bilingual Children in the Crossfire. Clevedon: Multilingual Matters. Curtain, H. and Dahlberg, C.A. (2010) Languages and Children – Making the Match: New Languages for Young Learners, Grades K-8 (4th edn). Boston: Pearson. Davies, P. (2009) Strategic management of ELT in public educational systems: Trying to reduce failure, increase success. TESL-J 13 (3), 1–22. Echevarría, J., Vogt, M. and Short, D.J. (2013) Making Content Comprehensible for Elementary English Learners: The SIOP Model. Boston: Pearson. Enever, J. and Moon, J. (2010) A Global Revolution? Teaching English at Primary School. London: British Council. Hamid, M.O. (2010) Globalisation, English for everyone and English teacher capacity: Language policy discourses and realities in Bangladesh. Current Issues in Language Planning 11 (4), 289–310. Hesse-Biber, S.N. (2010) Qualitative approaches to mixed methods practice. Qualitative Inquiry 16 (6), 455–468. Johnson, R.B., Onwuegbuzie, A.J. and Turner, L.A. (2007) Towards a definition of mixed methods research. Journal of Mixed Methods Research 1 (2), 112–133. Kaplan, R., Baldauf, R.B. and Kamwangamalu, N. (2011) Why educational language plans sometimes fail. Current Issues in Language Planning 12 (2), 105–124. King, N. and Horrocks, C. (2010) Interviews in Qualitative Research. London: Sage. Knell, E., Haiyan, Q., Miao, P., Yanping, C., Siegel, L.S., Lin, Z. and Wei, Z. (2007) Early English immersion and literacy in Xi’an, China. The Modern Language Journal 91 (3), 395–417. Kvale, S. (2007) Doing Interviews. Thousand Oaks, CA: Sage. Matear, A. (2008) English language learning and education policy in Chile: Can English really open doors for all? Asia Pacific Journal of Education 28 (2), 131–147. Muñoz, C. (2006) The effect of age on foreign language learning: The BAF Project. In C. Muñoz (ed.) Age and the Rate of Foreign Language Learning (pp. 1–40). Clevedon: Multilingual Matters. Páez, M. and Rinaldi, C. (2006) Predicting English word reading skills for Spanishspeaking students in first grade. Topics in Language Disorders 26 (4), 338–350. Patton, M.Q. (2015) Qualitative Research and Evaluation Methods (4th edn). Thousand Oaks, CA: Sage. Park, J.S. and Wee, L. (2012) Markets of English: Linguistic Capital and Language Policy in a Globalizing World. New York & London: Routledge. Pinter, A.M. (2006)Teaching Young Language Learners. Oxford: Oxford University Press. Ramírez Romero, J.L., Sayer, P. and Pamplón Irigoyen, E.N. (2014) English language teaching in public primary schools in Mexico: The practices and challenges of implementing a national language education program. International Journal of Qualitative Studies in Education 27 (8), 1020–1043. Richards, J. (2001) Curriculum Development in Language Teaching. Cambridge: Cambridge University Press.

288

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Sayer, P. (2015) ‘More & earlier’: Neoliberalism and primary English education in Mexican public schools. L2 Journal 7 (3), 40–56. Sayer, P. and Ban, R. (2013) What students learn besides language: The non-linguistic benefits of studying English as a foreign language in primary school. Mextesol Journal 37 (3), 1–17. Sayer, P. and Ban, R. (2014) Young EFL students’ engagements with English outside the classroom. ELT Journal 68 (3), 321–329. Secretaria de Educación Pública (SEP) [Minsitry of Public Education] (2010) Programa nacional de inglés en educación básica (PNIEB): Fundamentos curriculares de asignatura estatal lengua adicional. Mexico City: SEP. Shin, J.K. and Crandall, J. (2014) Teaching Young Learners English. Boston, MA: National Geographic Learning & Heinle-Cengage. Vargas, R. and Ban, R. (2011) Paso a paso con el PNIEB en las aulas [Step by Step with the PNIEB in the Classroom]. Melbourne, FL: Latin American Educational Services. Woodcock, R.W. and Muñoz-Sandoval, A.F. (1995) Woodcock Language Proficiency BatteryRevised: Spanish Form. Itasca, IL: Riverside Publishing. Wu, X. (2012) Primary English education in China: Review and reflection. In B. Spolsky and Y. Moon (eds) Primary School English-Language Education in Asia (pp. 1–22). New York & London: Routledge.

16 The Development of a Curriculum-Based C-Test for Young EFL Learners Raphaela Porsch and Eva Wilden

Introduction This study is set in the German context and relates to a significant change in the German educational system over the past 15 years: schools are legally bound to individually foster each pupil. A significant element of this policy is the continuous evaluation of learner achievements by teachers in order to gain knowledge about individual development and progress throughout a pupil’s school career. Teachers are to take into account individual differences among their pupils in their lesson planning. A decisive point of time for the assessment of individual pupils’ competencies is when teachers start teaching a new class as they need reliable information about their new pupils’ prior knowledge. In Germany, the majority of pupils are making the transition from primary to secondary school after year 4 (aged approximately 10 years old). Thus, teachers at the beginning of year 5 need a variety of practical methods and instruments to evaluate their new pupils. Addressing this need in the context of English as a foreign language (EFL) education, we explored the possibility of diagnosing the EFL competences of young learners after primary education with the help of a C-test developed for the target group within this project. The C-test is a variant of the Clozetest. In contrast to Cloze-tests, in C-tests only parts of words are deleted, not whole words. In all federal states of Germany, primary school children start learning a foreign language (mostly English) in year 3, in some states, including North-Rhine Westphalia (the context of this study), already in year 1 (for a detailed description of early foreign language education in Germany see Wilden & Porsch, 2015). Therefore, this particular C-test has been designed for the German context to evaluate pupils’ EFL competences at the beginning of secondary school in year 5. Aside from evaluating the 289

290

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

quality of the newly developed C-test for young EFL learners, the study to be presented also researches the usage of test-taking strategies in order to gain further information about the quality and appropriateness of the new test. After introducing prior empirical findings on C-tests, as well as on testtaking strategies, this chapter continues by sketching out the research design, in particular the development of a C-test for young EFL learners and the use of open questions within a questionnaire as a method of accessing young learners’ strategy use. Next, the findings are summarised to find tentative answers to the research questions. In doing so, there will be a particular focus on illustrating test-takers statements about their strategy use. The chapter concludes by discussing the findings in light of the general educational context of early EFL education and the applied method for gaining access to the pupils’ strategy use.

Theoretical Background C-tests aim at measuring general language ability (Asano, 2014) and have proven to be an effective instrument for assessing learners of different languages and in various age groups; however, they have not yet been adapted for young EFL learners. The construction of C-tests is based on the ‘reduced redundancy principle’ (Babaii & Ansary, 2001) and usually follows the deletion technique ‘rule of two’: ‘Beginning with the second word of the second sentence the second half of every second word is deleted’ (Grotjahn et al., 2002: 95). Ideally, C-tests consist of four to five short texts that are content neutral, authentic, in accordance to the target group, not fictitious and do not contain dialogues (Cronjäger et al., 2010). Each text consists of approximately 80 to 100 words with 20 to 25 gaps. In the test, the texts may be ordered in increasing difficulty. Along with the development of C-tests, studies have researched the use of test-taking strategies of testees in a C-test (e.g. Grotjahn & Stemmer, 2002; Salehi & Sanjareh, 2013). The usage of test-taking strategies has been interpreted from two perspectives: (1) as a compensation for a lack of knowledge or competence such as a lack of language proficiency; or (2) as a necessary precondition in times of frequent low- and high-stakes testing in schools. In this situation, pupils have to be familiar with various test formats in order to successfully display their achievements with regard to the constructs of interest, such as reading or listening comprehension. In this context, test-taking strategies can be a useful resource for testees. Cohen (2013: 893) makes a distinction between ‘strategies that might contribute construct-relevant variance to test results, namely language strategies and test-management strategies, and [test-wiseness] strategies that in contrast are likely to produce construct-irrelevant variance’. Language learner strategies are defined as ‘thoughts and actions, consciously chosen and

Development of Cur r iculum-Based C-Test for Young EFL Lear ners

291

operationalised by language learners, to assist them in carrying out a multiplicity of tasks from the very onset of learning to the most advanced levels of target-language performance’ (Cohen, 2011: 7). Examples are inferencing, paraphrasing or skimming to solve a reading comprehension task. Testmanagement strategies ‘are strategies for responding meaningfully to test items and tasks. … They include logistic issues such as weighing the importance of responding to different items or tasks, keeping track of time and determining where to look for answers’ (Cohen, 2013: 895). The third category of strategies applied in tests is test-wiseness strategies, which cover the use of knowledge of testing formats in order to perform well. For example, a testtaker chooses an answer because it accords with the wording of the passage. A difficulty for detecting the use of test-wiseness strategies in the context of language testing is that correct answers may not be attributed (completely) to language proficiency or, in other words, to the construct tested. For research purposes Cohen (2013: 900) points out: ‘In an effort to ensure the validity of their tests, assessment specialists would benefit from knowing just which strategies are likely to result in the correct answer.’ Knowledge about strategies that are usually applied by test-takers helps to ensure that tests are testing the constructs they claim to assess. There are only a few studies that have researched strategy use when solving C-tests, predominantly applying the think-aloud technique (e.g. Babaii & Moghaddam, 2006; Cohen et al., 1985; Fahim & Tabataba’ian, 2012; Grotjahn & Stemmer, 2002; Klein-Braley, 2002; Salehi & Sanjareh, 2013; Stemmer, 1991). Cohen (2011: 317) summarises the findings of research on C-test-taking strategies as follows: It has been found that response strategies predominantly involve microlevel processing. Since half the word is given, pupils who do not understand the macro-context have been observed to mobilize their vocabulary skills in order to fill in the appropriate discourse connector without indulging in higher level processing. This seems to be especially true for language learners at a beginning and an intermediate level of proficiency. However, findings have also shown that in test-taking strategy research not only test-takers proficiency ought to be considered, but also text features (such as the complexity of sentences causing C-tests to be more difficult), which in consequence are more likely to cause the employment of macro-level strategies.

Purpose of the Study and Research Questions The research design of this study on C-test design for young EFL learners has been based on the following principles: (1) meeting C-test criteria as

292

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

sketched out above and following the reduced redundancy principle; (2) taking into account the target group of young EFL learners, especially by referring to the relevant curricula with regard to familiar topics, vocabulary, grammar, reading and writing competencies; (3) developing an instrument that allows reliable and valid testing of EFL proficiency; and (4) including the children’s perspectives in the test development. Considering these criteria, the following research questions are addressed: I.

II.

Test properties: (a): Does the C-test show appropriate levels of item difficulty and item discrimination as well as a satisfactory reliability? (b) Is there at least a medium strength correlation between the C-test scores and the school grades in English providing a first indicator of the test’s construct validity? Testing strategies: (a) Are year 4 pupils able to state in written questionnaires which test-taking strategies they employ? (b) According to children’s statements which test-taking strategies do they apply when solving the C-test?

Methods Participants In 2014, data were collected at four primary schools (12 classes) from 201 year 4 pupils after three and a half years of EFL education in the German federal state of North-Rhine Westphalia. The pupils were aged between 9 and 11 years old, with the majority being 10 years old (71.7%), their mean age was 9.93 years (SD = 0.53). 49% of the children were male, 55.1% of them stated that German was the only language spoken at their homes, while 43.8% stated they spoke German and another language at home and 1.1% stated they did not speak German at home at all. Germany uses a six point-grading scale starting with ‘1’ as the best possible grade given for outstanding performance. The distribution of the grades for English (EFL) in the last reports in grade 4 reported by the children is as follows: ‘1’ (‘very good’); 13.7%, ‘2’ (‘good’); 50.5%, ‘3’ (‘satisfactory’); 28.9%, ‘4’ (‘sufficient’); and 6.8%, grades ‘5’ and ‘6’ (‘deficient’) were not given. The participation in this study was voluntary; permission by the parents had been obtained.

Measures After an introduction by the test administrator on how to solve C-tests, the four C-test texts were presented to the pupils (see Appendix). There was a time limit of five minutes for each of the four texts. The test booklets were handed out in two versions with texts placed in a different order in each of the versions; the participants were randomly assigned to these versions.

Development of Cur r iculum-Based C-Test for Young EFL Lear ners

293

Immediately after responding to the C-test, the children were requested to answer post-test written questions about: (1) the level of difficulty of the test; (2) their strategy use when solving the C-test; and (3) their personal backgrounds (gender, age, socioeconomic background, family languages and school grades in German and English). Altogether, the test and questionnaire took about 45 minutes to complete.

The C-test for young EFL learners Although Babaii and Moghaddam (2006) found that C-tests are easier when the number of missing letters is indicated, we decided to develop a ‘traditional’ version of the C-test. Since EFL learners at primary schools in North-Rhine Westphalia are not obliged to learn the spelling of English words from memory (MSW, 2008), we assume that indicating the number of missing letters would be a demotivating factor and as a consequence make the C-test more difficult. Furthermore, this approach might stimulate testtakers to use ‘letter counting’, which would defy the purpose of the test (Meißner-Stiffel & Raatz, 1996). The curriculum for English in North-Rhine Westphalia (MSW, 2008) states that children at the end of primary school ought to have achieved level A1 of the Common European Framework (CEFR), whereby their listening comprehension and speaking skills might be above that level. Since reading proficiency is essential for solving C-tests, these curricular requirements with regard to reading competences to be achieved at the end of class 4 in particular informed our development of the C-test: ‘When reading, students understand words and sentences and retrieve relevant information from simple, short texts with familiar vocabulary’ (MSW, 2008: 78, translated by authors). In addition, the curriculum states that the learners understand short (authentic) texts with familiar vocabulary. Since C-tests ought to consist of authentic texts, we explored standard EFL textbooks, children’s books, story cards and children’s stories published in various EFL journals. In addition to the text type, we also needed to find texts with vocabulary familiar to the pupils. The curriculum of North-Rhine Westphalia where the study took place refers to five different ‘fields of experience’ with sub-topics (MSW, 2008: 76). We aimed at covering at least three of the fields with our four texts. A further issue when assessing language proficiency with C-tests is which scoring method out of the two options to apply: (1) dichotomous coding that is widely used for analysing C-tests (see Eckes & Grotjahn, 2006); or (2) a differentiated evaluation leading to polytomous items (Cronjäger et al., 2010). The second method takes the idea of communicative language teaching into account, an approach that ‘emphasizes learning a language first and foremost for the purpose of communicating with others’ (Duff, 2014: 15). In this study, we followed the latter approach and answers were coded as follows: ‘0’ for answers that were incorrect regarding content; ‘1’ for correct answers regarding content, but including spelling errors; and

294

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

‘2’ for answers that were correct regarding content and spelling. The codes ‘1’ and ‘2’ were merged into one unique code. The piloting of the C-test (Ehmke, 2014) had shown that the ‘rule of two’ could not be adhered to, since a pre-trial with four children showed that it impeded understanding of the texts at that particular competence level. As an alternative, beginning with the second sentence, we deleted parts of every third word, with some exceptions, in order to facilitate understanding. Four texts each with 20 gaps were thus developed, amounting in total to 80 words with deleted letters. This version of the C-test and the scoring method was piloted with 144 year 4 pupils in 2013 (Ehmke, 2014). Based on these results, items that proved too difficult (mean of 0.05 or less) or too easy (mean of 0.95 or higher) for the target group were changed for the main study, by providing an additional letter or – in a few cases – changing the word.

Employment of strategies Immediately after having completed the C-test, the children were asked in a questionnaire about: (1) which gaps they perceived as difficult; and (2) which strategies they employed when filling the gaps. This was done by presenting a short passage from each of the four texts comprising three sentences along with the following questions (originally in German): Which gaps were difficult? Why? How did you find the solution? Below is an example of the stimulus used for asking children about perceived difficulties and strategy usage: It has got fluffy white fur. My rabbit is two years old. Every day I feed her carrots. Altogether, answers about the employment of strategies from 102 children were available (some children did not write an answer to every question); a total of 114 answers (from 82 children) could be used for analysis (some answers were illegible or unclear for the purpose of analysis). In qualitative methodology, coding as the process of analysing, naming and classifying data – in this context answers about strategy use – can be developed bottom-up (from the material) and/or top-down (from existing theories and/or prior research) (e.g. Schreier, 2012). In the context of this study, the pupils’ written answers were coded as either being a language learning (reading) or a test-management strategy; the testees did not state any test-wiseness strategies. In a further step, each answer was classified more precisely using a coding system. The codes were developed both topdown by referring to previous studies and bottom-up by inducing them from the data if the strategies named by the pupils had not been identified in prior research. The coding was completed by the authors using a consensual process.

Development of Cur r iculum-Based C-Test for Young EFL Lear ners

295

Results Research question Ia: Reliability, difficulty and discrimination Table 16.1 provides descriptive statistics for the four texts based on the answers by 201 pupils in the main study. The p-value indicates the medium difficulty of each item (here, each text), thus, the higher the value, the easier the item. As expected for this group, all texts were moderately difficult and within a range from 0.38 to 0.47, which indicates that all items could be used for further testing. The item discrimination (rit) is a value for the relationship between how well testees performed on the item and their total test score ranging from 0 to 1. The higher the value, the more discriminating the item is. An item that is highly discriminating indicates that test-takers with high scores in the test responded correctly, whereas pupils with low scores responded incorrectly. Discrimination values of 0.50 or higher (here: rit ≥ 0.71) mean that items can be regarded as discriminating well (Fisseni, 2003). The reliability – we used Cronbach’s alpha as an estimate of the internal consistency – of the test is sufficient as well (α ≥ 0.80; Fisseni, 2003). Table 16.1 Descriptive data, mean (of sum score), difficulty (p), discrimination (r it) and reliability (α) Title of text

Min

Max

M (sum)

SD

p

rit

α

Rats Party Friend Lola Total test score

0 0 0 0 2

20 20 20 18 76

8.78 7.57 9.31 8.61 34.27

4.68 4.68 4.36 4.26 15.01

0.44 0.38 0.47 0.45 0.44

0.71 0.77 0.77 0.71 –

− − − − 0.88

Research question Ib: Correlations with grades in English Table 16.2 shows the correlations between the scores of the C-test for each text and the overall score. As expected, significant correlations with a range of –0.40 to –0.50 were found, which can be interpreted as a first indicator of construct validity with the developed C-test measuring English proficiency. Table 16.2 Correlations of C-test scores and school grades in English

School grades in English

Text 1

Text 2

Text 3

Text 4

C-test score

–0.40

–0.48

–0.41

–0.45

–0.50

Note: All correlations are significant (p < 0.001).

296

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Research question II: Testing strategies According to the questionnaire responses, pupils used 14 different strategies when solving the C-test: 10 reading strategies and four test-management strategies, along with answers that were classified as ‘unspecific knowledge activation’, which is not considered to be a strategy. Although the stimuli for answering referred to the individual texts, in almost all cases children gave answers that reflect their general usage of strategies in all four texts – or, in other words, they did not indicate that their strategy use was different for each text or that it varied for each of the gaps. As a consequence, the frequency and proportion of applied strategies refers to the overall usage. Table 16.3 presents the four test-management strategies with examples and additional information about the frequently mentioned approach that we call ‘unspecific knowledge activation’, but which we do not consider being a strategy. As Table 16.3 shows ‘Guessing’ (S1) was the test-management strategy pupils most often named as having used during the C-test. This can be seen as a classical test-management strategy, which pupils in all probability employed because they know that in terms of test performance writing some answer promises more success than writing no answer. In contrast to ‘Unspecific knowledge activation (no strategy)’ we consider ‘Guessing’ a testmanagement strategy as the latter involves looking for clues in the text to find a possible solution. Similarly, the other test-management strategies identified seem to have been employed by the learners to cope with the logistic side of test-taking. They include, for example, answering easy items first and only then tackling the difficult ones to save time (S2) or reading the entire text before answering items to ensure understanding of the overall context (S4). Table 16.4 gives an overview of all 14 language learner strategies pupils applied according to their questionnaire answers. As expected, the language learner strategies identified include both micro and macro strategies. As micro strategies, for example, pupils stated to have translated words into their first language (S5) or to have tried out a number of possible words (S7) to find a solution. These can be classified as micro strategies as pupils mobilised their vocabulary skills and looked for a solution in the immediate context of the item. By employing macro strategies, on the other hand, they processed the text on a higher level. For example, they considered the overall context and topic of a text (S11) or they tapped other knowledge sources such as background and world knowledge (S13). In some instances, however, the distinction between micro and macro strategies was not clear-cut (S12) when pupils stated to have used the text preceding and following a gap to find the solution. As this might be a word or two in one case but two or three sentences in another, this category was coded as containing both micro and macro strategies. Despite of the occurrence of both micro- and macro-level strategies, it is significant that the young EFL learners tested in this study appear to have

Easiest-first approach

Following the item order

First reading-second answering

S2

S3

S4

The test-taker completed the items in the order of the text. First, the test-taker read the whole text. Second, she/he answered the items. The test-taker expresses that she/he thought about the solution without specifying any further actions

The test-taker explicitly states that she/he made a guess. The test-taker completed those items first that she/he considers as being the easiest.

Description

I read everything first and then I filled it in. I found the correct answers by thinking about them.

Ich habe erst alles gelesen und dann ausgefüllt. (# 4) Ich habe die Lösungen gefunden, indem ich nachgedacht habe. (# 45)

19

4

NA

3.6%

0.9%

5.4%

6

I simply filled in the easy bits and then I thought longer about the more difficult ones; but I thought for a long time. I filled in gap after gap.

Ich habe einfach das leichte ausgefüllt und dann habe ich beim schweren lange überlegt; aber ich habe lange überlegt! (# 1) Ich habe Lücke für Lücke ausgefüllt. (# 23) 1

12.6%

Proportion

14

Frequency

I mostly guessed.

Example (English translation)

Am meisten habe ich geraten. (# 150)

Example (German original)

Note: Proportion, percentage of all strategies mentioned by the C-test-takers. NA, not applicable.

Unspecific knowledge activation (no strategy)

Guessing

S1

Strategy

Table 16.3 Test-management strategies employed in the C-test

Development of Cur r iculum-Based C-Test for Young EFL Lear ners 297

Choosing the most suitable option out of some choices (micro)

In-depth-reading (micro)

S7

S8

S6

Translation into L1 (micro) German equivalent (micro)

S5

Strategy

The test-taker states that she/he read the text very thoroughly.

The test-taker states that she/ he tried out a number of possible words and made a choice by using her/his knowledge of words. She/he considered the item beginning and added letters.

The test-taker translated words/ word groups/the text into the L1. The test-taker was able to retrieve the word because she/ he recognised that there is a similarity between the English and German word.

Description

Ich habe die Lösung gefunden, weil ich mich im Kopf gefragt hatte was mit ‘F’ anfängt. (# 67) Indem ich mir die Sätze gut durch gelesen habe. (# 55)

Ich bin auf die Wörter gekommen, wenn ich die einzelnen Wörter ausprobiert habe. (# 63)

‘film’ kann man sich leicht merken weil es im Deutschen genauso geschrieben wird. (# 183)

Ich habe versucht mir den Text auf Deutsch zu übersetzen. (# 82) Das Englische hört sich fast genauso wie im Deutschen und wir haben das mit unserer Englischlehrerin besprochen. (# 88)

Example (German original)

Table 16.4 Language learner strategies employed in the C-test

I found the answer because I asked myself which word starts with ‘F’. By reading the sentences carefully.

‘film’ is easy to remember because it is spelled the same in German. I figured out the words when I tried out several words.

I tried to translate the text into German for myself. The English sounds almost like German and we also talked about that with our English teacher.

Example (English translation)

3

27

3

5

Frequency

2.7%

24.4%

2.7%

4.5%

Proportion

298 Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Using general meaning/word field (macro)

Using the co-text (micro/macro)

Using background knowledge (macro)

Rereading (macro)

S12

S13

S14

Retrieving the word by referring to the same lexical item given before (micro) Juxtaposition (macro)

S11

S10

S9

The test-taker states that she/he reread the text.

The test-taker has background knowledge about the topic that helped her/him to retrieve the word(s).

The test-taker made relations between words, clauses or sentences preceding or following the gap.

The test-taker identified the co-occurrence of familiar combinations of words. The test-taker sought to restore the mutilated word by choosing a matching word from the general meaning of the text or the word field.

The test-taker realised that the word was used before (referencemaking).

It was also helpful to read the text again.

I found it easy because I have got a rabbit myself.

Ich fand es leicht weil ich selber einen Hasen habe. (# 80) Hilfreich war auch den Text noch einmal durchzulesen. (# 50).

Because I know it as an English series.

Yes, I figured it out because of the words before.

Ja rausbekommen habe ich es wegen den Vorwörtern. (# 32) Weil ich das als englische Serie kenne. (# 28)

It helped me to read the things before and after!

I found My rabbit is two years old as a whole text and I found it really easy. I found the answer because this text is about vegetables only, so only the word ‘carrot’ fits.

With ‘feed’, I checked in the text/gaps already given.

Mir hat geholfen das davor und danach zu lesen! (# 167)

Ich habe My rabbit is two years old als ganzen Text gefunden und den fand ich total einfach. (# 17) Ich bin auf die Lösung gekommen, weil es in diesem Text nur um Gemüse geht, also ging nur das Wort ‘carrot’. (# 165)

Bei feed habe ich in den schon geschriebenen Sachen geguckt. (# 42)

4

3.6%

3.6%

13.5%

15

4

5.4%

1.8%

1.8%

5

2

2

Development of Cur r iculum-Based C-Test for Young EFL Lear ners 299

300

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

used micro-level processing more frequently. Moreover, the children stated that they had applied a number of test-management strategies, in spite of indicating that they were unfamiliar with the C-test in English. In the light of these results, the question arises whether young EFL learners actually benefit from employing test-taking strategies when completing a C-test in English – or in other words: ‘Do young test-takers score higher in a C-test English when they explicitly state having applied reading or test-management strategies?’ In this sense, the scores for the overall test were compared between those children who did explicitly state using strategies to retrieve words in the C-test – and thus very likely did use them – and those who did not. The findings indicate a relationship between the use of strategies and the test results; those children not stating any strategy (n = 119) achieved a mean score of 0.41 (SD = 0.18) and those who did state at least one strategy (n = 82) achieved a mean score of 0.47 (SD = 0.19); the difference is significant (t(167.59) = –2.225, p < 0.05).

Discussion This study aimed at testing the quality and validity of a newly developed C-test for young EFL learners at the end of primary school, as well as identifying children’s C-test-taking strategies by using written questionnaires. The findings presented in this chapter can be interpreted as a first indicator that the newly developed instrument allows a reliable and valid measurement of English proficiency (research question Ia/b). This conclusion can be drawn from the evaluation of various indices based on classical test theory; however, evidence from analyses using item response models ought to be generated in addition (see Cronjäger et al., 2010). For further validation, future studies should allow a comparison between different proficiency scores based on the C-test and other instruments assessing English skills like reading comprehension. Furthermore, this study attempted to access young EFL learners’ testtaking strategies by means of written questionnaires (research question IIa/b). The evidence demonstrated a wide range of strategies being used according to pupils’ written answers that are in accordance with previous work. As to the nature of test-taking strategies applied, pupils stated to have employed both micro- and macro-level reading strategies as well as test-management strategies. In particular, the latter is noteworthy as children stated this particular test type to be unknown to them previously. Most interestingly, additional analysis showed that strategic test-takers were more successful test-takers – supporting the idea of ‘the good language learner’ (e.g. Chamot, 2005; Cohen, 2006; Finkbeiner, 2005; Griffiths, 2008) and the significance of teaching strategic skills in the foreign language classroom.

Development of Cur r iculum-Based C-Test for Young EFL Lear ners

301

Conclusion Strategy research often uses think-aloud data or verbal report methods, which can hardly be applied in studies with 200 participants or more; however, the analysis of pupils’ written post-test questionnaire answers in this study provided considerable insights into young EFL learner’s use of various strategies. Moreover, it corresponded with findings from previous studies adopting verbal reports or think-aloud data. In particular, think-aloud protocols have been the primary tools for test-taking strategy researchers since the early 1980s. Although they have undergone some changes (see Cohen, 2012: 98–99), there are some limitations in collecting data from large groups. Since more than half of the children in this study responded meaningfully to the open-answer questions giving interesting insights into their strategyuse, we argue that for this particular purpose – questioning children in studies with 200 or more participants – the coding of written responses provides a possible alternative. Nevertheless, the authors aim at ensuring that all children write answers to all questions in a future study (see below). Furthermore, it is planned to validate the questionnaire approach through qualitative interviews with test-takers in a further study. Although the majority of the children tested responded meaningfully to the questions about their perceived difficulties while filling the gaps in the C-test, the answers were either very general (they said the texts were easy/medium/ hard) or stating single words. However, since we only presented parts of the texts with the solutions, a systematic analysis was not possible. Thus, in order to answer the question which gaps are perceived as rather easy and which ones as difficult by young EFL learners, a follow-up study testing and questioning pupils in year 5 – immediately after the transition from primary school – is planned. The children will complete the same C-test and subsequently will be asked about difficulties with the gaps and their strategy application in an open and closed format. Contrary to the approach described in this chapter, not only parts of the texts from the C-test will be presented again, but in the planned study the complete texts will be shown again. From this we hope to gain further information about the specific challenges of individual gaps in addition to the overall test results. In addition, we expect that the repeated questioning about strategy use with a larger group allows a differentiated analysis of strategy use since more recent works suggest that ‘learning success cannot be attributed to different frequencies of strategy use’ but it is rather the ‘quality of strategy use’ that is essential (Ludwig et al., 2013: 383).

Acknowledgements The authors are indebted to Janina Ehmke, former student at the RuhrUniversity of Bochum, Germany and Esther Sim, student at the University of Bielefeld, Germany, for their support in the data collection.

302

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Appendix Text 1: Linda likes rats Adapted from: Doms, C. (2009) ‘Why do you like rats?’ Grundschule Englisch 28, 22–25. Hi, I’m Linda and my favourite pet is a rat. Why? I li rats because the can run ve fast. I c hug a r and I can pl with a rat i the ho and in t garden. The on problem is, m friends hate r ! That’s why I ha got a rab . Her name i Gillian. It has g fluffy white f . My rabbit i two years o . Every day I fe her carrots. Carrots are her favourite food!

Text 2: Grandma’s birthday party Adapted from: Porsch, R. and Mellor-Schwartze, V. (2011) New clothes for grandma’s birthday. Eine Bildergeschichte mit Überraschungseffekt. Grundschule Englisch 37, 22–25. It’s Grandma’s birthday party on Saturday. Karen, Jake and Mum are go to the sh to buy n clothes. Jake is n happy. He wa to play foot . Karen is not ha . She wants t play w ith h friends. Mum i happy. She li shopping. They g to a b store. Mum picks a sh and trousers f Jake. Then the go to t changing rooms. Jake tries o the clo . He is sa . He says ‘I don’t like it. The shirt is too big and I look like Grandpa!’

Text 3: Diana’s best friend Adapted from: Waas, L. (2012) Diktate und Übersetzungen. Klasse 5. München: Hauschka. Diana’s friend is in her class. Her name i Amy and she li in Diana’s street. Diana and Amy a eleven years o . Their birthdays a in January, b on different da . Amy has, like Diana, gr eyes, but br hair. Diana i 139 cm and Amy is 141 cm ta . They both we glasses and th favourite clothes a jeans and pull . Sometimes they we earrings. They li going to t cinema together. Th favourite fi is ‘Harry Potter’. Ever ybody thinks they are sisters.

Development of Cur r iculum-Based C-Test for Young EFL Lear ners

303

Text 4: Lola does not like tomatoes Adapted from: Child, L. (2003) I Will Never Not Ever Eat a Tomato. Cambridge, MA: Candlewick. I have a little sister. Her name i Lola. She is sm fu . Sometimes I ha to give h dinner, but s not e everything. She do not like car car are for rab . I say ‘Wh do you li says ‘I don’t li potatoes or car or peas o or spag or eggs o sausages. I don’t e ban or oranges. And I never ever eat tomatoes!’ My tomatoes.

and very does . She says ?’ Lola mushrooms cauliflower, sister hates

References Asano, Y. (2014) C-Tests und ‘allgemeine Sprachkompetenz’: Theoretische Überlegungen und empirische Analysen. In R. Grotjahn (ed.) Der C-Test: Aktuelle Tendenzen/The C-Test: Current Trends (pp. 39–52). Frankfurt/Main: Peter Lang. Babaii, E. and Ansary, H. (2001) The C-test: A valid operationalization of reduced redundancy principle? System 29, 209–219. Babaii, E. and Moghaddam, M.J. (2006) On the interplay between test task difficulty and macro-level processing in the C-test. System 34, 586–600. Chamot, A.U. (2005) Language learning strategy instruction: Current issues and research. Annual Review of Applied Linguistics 25, 112–130. Cohen, A.D. (2006) The coming of age of research on test-taking strategies. Language Assessment Quarterly 3 (4), 307–331. Cohen A.D. (2011) Strategies in Learning and Using a Second Language. Harlow: Longman Applied Linguistics/Pearson Education. Cohen, A.D. (2012) Test-taking strategies. In C. Coombe, P. Davidson, B. O’Sullivan and S. Stoynoff (eds) The Cambridge Guide to Second Language Assessment (pp. 96–104). Cambridge: Cambridge University Press. Cohen, A.D. (2013) Using test-wiseness strategy research in task development. In A.J. Kunnan (ed.) The Companion to Language Assessment (pp. 893–905). Boston, MA: Wiley. Cohen, A.D., Segal, M. and Weiss Bar-Siman-Tov, R. (1985) The C-test in Hebrew. In C. Klein-Braley and U. Raatz (eds) C-Tests in der Praxis (pp. 121–127). AKS-Rundbrief Special Issue 13/14. Bochum, Germany: Ruhr-Universität Bochum. Cronjäger, H., Klapheck, K., Krätzschmar, M. and Walter, O. (2010) Entwicklung eines C-Tests für Lernanfänger der Sek. I mit Methoden der klassischen und probabilistischen Testtheorie. In R. Grotjahn (ed.) Der C-Test: Beiträge aus der aktuellen Forschung [The C-Test: Contributions from Current Research] (pp. 71–100). Frankfurt/Main: Peter Lang. Duff, P.A. (2014) Communicative language teaching. In M. Celce-Murcia, D.M. Brinton and M.A. Snow (eds) Teaching English as a Second or Foreign Language (pp. 15–31). Boston: National Geographic Learning. Eckes, T. and Grotjahn, R. (2006) A closer look at the construct validity of C-tests. Language Testing 23 (3), 290–325. Ehmke, J. (2014) Englischkompetenzen am Übergang von der Primar- zur Sekundarstufe: Eine Studie zur Erprobung eines C-Tests am Ende der Klasse 4 [English Competencies at the Transition from Primary to Secondary Level. A Study Testing a C-Test at the End of Class 4]. Ruhr-Universität Bochum, Germany: Master Thesis.

304

Par t 4: Evaluat ing Early L anguage Lear ning Programmes

Fahim, M. and Sadat Tabataba’ian, M. (2012) Concept maps, Cloze tests, and multiplechoice tests: A think-aloud approach to the comparison of the strategies utilized in different test formats. The Journal of American Science 8 (8), 131–138. Finkbeiner, C. (2005) Interessen und Strategien beim fremdsprachlichen Lesen: Wie Schülerinnen und Schüler englische Texte lesen und verstehen [Interests of and Strategies on Reading in a Foreign Language: How Students Read and Understand English Texts]. Tübingen: Narr. Fisseni, H.J. (2003) Lehrbuch der psychologischen Diagnostik: Mit Hinweisen zur Intervention [Textbook of Psychological Diagnostics: With Information on Intervention]. Göttingen: Hogrefe. Griffiths, C. (ed.) (2008) Lessons from Good Language Learners. Cambridge: Cambridge University Press. Grotjahn, R. and Stemmer, B. (2002) C-Tests and language processing. In J.A. Coleman, R. Grotjahn and U. Raatz (eds) University Language Testing and the C-Test (pp. 115– 129). Bochum: AKS. Grotjahn, R., Klein-Braley, C. and Raatz, U. (2002) C-Tests: An overview. In J.A. Coleman, R. Grotjahn and U. Raatz (eds) University Language Testing and the C-test (pp. 93–114). Bochum: AKS. Klein-Braley, C. (2002) Psycholinguistics of C-test taking. In J.A. Coleman, R. Grotjahn and U. Raatz (eds) University Language Testing and the C-test (pp. 131–142). Bochum: AKS. Ludwig, P., Finkbeiner, C. and Knierim, M. (2013) Effects of the adequacy of learning strategies in self-regulated learning settings. A video-based micro-analytical lab study. Journal of Cognitive Education and Psychology 12 (3), 374–390. Meißner-Stiffel, M. and Raatz, U. (1996) ‘ oder _ _ _ _?’ Zwei Grundlagenuntersuchungen zum C-Prinzip bei L1-Lernern. In R. Grotjahn (ed.) Der C-Test: Theoretische Grundlagen und praktische Anwendungen [The C-Test: Theoretical Foundations and Practical Applications] (pp. 173–183). Bochum: Brockmeyer. MSW (Ministerium für Schule und Wissenschaft Nordrhein-Westfalen) (2008) Richtlinien und Lehrpläne für die Grundschule in Nordrhein-Westfalen [Guidelines and Curricula for Primary School in North Rhine-Westphalia]. Düsseldorf: MSW NRW. Salehi, M. and Sanjareh, H. (2013) On the comparability of C-test and Cloze: A verbal protocol approach. English for Specific Purposes World 14 (39), 1–18. Schreier, M. (2012) Qualitative Content Analysis in Practice. London: SAGE. Stemmer, B. (1991) What’s on a C-test taker’s Mind? Mental Processes in C-test Taking. Manuskripte zur Sprachlehrforschung 36. Bochum: Brockmeyer. Wilden, E. and Porsch, R. (2015) Learning EFL from year 1 or year 3? A comparative study on children’s EFL listening and reading comprehension at the end of primary education. In M. Nikolov (ed.) Assessing Young Learners of English: Global and Local Perspectives (pp. 191–212). New York: Springer.

17 Mixed Methods in Early Language Learning Research Janet Enever and Eva Lindgren

Introduction In this concluding chapter we critically examine our understandings of mixed methods research (MMR), particularly in relation to the field of early language learning (ELL). We take as our starting point the selection of research studies included in this volume and consider the paradigmatic nature of the field represented by this collection. We should state at the outset that this collection is by no means intended to provide a comprehensive overview of mixed methods research available in the field of ELL; rather, the collection may be regarded as an initial attempt to gather together a sample of recent studies that have adopted a mixed methods approach, in some or many aspects of their research design. It is perhaps important to note also that contributors to this volume were not requested to engage specifically with mixed methods, rather, they were invited to submit a chapter that included a mixed methods approach in some or all aspects of the study design. In the following sections we briefly continue the discussion of some relevant design aspects of MMR begun in the introductory chapter, considering their significance for research in ELL. We discuss the particular contribution that chapters in this volume make to consolidating our understandings of the relevance of MMR for ELL research. Finally, we will outline some possible future directions for ELL research, drawing on the particular merits that an MMR framework may offer.

Mixed Methods in Early Language Learning Research To the best of our knowledge, this is the first publication in the field of ELL that focuses entirely on the use of MMR in ELL. We can identify no publication attempting to provide a ‘state of the art’ overview of MMR in ELL, although a number do provide a partial perspective on this. Indeed, in the much broader field of applied linguistics in general, it is only quite 305

306

Early L anguage Lear ning

recently that such a publication has emerged (see Riazi & Candlin, 2014). We also note the contribution of a recent publication by Zha and Tu (2015) reviewing the growth of MMR in comparative education research over the past 20 years, which they attribute to a gradual paradigm shift. Given the relatively recent establishment of MMR as an acknowledged approach to research investigation (Creswell & Plano Clark, 2011) and the very recent growth of research in the field of ELL, first charted in Europe by Blondin et al. (1998) and later more extensively by Edelenbos et al. (2006), it is perhaps unsurprising that the theme of MMR has not yet received extensive coverage in ELL research. It should also be noted that currently there exists no internationally published journal devoted entirely to research in ELL, a repository which might arguably have provided a natural home for the discussion of research methodologies. Before engaging in a discussion of what the chapters in this volume can tell us about the potential contribution of a mixed methods approach to research in the field of ELL, we will first provide a brief discussion of some recently published studies in ELL that explicitly claim to have adopted an MMR approach. We do not intend this as an exhaustive summary of current publications, rather as an indicative reflection on the state-of-the-art. Notably, a study by McWayne et al. (2013) is one of very few related to early language learning where MMR is explicitly mentioned in the title, suggesting that the inclusion of this research framework was a central element of the study. Significantly, the paper refers to the recent employment of MMR in school psychology studies as offering ‘an exciting approach for new scale development that is sorely needed’ (McWayne, 2013: 594). This study of Latino children growing up bilingually in the United States explores how families support their children’s educational experiences using an emic approach, which assumes that, ‘beliefs, values, and behaviours, are situated in, and reflective of, the particular cultural and psychosocial realities of individuals and groups’ (McWayne, 2013: 594). As a first stage in the study, qualitative data were collected through semi-structured conversations with groups of parents. From this, a self-report scale was developed and used to formulate a questionnaire. Finally, the questionnaire was used to gather large-scale data on parents’ engagement with their pre-school children’s development that could be construed to be in support of educational development in specifically culturally-defined ways. The analysis revealed concerns related both to their children’s socioemotional development and to their cognitive-linguistic development. In a quite different context, Mihaljević Djigunović (2015) explored the nature of individual differences among young EFL learners in Croatia using a mixed methods approach. Here, the author refers to the complementary nature of combining a longitudinal case study with quantitative data, proposing that it can offer ‘deeper and richer insights into affective processes’ (Mihaljević Djigunović, 2015: 32). In seeking out evidence of individual

Mixed Methods in Early L anguage Lear ning Research

307

differences, the research team sought to explore their relationship to age and language proficiency, adopting a situated approach to investigate how learners’ attitudes, motivation and self-concept interacted with contextual variables both in and out of school. Data were collected over a three-year period from two samples – one who had begun learning English at six years of age and the second who started later at 10 years. Throughout the study, both quantitative and qualitative data were collected. The findings were first analysed quantitatively, then comprehensive individual profiles for six learners were constructed from the data, including detailed knowledge of their contextual backgrounds. In a justification of this approach the author notes that quantitative data may facilitate some generalisations, but is less able to offer insights of fluctuations in affective processes over time. This combination of methodological approaches offers insights to the ‘non-linear relationships between many factors’ (Mihaljević Djigunović, 2015: 31) that may occur in the processes of language learning. Copland et al. (2014), in their report on some aspects of a large-scale survey for teachers of English to young learners, engage in a quite detailed discussion of their MMR framework. Adopting the term ‘QUAN-qual design’ (Morse, 2010), they identify two main strands of data in their study (note: The capitalisation used by the authors is intended to highlight which aspect of the research design is more dominant). First, they refer to ‘the core component’, comprising the survey data and secondly, a smaller amount of qualitative data collected from interviews, observations and an open question contained in the survey. They describe this as ‘the supplementary approach’ (Copland et al., 2014: 742). Implied in this choice of terminology is the view that the large-scale survey itself reveals some valuable quantitative findings, but the addition of some smaller-scale qualitative data supplements and enriches the analysis. Finally in this section we include an MMR study of reading skills development in English reported by Jang et al. (2015). Their investigation drew on a sample of 44 children in grades 5 and 6 in Canada, of which 18 had exposure to languages other than English at home. This two-phase developmental study employed quantitative methods, first to determine relationships between variables for children’s learning characteristics, with subsequent qualitative evidence to describe the processes underlying these variables. For this sequential design, Jang et al. (2015) drew on Maxwell and Loomis (2003) to identify this approach as a developmental MMR design. With the sample of these four MMR studies it is already apparent that a variety of interpretations of MMR design lend themselves to ELL research. The four studies briefly outlined above include the following approaches: (1) Emic collection of qualitative data, coded to formulate a questionnaire, which was subsequently administered to a larger-scale sample, reflecting a sequential design pattern.

308

Early L anguage Lear ning

(2) Parallel collection of quantitative and qualitative data. Initial quantitative analysis facilitated complementary development of comprehensive case studies. (3) A sequential design including large-scale survey data, followed by supplementary small-scale interviews which contributed to enriching the study. (4) Sequential design including collection of quantitative data to determine relationship between variables, followed by qualitative preparation of learner profiles as a tool for understanding the impact of diagnostic feedback. These varied approaches to mixed methods design are amply reflected in the range of chapters included in this volume. In the next section we discuss these further, giving consideration to whether it is helpful to attempt some kind of categorisation, as proposed by Greene (2008), Creswell (2009) and many others, or whether an over-emphasis on systematic approaches might limit the opportunities for the formation of innovative new approaches.

The Contribution of this Collection The three chapters included in Part 1, Overviews of Research Findings, together describe a variety of methods. Chapters 2 and 3 are principally issues-related (Zha & Tu, 2015), concerned with the complex issues raised by many research studies in relation to, first, African contexts and second, cultural learning in ELL. Chapter 4 reflects on the divided nature of research in education, where researchers have historically held ‘pre-existing commitments to certain paradigms’ (Zha & Tu, 2015: 167). Reflected here are traditions both of quantitative (quasi-experimental) research and qualitative research in literacy development in the field of English as an additional language (EAL). Murphy notes here that while MMR might offer more comprehensive insights for EAL research, its potential is, as yet, substantially under-explored. Together, these three chapters provide a useful overview of the current state of knowledge in particular sub-fields of ELL, also reflecting areas of continuing debate regarding the nature of research design. As mentioned in the Introduction to this volume, the five chapters contained in Part 2 fully position themselves as MMR, including a variety of MMR patterns. The study included in Chapter 5 is unusual, in that it took the form of action research, whereby the teacher-as-researcher was positioned ideally to collect quantitative evidence on vocabulary learning and literacy development, with qualitative data collected from children on a regular basis. The MMR structure follows a sequential pattern where the children’s weekly reflections, informal classroom assessments, video and audio recordings illuminated the quantitative data and documented progression over time.

Mixed Methods in Early L anguage Lear ning Research

309

In contrast, the second study in this part (Chapter 6) was conducted by a team of five, working on a substantial data set produced from questionnaires and the novel approach of elicited metaphor analysis (EMA) to explore learner motivation. The scale of complexity in combining these two data sources for this study led the research team to describe their study as exploratory in nature. Here the MMR framework proved valuable in addressing the diverse elements of motivation, offering a new approach to revealing motivational factors for ELL. The third empirical paper in this part (Chapter 7) draws on data from a large-scale investigation on the potential for using multilingual electronic storybooks for children’s enjoyment and development of multilingual competency. Drawing on a sub-set of quantitative data from a larger study, two approaches to analysis are included. First, a quasi-experimental approach of administering a follow-up comprehension test to three sample groups and second, a qualitative analysis and categorisation of emotional responses from video material. Hilton’s chapter (Chapter 8) on individual L2 learning differences sets out to gather a substantial amount of quantitative data which can be examined by means of various statistical analysis programmes. In addition, parents’ and children’s elicited responses to a questionnaire are coded for statistical analysis – facilitating the inclusion of subjective viewpoints in a quantitative (numerical) analysis procedure and allowing for the subsequent construction of individual profiles for a small sample of the study participants. The final chapter in this part (Chapter 9) interrogates the effectiveness of a CLIL model for the teaching of three primary school subjects, employing a quasi-experimental framework with the inclusion of a control group. A battery of pre- and post-test qualitative measures were also taken to gauge the pupils’ attitudes to learning and to the administration of a final test. In this study, a mixed methods design is fully integrated at each stage of the procedure, enabling each procedure, in turn, to shed light on the other. These five chapters can be summarised as offering the following MMR frameworks: Chapter 5: Quantitative and qualitative data collected on a weekly sequential basis, facilitating both weekly small-scale analysis and subsequent, more substantial analysis including data sourced from parallel video recordings. Chapter 6: Quantitative questionnaire data, sequentially combined with qualitative analysis of elicited metaphors. Chapter 7: Quantitative data collection and analysis, followed sequentially by further qualitative analysis of smaller sub-sample. Chapter 8: A sequential design with qualitative analysis of video material complementing the initial collection of quantitative data. Chapter 9: Quantitative data combined with pre- and post-test qualitative data collection, facilitating an integrated analyses procedure.

310

Early L anguage Lear ning

The third part of the book contains four chapters adopting a longitudinal research framework. Noting the similarity of their approaches we will consider these four chapters as a whole, given that they all look at development over time. In each of the chapters, explanations for the authors’ choices of MMR vary, with Butler (Chapter 10) emphasising the importance of complementarity and triangulation in attempting multiple level analyses and Lopriore (Chapter 11) similarly reflecting this potential in her study. Lindgren and Enever (Chapter 12) discuss the advantages of parallel data collection for supporting analysis that can provide both general insights and shed light on the nature of more individualised pathways to learning. Mihaljević Djigunović (Chapter 13) notes particularly the challenges for investigating the complexity of contextual and individual learner factors longitudinally, finding it important to use mixed methods throughout her study to gain further insights. The similarity of research design is evident in all four of these chapters, leading us to suggest that longitudinal studies may find the mixed methods framework particularly advantageous. Together, these four chapters can be summarised as adopting the following mixed methods design: Chapters 10, 11, 12 and 13: Simultaneous collection of quantitative and qualitative data, with whole group analysis, focus group and/or individual case study analysis conducted at regular intervals. A final summative analysis is generally provided in both quantitative and qualitative formats, often including longitudinal case studies. In the fourth and final part of the book we include three chapters that focus on the evaluation of some of the major themes for ELL in instructed contexts today – interactive task design, educational outcomes and assessment test design. An outline of each MMR theme is given below, followed by summative statements of their MMR design. The part begins with Chapter 14, in which García Mayo and Imaz Agirre investigate conversation interaction in CLIL and non-CLIL environments to evaluate learner strategies in learner-learner interaction for two age groups. A longitudinal quantitative (quasi-experimental) research design including four learner groups was selected. Data collection for the four groups was based on a picture placement task, with subsequent transcriptions from video recordings of learner interactions during the task. The use of conversation adjustments was examined over time. Quantitative analysis was conducted, followed by supplementary small-scale qualitative analysis. Chapter 15 adopts a mixed methods design for a large-scale impact study of the introduction of primary English in one state in Mexico. In this, Sayer, Ban and López de Anda set out to assess students’ language progress and to evaluate the overall impact of the programme. While Sayer et al. identify interviews as their primary source, they include also a range of quantitative

Mixed Methods in Early L anguage Lear ning Research

311

measures in their data collection. Their account indicates that an integrated MMR design was adopted where quantitative data were used to inform qualitative findings and vice versa, thus serving to answer the initial research questions. The collection of MMR studies in this volume concludes with a chapter by Porsch and Wilden (Chapter 16) in which they evaluate the design of a language test developed for use in Germany, to be administered at the beginning of secondary school (grade 5). The MMR design included two strands: the administration of the test, followed by a post-test questionnaire on the test-taking strategies used by participants. Here, research design was mixed at the points of instrument design, data collection and software analysis, while the findings were interpreted as two separate strands for discussion. These final three chapters can be summarised as offering the following MMR frameworks: Chapter 14: Quantitative data collection and analysis, with initial qualitative analysis of smaller sub-sample. Chapter 15: Wide range of quantitative and qualitative data collection with qualitative perspectives substantially driving analysis procedures. Chapter 16: Sequential quantitative and qualitative data collection with statistical analysis, followed by presentation and discussion of findings in two separated strands.

Discussion and Future Directions for MMR in ELL Close readers of this chapter will no doubt have recognised our hesitancy in attempting any kind of categorisation of the variety of MMR frameworks adopted by the authors included in this volume. Indeed, we freely admit that we may well have failed to fully reflect the mixed nature of some studies in our efforts not to misrepresent them (for which we apologise sincerely to all authors). Our difficulties relate in part to the contested nature of the continuing debate around the question of categorisation itself. Since the emergence of MMR as an acknowledged research approach there have been many attempts to develop a system of categorisation that might help to legitimise it and provide what Guest (2013: 142) has described as creating ‘a common language’. According to Creswell and Plano Clark (2011) there have been at least 15 published attempts to provide typologies for MMR, evidence which surely indicates that we are a long way from reaching a shared view on this topic – if indeed this is our aim. Riazi and Candlin (2014) suggest that Greene et al.’s (1989) five purposes of MMR may help in identifying purposes. The listed purposes are: triangulation, complementarity, development, initiation and expansion. According to Onwuegbuzie and Collins (2007), a further refinement of this framework might be to also

312

Early L anguage Lear ning

consider whether the mixing of methods occurs concurrently or sequentially. However, Riazi and Candlin (2014: 145) go on to note that the absence of any link between these purposes and an underlying paradigm ‘makes conceptualising and differentiating between the proposed purposes for MMR studies difficult’. In conclusion to these concerns regarding categorisation we return to our comments made earlier in this chapter, restating our position that attempts to establish a systematic framework for MMR in ELL at this early stage seem likely to be counter-productive in that they may limit the exploration of innovative new approaches to design. Turning to other themes that have surfaced during the development of this volume we would like to raise two points for reflection and possible future exploration. First, we consider research expertise as a contributory factor to both the initial design and subsequent administration of any research study. In Chapter 15, Sayer et al. note that they viewed themselves as essentially qualitative researchers. This leads us to reflect on whether a complex, multi-layered research investigation might often benefit from the selection of a multidisciplinary research team, bringing with them substantial backgrounds in the contributory paradigms of quantitative and qualitative methodologies respectively. No doubt with this mix of traditions, many debates on design and implementation would ensue. With a shared commitment to the design of a comprehensive, multi-layered study however, this mix of expertise is highly likely to succeed in achieving valuable outcomes to the research. It may also contribute towards the recognition of what Zhu and Tu (2015: 167) have described as ‘a continuum of scientific enquiry’. Following this point, we suggest that a multidisciplinary team may be particularly relevant in the design of a longitudinal study, where gathering evidence on a number of levels at consecutive points in time may bring complex challenges in design which will require the shared expertise of a multidisciplinary team. Of the four longitudinal studies that are included in this volume we are aware that at least three did include multidisciplinary teams at some stages of their design and implementation. From our personal involvement in the initial team that developed the ELLiE study (Enever, 2011) we can confirm the benefits of a multidisciplinary team for that study. Related to this point, a recent concern in ELL research has been to understand the dynamic nature of the learning process for these young beginners. Charting this progress across a number of years enables the research team to evaluate progress at interim points throughout the study period. Such interim analysis (perhaps on a yearly basis) offers opportunities to take stock of the research design at regular intervals, adapting research instruments and/or including new ones, to provide more detailed and precise insights as to the nature of this complex process. As researchers, our initial reflections on the experience of conducting longitudinal studies with mixed methods designs lead us to propose that the establishment of an iterative procedure for systematic revisiting of the

Mixed Methods in Early L anguage Lear ning Research

313

quantitative data set may offer a framework for identifying the most useful pathways to pursue through qualitative analysis during the whole longitudinal timeframe of the study. However, this viewpoint needs further testing to establish the extent to which it may be applicable to ELL studies. The two questions raised here are clearly speculative and would both benefit from further investigation, as would the many other concerns related to the legitimisation of MMR as a valid research methodology. We hope that this chapter and all that is contained in this volume has contributed to opening up the field for further exploration of many themes in relation to ELL and to ways in which they may interrogate the opportunities afforded by an MMR design. The emerging picture of a multi-dimensional MMR in ELL today is a position that we wholeheartedly encourage. We would thus emphasise the importance of remaining open-minded to new theories and to welcoming new approaches for the study of ELL as the field continues to expand.

References Blondin, C., Candelier, M., Edelenbos, P., Johnstone, R., Kubanek-German, A. and Taeschner, T. (1998) Foreign Languages in Primary and Pre-School Education: Context and Outcomes. A Review of Recent Research within the European Union. London: CILT. Copland, F., Garton, S. and Burns, A. (2014) Challenges in teaching English to young learners: Global perspectives and local realities. TESOL Quarterly 48 (4), 738–762. Creswell, J.W. (2009) Research Design: Qualitative, Quantitative and Mixed Methods Approaches (3rd edn). Los Angeles: Sage Publications. Creswell, J.W. and Plano Clark, V.L. (2011) Designing and Conducting Mixed Methods Research. Thousand Oaks, CA: Sage. Edelenbos, P., Johnstone, R. and Kubanek, A. (2006) The Main Pedagogical Principles Underlying the Teaching of Languages to Very Young Learners, Languages for the Children of Europe, Published Research, Good Practice & Main Principles. Brussels: European Commission. Enever, J. (ed.) (2011) ELLiE. Early Language Learning in Europe. London: British Council. Greene, J.C. (2008) Is mixed methods social inquiry a distinctive methodology? Journal of Mixed Methods Research 2 (1), 7–22. Greene, J.C., Caracelli, V.J. and Graham, W.F. (1989) Toward a conceptual framework for mixed method evaluation designs. Educational Evaluation and Policy Analysis 11, 255–274. Guest, G. (2013) Describing mixed methods research: An alternative to typologies. Journal of Mixed Methods Research 7 (2), 141–151. Jang, E.E., Dunlop, M., Park, G. and van der Boom, E.H. (2015) How do young students with different profiles of reading skill mastery, perceived ability, and goal orientation respond to holistic diagnostic feedback? Language Testing 32 (3), 359–383. Maxwell, J. and Loomis, D. (2003) Mixed method design: An alternative approach. In A. Tashakkori and C. Teddlie (eds) Handbook of Mixed Methods in Social and Behavioral Research (pp. 241–271). Thousand Oaks, CA: SAGE Publications. McWayne, C.M., Melzi, G., Schick, A.R. and Kennedy, J.L. (2013) Defining family engagement among Latino Head Start parents: A mixed methods measurement development study. Early Childhood Research Quarterly 28 (3), 593–607. Mihaljević Djigunović, J. (2015) Individual differences among young EFL learners: Age- or proficiency-related? A look from the affective learner factors perspective. In J. Mihaljević Djigunović and M. Medved Krajnovic (eds) Early Learning and Teaching of English. New Dynamics of Primary English (pp. 10–36). Bristol: Multilingual Matters.

314

Early L anguage Lear ning

Morse, J.M. (2010) Simultaneous and sequential qualitative mixed method designs. Qualitative Inquiry 16, 483–491. Onwuegbuzie, A.J. and Collins, K.T. (2007) A typology of mixed methods sampling designs in social science research. The Qualitative Report 12 (2), 281–316. Riazi, A.M. and Candlin, C.N. (2014). Mixed-methods research in language teaching and learning: Opportunities, issues and challenges. Language Teaching 47 (2), 135–173. Zha, Q. and Tu, D. (2015). Doing mixed methods research in comparative education: Some reflections on the fit and a survey of the literature. Annual Review of Comparative and International Education 2015 28, 165–191.

Index

Achievement 4, 25, 33, 43ff, 74, 80, 96, 155, 169ff, 187ff, 201ff, 225, 227, 235, 289, 290 Affective 28, 128ff, 168ff, 184, 186, 203, 222, 224, 238, 239, 306, 307 Factors 131, 203, 238 Trajectories of affective development 239 Africa 3, 9, 10ff Attitudes: Learners’ (students) 94ff, 129, 152, 153, 188–189, 207, 212, 217, 266 Teacher 26

Early exit model 11ff Elementary school 251, 271, 284 Elicited imitation 69ff, 133, 134, 136 Elicited metaphor analysis 85, 90, 309 English as a foreign language (EFL) 4, 5, 49, 68, 105ff, 125, 130, 146ff, 187, 199, 201ff, 222ff, 249ff, 266, 270, 289ff, 299 Achievement 4, 74, 80, 86, 96, 156, 169, 170, 180ff, 187ff, 201ff, 222ff, 289, 290 Interaction with contextual factors 167, 171, 222, 234, 225, 236 Interaction with individual factors 170 Out-of-class use of 158, 159, 191, 193, 194, 197, 198, 224, 226, 229, 234, 237, 238 Experiment/experimental group 2, 13, 43, 57, 110, 117, 146ff, 282, 308ff Exposure to FL 19, 65, 70, 127, 148, 158, 159, 190ff, 204ff, 224ff, 249, 254, 255, 270, 275, 284

C-test 5, 289ff China 3, 4, 86ff, 167ff CLIL content and language integrated learning 3, 4, 5, 145ff, 249ff, 281, 286, 309, 310 Code-switching 111, 122ff Cognitive load 112, 160, 231 Complementarity 2, 172, 182 Complex/dynamic systems 2, 87, 128, 139, 145 Comprehension 4, 11, 12, 47ff, 68ff, 91, 102, 112ff, 150ff, 171, 187, 189, 197, 198, 204, 215, 261, 290ff, 309 Computer 81, 110, 157, 159, 208, 211, 225ff, Contact with native speakers 223 Context 1, 3, 5, 9ff, 33ff, 42ff, 65, 66, 81, 86ff, 108, 111, 118, 127ff, 146, 158, 159, 167ff, 186ff, 201ff, 222ff, 255, 261, 266, 269ff Control group 110, 113, 148ff, 158ff, 309 Croatia 3, 5, 169, 170, 203, 224ff, 306

Facebook 282, 283 France 3, 4, 31, 36, 127, 142 German/Germany 3–5, 31, 36, 109ff, 169, 289ff, 311 Imitation 69ff, 102, 132ff, Individual differences 30, 65, 66, 74, 80, 81, 83, 127ff, 222, 306 Individual factors 170 Instructional setting 18, 249, 253, 254, 259 Interaction 27ff, 55, 56, 87, 88, 109ff, 135, 147, 193, 195, 201ff, 236ff, 249ff, 310 Interview/interviewed students 3, 25ff, 55, 91, 131ff, 150, 169ff, 187ff, 203ff, 226ff, 274ff, 301, 307, 308, 310 Italy 3, 4, 36, 188, 189

Diagnosing, diagnostic 5, 289, 308 Digital 108, 111, 116, 218, 275

315

316

Early L anguage Lear ning

L1 13, 41ff, 65ff, 108, 109, 127ff, 170, 211, 229, 232, 250ff, 276, 296 L2 1, 3, 5, 13, 42ff, 67ff, 86ff, 108, 109, 127ff, 159, 177, 184, 190, 201ff, 269, 280, 281, 309 L3 103 Late exit model 14, 16 Linguistic outcomes 65, 269, 280, 281 Listening skills 159, 212 Literacy 4, 21, 22, 41ff, 65ff, 93, 109, 125, 126, 144, 161, 187, 199ff, 275, 287, 306 Longitudinal 4, 5, 25, 26, 33, 87, 133ff, 166, 167ff, 186ff, 201ff, 254, 255, 306, 310, 312 Majority language 10, 19, 42, 44, 45 Minority language 10, 19, 20, 41ff Mother tongue 1, 9ff, 125, 147 Motivation 3, 4, 5, 34, 85ff, 108, 112, 127ff, 137ff, 160, 167ff, 187ff, 222ff, 252, 263, 266, 307, 309 Multilingual 3, 4, 9ff, 31, 34, 108, 109, 111, 112, 134, 272, 273, 309 Multilingual education 9, 20 Negotiation 112, 117, 118, 121, 123, 250ff

Questionnaire 3, 4, 25ff, 85ff, 114, 131ff, 150ff, 186ff, 203ff, 226ff, 266, 290ff, 306ff Reading 3, 4, 11ff, 25, 37, 43ff, 66ff, 86, 109ff, 139ff, 149ff, 187ff, 202ff, 226, 229ff, 252, 275, 290ff, 307, 313 Receptive 50, 67ff, 108ff, 133, 136, 204ff Self-concept 4, 5, 169ff, 203, 217, 223ff, 307 self-determination theory 168 Self-perception 188, 191, 192, 198, 233, 234 Slovenia 3, 147ff Socioeconomic status (SES) 44, 65, 167, 170, 203, 226, 229, 281 Storybook 4, 108, 110, 135 Sweden 1, 3, 4, 202ff Test development 292 Test-taking strategies 290ff The 4Cs conceptual framework 146 Thick description 4, 201ff Triangulation 2, 172, 182, 310, 311 Turkish 109ff UK 3, 4, 12, 27ff, 45ff, 158

Oracy 4, 66, 71, 80, 202, 204, 207, 225 Perceptions 30ff, 45, 88, 90, 155, 157, 188ff Plurilingual 24, 27, 30, 83, 109, 115, 131, 135 Productive 67, 110, 111, 140, 141, 202, 210

Verbal working memory 65ff Video data 117 Writing 4, 25, 37, 46, 53ff, 68, 77ff, 150, 158ff, 187ff, 204ff, 231, 238, 275, 283, 292, 296