Die effektive Bibliothek: Roswitha Poll zum 65.Geburtstag [Num. figs. ed.] 9783110965209, 9783598116698

Citation preview

Die effektive Bibliothek

Die effektive Bibliothek Roswitha Poll zum 65. Geburtstag

Herausgegeben von Klaus Hilgemann und Peter te Boekhorst

K G · Saur München 2004

Illustrationen und Umschlaggestaltung: Erwin Lohr

Bibliografische Information Der Deutschen Bibliothek Die Deutsche Bibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.ddb.de abrufbar.

Θ Gedruckt auf säurefreiem Papier © 2004 by Κ. G. Saur Verlag GmbH, München Printed in Germany Alle Rechte vorbehalten. Das Werk ist in allen seinen Teilen urheberrechtlich geschützt. Jede Verwertung ist ohne Zustimmung des Verlages unzulässig. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung in und Verarbeitung durch elektronische Systeme. Satz: Dr. Peter te Boekhorst, Andreas Rafalski Druck/Bindung: Strauss Offsetdruck GmbH, Mörlenbach ISBN 3-598-11669-1

Inhalt

Grußwort des Rektors der Westfälischen Wilhelms-Universität Münster

7

Grußwort des Kanzlers der Westfälischen Wilhelms-Universität Münster

10

Rowena Cullen An Integrative Model for the Evaluation of Information Services

13

Peter Brophy The Quality of Libraries

30

Maurice B. Line An Agenda for Overdue Change

47

Geoffrey Ford and Suhair Abdinnour Art or Science? Two Approaches to Materials Fund Allocation in Academic Libraries

54

Sigrid Reinitzer und Roswitha Karpf Wissenschaftliche Bibliotheken im Kontakt zu Schulen und Schulbibliotheken

69

Alex C. Klugkist Die Bibliothek von Babel oder die digitale Informationsversorgung

81

Aase Lindahl Renewals in Library Statistics - Report of a Danish Discussion

89

Klaus Ceynowa Kennzahlenorientiertes Bibliothekscontrolling Perspektiven und Probleme

91

Klaus G. Saur Das Verzeichnis lieferbarer Bücher

110

Renate Vogt Vermittlung von Informationskompetenz im Rahmen der Hochschullehre . . .

117

Dieter Stäglich Der Wandel nordrhein-westfälischer Bibliothekspolitik am Beispiel des Hochschulgesetzes vom 14. März 2000

129

5

Inhalt Wilhelm Held Information, Kommunikation und Medien - Durch Kooperation und Konsolidierung zu Fortschritten

140

Bibliographie der Veröffentlichungen von Roswitha Poll

148

Verzeichnis der Autorinnen und Autoren

155

6

Grußwort des Rektors der Westfälischen Wilhelms-Universität Münster

Universitätsbibliotheken stellen heute innerhalb der Hochschulen hochmoderne Plattformen für die Beschaffung, Erschließung und Vermittlung von Fachinformationen für die Zwecke von Forschung und Studium dar und zählen daher zu den wesentlichen Voraussetzungen sowohl für erfolgreiche wissenschaftliche Arbeit als auch für eine hochwertige Ausbildung. Die Universitätsbibliothek Münster, die als Landesbibliothek in der Region zusätzliche Aufgaben über den wissenschaftlichen Bereich hinaus wahrnimmt, hat in den vergangenen 16 Jahren unter der Leitung von Frau Dr. Roswitha Poll eine hervorragende Entwicklung genommen; die Leistungsfähigkeit und das Profil der ULB konnten entscheidend vorangetrieben werden. So wurden in dieser Zeit zukunftsweisende Neuerungen eingeführt und die Informationsversorgung wie -kompetenz der Nutzerinnen und Nutzer konnte weiter erhöht werden. Zu den wichtigen Innovationen sind vor allem drei Bereiche zu zählen: Zum einen die Einführung der sog. „funktionalen Einschichtigkeit", die auf der Grundlage eines einheitlichen Online-Nachweises der Bestände und der Abstimmung von Neuerwerbungen zu einer Zusammenlegung kleinerer Bibliotheken zu leistungsfähigeren Einheiten geführt hat: So konnte die Zahl von ursprünglich 215 Bibliotheken im Jahre 1988 auf heute 163 verringert werden. Ferner die Anwendung und Entwicklung neuer Managementtechniken: Die ULB Münster gilt heute international als führend in den Bereichen der Leistungsmessung, des Controlling, der Kosten- und Leistungsrechnung und des Benchmarking. Und letztlich erwähne ich den konsequenten Ausbau des elektronischen Dienstleistungsangebots im Bereich der Kataloge, der Datenbanken und der Volltexte, der unter der Ägide von Frau Dr. Poll vehement betrieben worden ist. In diesem Zusammenhang stehen für die Zukunft der ULB weiterhin Fragen auf der Tagesordnung, die der Wissenschaftsrat mit seinen „Empfehlungen zur digitalen Informationsversorgung durch Hochschulbibliotheken" aufgeworfen hat. Es geht dabei u. a. um eine verstärkte Kooperation zwischen der Universitätsbibliothek und dem Zentrum für Informationsverarbeitung (ZIV) der Universität. Der Wissenschaftsrat hält es für geboten, aus diesen Bereichen ein auf die lokalen Bedürfnisse abgestimmtes Informationsmanagement als Einheit aufzubauen, das aus der Sicht der Studierenden und Lehrenden alle integrierten Dienstleistungen anbietet. Denkt man diese Entwicklung zu Ende, kann dies von einer stärkeren Koordination der beteiligten Einheiten bis hin zu einer organisatorischen Verschmelzung von insbesondere Bibliothek(en) und dem Zentrum für Informationsverarbeitung unter einer einheitlichen Leitung führen. Auch dieser Herausforderung kann die ULB Münster dank der hervorragenden Arbeit von Frau Dr. Poll gelassen entgegensehen.

7

All dies macht deutlich: Frau Dr. Poll hat sich als Direktorin der ULB in ganz besonderer Weise für unsere Universität eingesetzt und sich dabei herausragende Verdienste erworben. Die Universität ist ihr daher zu großem Dank verpflichtet.

Prof. Dr. Jürgen Schmidt

8

Grußwort des Kanzlers der Westfälischen Wilhelms-Universität Münster

Regelmäßige und enge Kontakte zwischen dem Kanzler einer Universität und der Direktorin der Universitätsbibliothek sind ganz selbstverständlich. Es gibt vielfältigen dienstlichen Anlass dafür. Und natürlich geht es dabei häufig ums Geld. Dass sich aus solch langjährigen Beziehungen zwischen Frau Dr. Poll und mir auch ein besonderes Vertrauensverhältnis entwickelt hat, sollte eigentlich niemand überraschen. Es reicht zurück in die Zeit ihres Amtsantritts an der Universität Münster. Ich war damals noch Haushaltsdezernent und hatte schon als solcher viel mit der Bibliothek und der Bibliothekskommission zu tun. Das Geschäft galt bis dahin nicht gerade als aufregend. Das sollte sich mit der neuen Direktorin gründlich ändern. Man tut ihren liebenswürdigen Amtsvorgängern posthum sicher keinen Tort an, wenn man feststellt, dass Roswitha Poll die Universitätsbibliothek, aus der auf ihr Betreiben die Universitäts- und Landesbibliothek Münster wurde, ganz neu und nachhaltig geprägt hat. Sie brachte die Idee von der Bibliothek als einer „Informationsdienstleistungseinrichtung" und sehr konkrete Vorstellungen über spezielle Bibliotheksprojekte mit, mischte die Mannschaft der Referenten gehörig auf, was nicht ohne Irritationen abgehen konnte, und verstand es, bei Leuten wie mir erstmals so etwas wie Begeisterung für die Sache der UB zu wecken. Wer Roswitha Poll kennt, weiß, dass sie auch dabei mit der ihr eigenen Sachlichkeit und Nüchternheit zu Werke geht. Aber es wirkt. Viele der Beteiligten begannen damals zu begreifen, dass Bibliothek, Rechenzentrum und Verwaltung, getragen von einem entsprechenden Impuls des Rektorats, gemeinsam die Verantwortung für das Gelingen von Information und Kommunikation quer durch die Universität und weit über deren Grenzen hinaus zu übernehmen haben. Und wer hätte Roswitha Poll schon widersprechen mögen, wenn sie insofern für die Bibliothek eine führende Rolle reklamierte - bei aller Aufgeschlossenheit für die enge Kooperation mit den anderen Partnereinrichtungen in der Universität. Sie ist bis heute die Pragmatikerin geblieben, die immer das Ideale im Sinn und das Machbare vor Augen hat. Die Formel von der „funktionalen Einschichtigkeit" des Bibliothekssystems der Universität Münster gibt beredt Zeugnis von dieser Grundeinsicht. Die systematisch reorganisierten Dienste der ULB erhielten ein neues Gesicht. Auf breiter Front zog die IuK-Technik in die Bibliothek ein, wurde die Rechnerausstattung bedarfsgerecht verbessert. Sehr früh schon wurden sämtliche Neuerwerbungen der Zentralbibliothek, nach Möglichkeit aber auch der dezentralen Bibliotheken sowie im Wege der Retrokonversion ein großer Teil der Altbestände online erfasst. Der gute alte Zettelkatalog hatte ausgedient. Das Fenster zur „Digitalen Bibliothek" wurde weit geöffnet, doch nicht weiter, als es dem Benutzerinteresse diente. Die andernorts mit Vehemenz geführten ideologischen Grundsatzdiskussionen fanden in 10

Münster nicht statt. Implementierung von Verfahren und Investition von Ressourcen folgten einer rationalen Raison und bewiesen das notwendige Augenmaß. So entsprach es dem persönlichen Stil der Bibliotheksleiterin. Klassische bibliothekarische Aufgaben wie die Konservierung der Altbestände, die geradezu liebevolle Pflege der Handschriften, Inkunabeln, Rara kamen dabei nicht zu kurz. In der Ära Poll gelangen einige durchaus spektakulär zu nennende Neuerwerbungen wie die der Bibliothek Fürstenberg-Stammheim, der Dyckschen Handschrift mit dem Epos „Van den vos Reynaerde", des Codex Henrici („Soester Bibel"), mehrerer wichtiger Droste - Autographen, wie zuletzt noch vor wenigen Wochen, oder die Übernahme der Altbestände der Bibliothek des Regierungspräsidenten Arnsberg. Roswitha Polls Phantasie beim Auftun von öffentlichen und privaten Geldgebern war unerschöpflich, ihre Hartnäckigkeit beim Zusammenbringen der beträchtlichen Summen glücklicherweise auch. Sie konnte mit Erfolg lästig fallen. Die ULB und nicht zuletzt die Universität haben davon profitiert. Neuland betreten hat die ULB unter dem maßgeblichen Einfluss von Roswitha Poll mit dem Einstieg in das Bibliothekscontrolling. Im Rahmen eines von ihr eingeworbenen DFG-Projekts war die ULB Münster schon vor 1998 gemeinsam mit der ULB Düsseldorf und der UB Paderborn Modellstandort für die Einführung des Prozesskostenmanagements in wissenschaftlichen Bibliotheken. Projektziel war die Entwicklung einer an die finanzwirtschaftlichen Rahmenbedingungen zentraler Serviceeinrichtungen der Hochschulen angepassten Kosten- und Leistungsrechnung. Sie soll, wie Roswitha Poll es im Vorwort der von K. Ceynowa und A. Coners als Buch vorgelegten Veröffentlichung der Projektergebnisse beschrieb, Hochschulbibliotheken in die Lage versetzen, die Kosten ihrer Leistungen transparent zu machen und ihre Ressourcen kostenoptimal einzusetzen. In der Übernahme dieses Projekts durch die ULB Münster sah deren Direktorin die folgerichtige Weiterführung ihrer Arbeiten in den Bereichen Leistungsmessung und Qualitätsmanagement. Mit der Vorlage der Projektergebnisse sind von Münster aus Standards gesetzt worden, die inzwischen (nach Übersetzung ins Englische und ins Spanische) europaweit Anwendung finden. Mit dem wiederum von der DFG geförderten Nachfolgeprojekt „Balanced Scorecard für Wissenschaftliche Bibliotheken" hat die ULB Münster den eingeschlagenen Weg konsequent fortgesetzt, diesmal gemeinsam mit der Staats- und Universitätsbibliothek Bremen und der Bayerischen Staatsbibliothek München. Das unter der Verantwortung von Roswitha Poll durchgeführte Projekt, dessen Ergebnisse abermals von Ceynowa und Coners publiziert wurden, war auf die Entwicklung eines strategisch orientierten Kennzahlensystems zur Leistungsevaluation wissenschaftlicher Informationsversorgung gerichtet. Es verfolgte das Ziel, die ausschließlich monetäre Analyse der Bibliotheksleistung durch die umfassende Einbeziehung nicht-monetärer Faktoren zu erweitern. So ist auf dem Gebiet des Universitätscontrollings unter der Leitung und persönlichen Mitwirkung von Roswitha Poll landesweit, aber auch national und international mit großer Aufmerksamkeit beachtete Pionierarbeit geleistet worden, die sich fest mit dem Standort Münster verbindet.

11

Die ULB ist als zentrale Betriebseinheit nach Aufhebung der ehedem unmittelbar aus der Titelgruppe 95 des Landeshaushalts gesicherten Mittelzuführung aus der Perspektive des die Ressourcen der Universität jetzt eigenständig verteilenden Rektorats natürlich in erster Linie Bedarfsträgerin, wie es auch die anderen zentralen Einrichtungen, aber auch die Fachbereiche sind. Sie ist allerdings auch Leistungseinheit und inzwischen insofern einem Fachbereich durchaus vergleichbar. Die Fachbereiche stehen bei der Mittelverteilung miteinander im Wettbewerb und können ihre Position aufgrund erbrachter Leistungen im Rahmen des in der Universität praktizierten Kennzahlenschlüssels verbessern. Die ULB kann das nicht, obwohl sie in ihrem Geschäftsbereich beachtliche Leistungen vorzuweisen und namentlich nicht unerhebliche Drittmittel und - mit viel Geschick ihrer Direktorin - über die Jahre hin noch mehr sogenannte unechte Drittmittel insbesondere vom Wissenschaftsministerium des Landes eingeworben hat. Vielfach ist dabei auch nach dem bewährten Prinzip der „matching funds" verfahren worden, d.h. den von außen bereitgestellten Mitteln mussten universitäre in vergleichbarer Höhe zur Seite gestellt werden. Dabei hat es sich in aller Regel um zusätzliche Investitionsmittel gehandelt. Sie sind unter der Verantwortung der Bibliotheksleiterin garantiert gut angelegt worden. All die anderen Initiativen und Aktionsfelder Roswitha Polls, die eine ausführliche Würdigung verdienten, können hier allenfalls angedeutet werden: ihre Rolle als Planerin und unermüdliche Antreiberin der baulichen Erweiterung der ULB, als Veranstalterin vieler bedeutender Ausstellungen in dem von ihr für diese Zwecke realisierten schönen Pavillon zur Aa hin, als Ideengeberin und tatkräftige Akteurin für das gemeinsam mit dem Zentrum für Informationsverarbeitung (ZIV), der Koordinierungsgruppe für die computergestützte Hochschullehre (cHL) und der Verwaltung der Universität erarbeitete und vom Rektorat verabschiedete Konzept für den IKMService im Rahmen eines umfassenden Informationsmanagements der Universität Münster. Zu erwähnen bleiben ferner ihr persönlicher Anteil an der Formung eines kompetenten Teams von Mitarbeiterinnen und Mitarbeitern, die sie gefördert, aber auch immer wieder gefordert hat, und schließlich ihre für die Bibliothek so außerordentlich wichtigen und ergiebigen internationalen Kontakte. Sie ist eine im In- und Ausland gefragte Referentin, wo immer es um Bibliotheksbelange geht. Als Kanzler der Universität Dortmund und als Mitglied des Vorstands des Verbandes der Bibliotheken in Nordrhein-Westfalen hatte ich eine Reihe von Jahren die Gelegenheit, gewissermaßen aus der Außensicht zu beobachten, wie sich die ULB Münster unter der Leitung von Roswitha Poll entwickelt und zu welch zusätzlichem Ansehen sie es dabei gebracht hat. Sie zählt gerade auch unter Wettbewerbsaspekten heute unbestritten zu den besten Adressen im Lande. Roswitha Poll hat die Universitäts- und Landesbibliothek Münster für die Zukunft sehr gut gewappnet. Für ihre große Leistung gebührt ihr ebenso großer Dank.

Dr. Klaus Anderbrügge

12

Rowena Cullen

An Integrative Model for the Evaluation of Information Services

1. Introduction Evaluation has been seen as a core activity for libraries since the origins of public and academic libraries in the late nineteenth and early twentieth centuries. The second half of the twentieth century has seen extensive research carried out on methods for evaluating libraries and information services, that has ranged across a wide spectrum of approaches, each of which has been dependant on a particular view of the library or information service.1 At the same time, within libraries themselves the use of evaluation as a management tool has remained conservative—a narrow range of methods being used by most libraries, underpinned by a generally inadequate understanding of the importance of evaluation, or the underlying conceptual basis of any particular method. However, as information service emerges into the twenty-first century as a complex range of networked electronic services, based on multiple ownership, multiple points of access, and multiple accountabilities, it is imperative that library and information service managers should not only look to a wide range of methods to evaluate the services they offer, but also have a more profound understanding of the complex environment that they are operating in, and are seeking to evaluate. 1.1 Existing Models of Evaluation of Information

Services

There are two key factors that inevitably influence evaluation of information services. The first is the view held by the stakeholders of the service, (that is, the library or information service itself, its managers, its funders, and users) as to what its function is. The second factor is the conceptual model of evaluation held by whoever is responsible for evaluating its effectiveness. Any framework for evaluation, whether originating internally in the institution, or from external agencies will be based on the construct of the service held by that particular stakeholder. For example, a city council, seeking to evaluate public library services may take the view that its primary task is to provide recreational reading and social contact for citizens, and see the library as part of its leisure services, along with parks and other recreational facilities. Despite the fact that this limited view may not be held by the library staff and managers, this view of the library determines the way the council seeks to evaluate library services, and may lead to over-reliance on use levels (or market penetration, as

Cullen, Rowena. "Does Performance Measurement Improve Organisational Effectiveness? A Postmodern Analysis." Performance Measurement and Metrics: the International Journal for Library and Information Services. Sample issue, 1999, 9-30.

13

Rowena Cullen Poll defines it)2, and user satisfaction, as key performance indicators. Another council may see library and information services as a major contributor to the knowledge society, a source of life-long learning, and competitive intelligence for local businesses. It will be more likely to seek evidence of the use and value of these activities. Thus the weight given in evaluation to the various aspects of an information service will vary according to the views held by the institution itself of its core purpose and roles, and that of its main external stakeholders, and the balance achieved between the two. A wide variety of such constructs are reflected in the extensive literature on library and information service evaluation, and the different perceptions held by researchers and practitioners about evaluation over time and in different contexts. These various constructs are reflected in the many definitions of effectiveness used in the Library and Information Science literature. A definition used by Lynch in 1983, and cited frequently since, however, may serve to provide some consistency. Lynch defines measurement as "the process of ascertaining the extent or dimension of quantity of something", performance as "the doing of something, an activity", evaluation as "the process of determining whether something is what you want it to be", and effective as referring to "something which does well that which it is supposed to do". All four terms are combined in a succinct self-explanatory statement. 'The results of measurement can be used to evaluate the performance of a library and thereby determine whether or not it is effective",3 Over the years effectiveness has remained the most controversial term of the three and its use has changed somewhat to encompass a more direct response to the organisation's environment and stakeholders. For example, Childers and Van House maintain effectiveness deals with "goodness, achieving success, and the quality of performance." When assessing an organisation's effectiveness, they suggest the following questions should be addressed: • • • •

To what extent does the organisation achieve its goals (input, process, output or outcome goals)? To what extent is the organisation a healthy operating unit? To what extent can the organisation capture from the external environment the resources needed to survive or thrive? To what extent are the various stakeholders'1 priorities met?4

This is a useful starting point for an overview of concepts of evaluation adopted by the LIS profession and researchers over the years, ranging as it does over key concepts of inputs, outputs and process, and the general systems approach to organisational effectiveness, as well as the multiple constituencies approach that reflects the role of stakeholders in evaluation.

3

4

Poll, Roswitha. "Performance Measures for Library Networked Services and Resources." The Electronic Library, 19,5 (2001), 307-314. Lynch, Mary Jo. "Measurement of Public Library Activity: the Search for Practical Methods." Wilson Library Bulletin, 57 (1983), 388. Childers, Thomas A and Nancy A. Van House. What's Good: Describing your Public Library's Effectiveness. Chicago, American Library Association, 1993, 7.

14

Integrative Model for the Evaluation of Information 1.2 Major evaluative

Services

paradigms

The general systems model of an organisation is a useful framework for exploring the development of evaluative approaches to library and information services (see Fig. I). Most evaluative approaches developed up until nineteen eighty fit into this model, focusing on inputs, processes or outputs, all of which are easily quantified. Early forms of evaluation, for example, focused heavily on inputs such as the size and use of the collection, 5 and remain a popular measure. University and public library statistics covering the number of items added to the collection, the total number of items in the collection, and more recently the number of electronic services made available are still collected, published, and used for benchmarking. Outputs such as the number of book loans registered, users' questions answered, 6 or databases accessed, 7 are used in much the same way. Internal process measures, focused in the past on the delays in making books available to users that occurred in the acquisition or cataloguing process, or on delays in obtaining interlibrary loans, or reshelving materials, 8 are now translated into measures of system and network availability, download times for electronic documents etc.9 Management Information Systems (MIS) and Decision Support Systems (DSS) are increasingly being used for measuring processes, now that automated library systems can routinely report on library activities, and use. 10 All of the modes of evaluation described above, from input studies to outcomes, whether they are direct outcomes (described by some as impacts) or indirect outcomes, such as user satisfaction and research income earned on the basis of research output, can be aligned with the general systems model of information services (see Fig. I). These are all important measures, and they all contribute in some way to the ability of a library to meet the needs of its users. But they remain simple measures and tell little about the adequacy of the inputs, processes or the user's satisfaction. The model does not reflect the varying roles and expectations of users and their differing needs at different times, or the varied characteristics of information services themselves. Most importantly, the systems model does not reflect the importance of stakeholders, and their expectations in both the design and evaluation of information services. An ability to incorporate these multiple constituencies and complex relationships into an evaluative model is a key issue in the acceptability of performance measurement and evaluative frameworks.

6 7

8

9 10

Baker, Sharon and F. Wilfred Lancaster. The Measurement and Evaluation of Library Services. 2 nd ed. Arlington, VA: Information Resources Press, 1991, 55, 79. Baker and Lancaster, 1991, 229. Bertot, John. "Measuring Service Quality in the Networked Environment: Approaches and Considerations. " Library Trends, 49,4 (2001), 758-775. Kantor, Paul. Objective Performance Measures for Academic and Research Libraries. Washington, D.C: Association of Research Libraries, 1984. Bertot, 2001. Cullen, Rowena. "A Bottom-Up Approach from Down Under: Management Information in Your Automated Library System." Journal of Academic Librarianship, 18,3 July (1992), 146-150.

15

Rowena Cullen

Feedback

Outcomes

Processes Ress . 'ces (staffing, financial)

Input ι Output rates arw benchmarking

Statistic·.

O S S I MIS

Benchmarking

Efficiency

Collection evaluation

C o M I B a n i f t analysis

I

l o a n s ι documents delivered

Health decisions

Database» accessed

B u s i i « » decisions

Reference sra^tWes an^.vered

Student learning outcomes

Classes attended

Research produetMty CKafiOn studies

NsK'IQrif & . d -il:|! , Download times

Figure 1: Models of Evaluation Used in Library and Information Services 2. An Integrative Model for the Evaluation of Information Services An integrated approach to the evaluation of information services must take into account both formal and informal processes relating to the life cycle of information, its creation, procurement, organisation, distribution, dissemination, use, retirement and preservation, and the manifestations of such processes in a number of different contexts such as libraries, online database systems, digital libraries and the World Wide Web. A process of evaluation must therefore be able to adjust to the individual parameters of these different models of information service, many of which now overlap and interact. However, even these new information services and information retrieval systems operate within a specific social construct in which the dominant stakeholders determine both what knowledge or information is deemed to be of value, and the basic model underlying the creation, organisation, dissemination and use of information in that system, what will be preserved and how it will be archived. Any evaluative framework for information services must therefore either take into account the multiple factors of the construct, that is, it must attempt to take into account all the identifiable stakeholders of the system and their various needs, or alternatively, it must overtly focus on a particular perspective of the service for the practical purpose of designing an evaluative system for a specific social, administrative or political purpose. A simple representation of an integrated model of information services is presented in Figure II where the main elements of an information service are demonstrated. Central to the model is the information service itself, which may be a single service or a complex group of services, in both print and digital formats, provided by a large library system. It may be a formal information service designed to meet the needs of a large group of stakeholders, or an informal service that has grown up to meet the needs of one specific group. The service may be a simple online database created by an organisation for its members, such as the original subjectbased databases Psychological Abstracts or Chemical Abstracts, which are now disseminated by online database system vendors as Psychlit and CA Abstracts. Other 16

Integrative Model for the Evaluation of Information

Services

examples might include the provision of clinical information from the research literature to the family physician, or government information disseminated through the World Wide Web.

Mission, Purpose and Attributes

User Groups

Information Products

Stakeholders

Figure 2: Key Elements of an Information Service The model can also be used to analyse a library system, from a one person library designed to serve the needs of a small company, although even this may have several different user groups, with differing needs, to a major research library expending tens of millions of dollars and employing several hundred staff. A large library incorporating both print and digital resources has many different user groups who have multiple needs, as well as several different stakeholders. A wide range of evaluation methods will be needed to encompass all its facets, and uses.11 2.1 Attributes of the elements of an information

service

Each of the elements of an information service has sub-elements, and these have many and varied attributes. An integrated evaluative model must be capable of taking all of these into account. Some of the attributes pertaining to each element are described below. 2.1.1. Mission, purpose and attributes There are several attributes of the information service itself which have a significant impact on its performance, and which are integral to any comprehensive evaluation of the service. One factor that is crucial to both the design and evaluation of any infor-

11

Cullen, Rowena J. "Evaluating digital libraries in the health sector. Pts 1 and 2." Health Information and Libraries Journal, Dec 2003 [in press].

17

Rowena Cullen mation service is the purpose for which it is created, or has arisen. Purpose, in conjunction with the incorporation of feedback into organisational planning and decisionmaking is critical to determining the effectiveness of information services. Contingent on purpose is the actual configuration of the service itself—what services are offered, how, to whom, at what cost etc. Significant attributes of these components of the service also include the service quality factors such as the relevance and timeliness of services, and the helpfulness and courtesy of staffing delivering them. The level of resources available to the organisation or service, the source of these, and the characteristics of these resources are also significant attributes of the information service that will have an impact on its evaluation. In most cases financial resources will dictate the amount of money that can be expended on staffing and buildings as well as the collection itself, which in a complex information service such as a library comprises a number of component services (e.g. book loans, reservation services, reference services, information literacy education, access to indexes and abstracts, and e-journals). Input-based, process-based and output-based models of evaluation focus heavily on these components, and their attributes, and make them the basis of their performance measurement. Staffing is a key issue; the overall quality, commitment, basic training and continuing professional education of staff will all impact on the type and quality of services offered. In different countries and cultures, staffing resources may be constrained by employment policies and laws, 12 and, in developing countries, by the availability of trained staff. In countries in which information service staff are mostly professionally trained to degree level they bring their own conception of the information service, and what its purpose and values should be. These include professional ethics, and externally derived standards, that may or may not agree with other stakeholders, such as councils, or university administrators. Organisational culture, and the responsiveness of staff to change, especially, but not exclusively in technology, are also important considerations. 13 All of these will have a significant impact on service delivery and evaluation. There are, in addition, newly identified organisational attributes which have a significant impact on information services including Lakos's 'culture of assessment', 1 4 and qualities of leadership which are now perceived to be important aspects of organisational culture, and to have a considerable impact on the overall effectiveness of the organisation, and inter-relate closely with its ability to acquire resources, respond to stakeholders, etc. 15

12

13

14

15

Bunzel, Jürgen and Roswitha Poll. "German Academic Libraries: Tradition and Change." Journal of Academic Librarianship, 28,6 (2002), 418-425. Lancaster, F. W. and Beth Sandore. Technology and Management in Library and Information Services. London: Library Association Publishing, 1997, 13-26. Lakos, Amos. "Building a Culture of Assessment in Academic libraries - Obstacles and Possibilities." Conference paper presented at Living the Future //, Tuscon, Arizona, April 22 1998. Accessed 12/8/03. Hernon, Peter, Ronald R. Powell and Arthur P. Young. "University Library Directors in the Association of Research Libraries: Next Generation, Part one." College and Research Libraries, 62,2 (2001), 121.

18

Integrative Model for the Evaluation of Information Services 2.1.2 Stakeholders A different set of variables relates to the second key element of an information service, the stakeholders of the organisation, and any other group which asserts some ownership of its activities or the results of those activities. External stakeholders include resource allocators, that is, city councils, university administrators, company directors, or executives making decisions about resource allocation and company infrastructure. Consortium partners, and any other organisations with which the institution has entered into collaborative arrangements will also need to be taken into account. Collaborative relationships modify many aspects of organisational performance, and culture, and can alter the relative power of existing stakeholders. External stakeholders also include secondary consumers, at one remove, of the product the organisation provides—parents, and employers of students with degrees in which high quality information services should be a significant component, alumni of a university who value its library, clients of a law firm who need to know that legal advice was up-to-date and based on good information. The power of external stakeholders, albeit claiming to represent the interests of internal stakeholders, or customers, is often shown when groups such the 'Friends' of a public library are able to have decisions about opening hours or user charges reversed, as sometime happens; for example, in New Zealand a powerful 'Friends' association of a major research library succeeded in using court action to enforce a change in legislation to reflect its interests.16 Internal stakeholders are often considered to include the users, or customers of the service. These would include graduate and undergraduate students, faculty, parents, children and adult users of a public library, or lawyers dependant on the company law library. In the model presented here users are considered to be such a significant stakeholder group that they are treated as a separate element. Hernon and Altman argue that there is a clear distinction between customers, and stakeholders, "who have an interest in the organization, usually related to funding . . . . Stakeholders may exert influence, primarily through funding or legislation, but they are not customers".17 Staff are also an important group of stakeholders, as well as a resource to be considered under service components that relate to inputs. The construct of an ideal information service held by both external and internal stakeholders, and their expectations of service derived from these, will inevitably impact on the design of the service, and the forms of evaluation put in place to assess performance.

16

17

Underwood, Rachel. "David and Goliath? - or Perhaps Lucy Lawless! A Recent History of the Alexander Turnbull Library." A paper presented at Past, Present, Future, a forum hosted by La Trobe Library, held at the State Library of Victoria, May 24 2001. The New Zealand National Library Act (2003), passed subsequently, enshrines in law the independence of the Alexander Turnbull Library, prohibits the sale of its historic collections, and appoints Guardians to protect its interests. Hernon, Peter and Ellen Altman. Service Quality in Academic Libraries. Norwood, N.J.: Ablex. 1998, 5.

19

Rowena Cullen 2.1.3 User groups Turning, then, to users of the service, a large number of variables should influence evaluation of the service whether or not they have been taken into account in the design of the service. Users of any particular information service may be a homogeneous group with a narrow range of needs and requirements of the service, or they may include many varied groups, or 'audiences' as they are referred to in the world of web evaluation, who bring different and often conflicting demands to a service. They may in themselves have a variety of needs depending on the multiple roles they play in the organisation (physician, researcher, teacher, or professional), or in their personal lives. They will have characteristics derived from their personalities, ethnic background, and cognitive preferences as well as their organisational or societal role. Cognitive preferences, for example, may lead users to select print over electronic resources, (or vice versa) and to prefer mediated or unmediated services, or information with specific characteristics as well as information services with specific characteristics.18 Cultural differences affect how knowledge is created, stored and disseminated, and may lead users to prefer oral and visual information sources over print.19 Barriers to use of an information service may be physical or intellectual,20 and may be dependant on system variables or user demographics and other characteristics.21 Research continues to reinforce Mooers law, that someone in search of information will seek it only if it is more troublesome not to have it;22 an effective information service is therefore one that overcomes this natural tendency of information seekers, and in which the benefits of possessing information clearly outweigh the disadvantages of not having it, and the effort of obtaining it. All the factors listed above can be seen to affect the user's satisfaction with individual aspects of service delivery and their view of the information service overall. While few information services are able to be so flexible and responsive as to meet such a wide range of requirements, the extent of knowledge of these requirements and the organisation's willingness to engage with a range of user characteristics should affect system design and be included in any model of evaluation.

18

19

20

Cullen, Rowena. "The Medical Specialist: Information Gateway or Gatekeeper for the Family Practitioner." Bulletin of the Medical Libraries Association, 85,4 (1997), 348-355. Cullen, Rowena. "Biculturalism and librarianship in New Zealand: a more fundamental change than Information Technology." Proceedings, 62nd IFLA General Conference, 1996. Wyatt, Jeremy. The Effects of Manual Paper Reminders on Professional Practice and Health Care Outcomes: a Systematic Review of 22 Randomised Controlled Trials Including 41705 Patients. Revised Final Project for NHS R&D Project no IMP 15-11, 14/12/2000.

2000. 21

22

Wilson, Tom and Christina Walsh. Information Behaviour: an Interdisciplinary Approach. British Library Research and Innovation Report 1996. Mooers, Calvin N. "Mooers' Law, or Why Some Retrieval Systems are Used and Others are Not." American Documentation, 11,3 (1960), ii.

20

Integrative Model for the Evaluation of Information 2.1.4 Information

Services

products

Unless an information service is defined as a single information resource, such as a database, it is dependant on the availability of information products and services produced by a number of different providers, publishers, and vendors. Each of these, and their products, have characteristics that impact on service design and evaluation. Each single information resource will have specific characteristics that impact on service delivery, dependant on the type of information, the audience for which it is intended, its scope etc. The subsidiary attributes of the various information products are dominated by the main attribute, the content, or the knowledge that the system is designed to disseminate. This will depend in part, but not exclusively, on its creation, or circumstances of its origins - the creators, their purpose, their intended audience etc. Information content has its own lifecycle which varies according to the subject matter or discipline.23 For example, knowledge in the humanities tends to have a longer 'shelflife' than knowledge in the sciences (texts of interest can be hundreds if not thousands of years old), and libraries serving students and scholars in the humanities must provide access to older material. 24 Public libraries will provide multiple copies of recent fiction in high demand, and then weed their collections to be able to maintain good coverage of an entire literature. A law company library may require only current legislation compared with an academic law library which has to take a historic approach. Services will therefore differ in each of these contexts, and systems of evaluation need to be sensitive to such contexts. Knowledge provided by public institutions is generally required to comply with the standard criteria for knowledge sources, that is, it is expected to be accurate and up-to-date and be able to demonstrate that is from an authoritative source. For example, in the health sector, the National electronic Library for Health (NeLH) focuses exclusively on accredited and evaluated sources to support evidence-based healthcare. 25 Variations from these criteria occur primarily only in major research and heritage collections, whose collections may cover all aspects of knowledge in society. The standard criteria for knowledge sources would therefore be expected to form part of an evaluation framework, as they do, for example, when reference collections are evaluated for their authority, accuracy, currency and scope, alongside reference staff competence and service delivery. Knowledge also has a cost (high or low) in terms of the difficulty and any exclusivity.26 For example, information in a manually indexed subject specific database has a high development cost that is reflected in charges made for the service and the value of that service to researchers in the field. Information in a company register 23

24

25

26

Vickery, Brian and Alina Vickery. Information Science in Theory and Practice. London: Butterworths, 1987, 245 Stone, Sue. "Humanities Scholars: Information Needs and Uses." Journal of Documentation, 38,4(1982), 292-313. Turner, Alison, Veronica Fraser, J.A.Muir Gray, and Ben Toths. "A First Class Knowledge Service: Developing the National electronic Library for Health." Health Information and Libraries Journal, 19,3 (2002), 133-145. Vickery, Brian and Alina Vickery. Information Science in Theory and Practice. London: Butterworths, 1987, 22-23.

21

Rowena Cullen database may have low collection costs, but a high commercial value that is also reflected in charges made for its use. Information on the World Wide Web may be provided free of charge to users for political or social reasons. Both of these lie behind the provision of free public access to the high-cost database Medline where costs of acquiring and indexing the journal literature have always been borne by the National Library of Medicine as part of its appropriation from the United States government, but to which the Web now permits universal access, along with access to all other US government information. Alternatively knowledge may be provided at low cost by a research team who wish to disseminate their findings, or it may be cheaply gathered and freely disseminated on the Web by an individual, a company, or a government agency wishing to circulate the information to the widest possible audience. An additional factor that must be taken into account is the number of books and journals now available electronically, and incorporated, through networked services, into digital library services. A very large proportion of e-journals are now part of major library subscriptions, and users would be unaware of whether they are accessing a 'free-to-air' journal or one for which their institutional library or information services has contracted for access. The situation is further complicated by the surge in interest in Open Access to journals, and the deposit by some authors (especially in medicine, science and technology) of their papers in some form of Open Archive.27 Book publishers, by contrast, do not yet use this medium as extensively since cost structures of e-books have not been well established, although a significant number of textbooks, and older literary texts can now be found online. The rise in both e-books and e-journals suggest that new economic structures for the dissemination of information on the Internet are emerging, that will inevitably affect evaluation in some way. A second very important set of attributes of information products that form part of an information service, and need to be evaluated, come under the general concept of intellectual access. In print format this may encompass the way in which an information tool, for example a major multi-volume encyclopedia, or an index is arranged, and whether the index terms have been well structured into a controlled vocabulary. The general searchability of the tool is very important, and carries both a cost in its creation, and an added value to users.28' 29The related concept of bibliographic organisation and control, including the use of metadata to describe electronic information, is equally important in providing access to information sources, and needs to be included in evaluative processes.30 Both concepts are critically important when trans-

27

28

29

30

For current information on these initiatives readers should visit the Public Library of Science web site at: ; and the SPARC web site (of the Scholarly Publishing and Academic Resources Coalition) at Baker, Sharon and F. Wilfred Lancaster. The Measurement and Evaluation of Library Services. 2 nd ed. Arlington, VA: Information Resources Press, 1991, 273. Vickery, Brian and Alina Vickery. Information Science in Theory and Practice. London: Butterworths, 1987, 262. Baker, Sharon and F. Wilfred Lancaster. The Measurement and Evaluation of Library Services. 2 nd ed. Arlington, VA: Information Resources Press, 1991,181.

22

Integrative Model for the Evaluation of Information Services ferred to the electronic context, and are emerging as key issues in evaluating webbased information services and products. Each service will also have an intended audience or user group (which may not be the same as the intended audience for any single item disseminated by the service), and there will be constraints at times on the availability of the content. All of these factors have an impact on the design of the system and on appropriate systems for evaluating its effectiveness. Value and cost both interact extensively with service attributes (e.g. purpose and resources), and with stakeholder and user group attributes. 2.2 Integrating attributes of each component of an information service into the model. Fig. II can then be re-drawn with attributes of each of the components added (see Fig. Ill), making clearer the interrelationship between each of the components. Fig. Ill presents a more complex representation of the four key elements of information service, with selected attributes added. Each of the elements of an information service, and their individual attributes can be evaluated or form the basis of an evaluation system, or, as in the two examples outlined below, will all be touched on to some extent. Criteria based on the attributes outlined above, and the extent to which services meet needs defined in terms of these can become measures of performance. So also can the interrelationships between these various dimensions— for example, the interrelationship between service attributes and user needs, cultural or cognitive preferences, the interrelationship between information attributes and user needs, and so on. These, as we have seen, are clearly linked to service quality and satisfaction. The majority of the evaluative models which have been used in the past are based on the attributes of one or more of the elements of an information service as shown in Figure III, and the relationships between these attributes. For example, explorations of the perceptions of various groups of stakeholders, as in the work of Van House and Childers31 and Cullen and Calvert, 32 focus on the interrelationship of the attributes of stakeholders and service attributes but also explore users' requirements of information services, both implicit and explicit, and their perceptions of the quality of these services. Analysis of the perceptions of these various stakeholders highlight the different constructs of an information service operating for each group, and reveal a significant gap between user perceptions and information professionals' perspectives on aspects of information service and delivery which are the focus of current research on service quality. Research into service quality literature carried out by the ARL 33

32

33

Van House, Nancy A. and Thomas Childers. "Dimensions of Effectiveness II: Library Performance." Library & Information Science Research , 12,2 (1990), 131-153. Cullen, Rowena J. and Philip J. Calvert. "Further Dimensions of Public Library Effectiveness: Report on a Parallel New Zealand Study." Library & Information Science Research, 15(2) 1993, 143-164. Cook, Colleen and Bruce Thompson. "Reliability and Validity of SERVQUAL Scores Used to Evaluate Perceptions of Library Service Quality." Journal of Academic Librarianship, 26,4 (2000), 248.

23

Rowena

Cullen

and by Hernon and Nitecki 34 further explores these complex relationships. Only by placing this research within an integrated model however, can we see its relationship to other models of evaluation, a relationship which enhances the value of each model in turn. User Groups

Mission and Purpose ] I Purpose and type of service \Miat. to whom why &

Homogenous or varied groups

how

Single or multiple roles

I Organisational culture I Component services offered Timeliness - Relevance Helpfulness etc I Ressources -1 inancial - Staffing - Physical

Individual Users Personal information seek ng styles Needs/expectations - Cuiural influences aarriers

Stakeholders

Informal Internal

High I low production costs

- Staff Parent Organisation

High I low quality High / low value

External Availability - Consortium partners - Friends groups - Community

Audience

Figure 3: Attributes of the Elements of an Information Service that Impact on Evaluation Similarly the literature on information needs, cognitive preferences, and barriers to information is enhanced and made more meaningful by finding a place within an integrated model of evaluation, instead of remaining isolated within a specific discipline. The five criteria for preferring one information source over another which have been shown to determine physicians' choice of information source, availability, searchability, understandability, credibility, and applicability, 35 can be generalised to other user groups, to form the basis of an integrated evaluation system. Similarly, information services which take into account the known barriers to information use in their design and implementation, have been shown to lead to better outcomes than

34

35

Hernon, Peter and Danuta A. Nitecki. "Service Quality: a Concept not fully Explored." Library Trends, 49,4 (2001), 687-708. Connelly, Donald P., Eugene C. Rich, Shawn P. Curley, and John T. Kelly. "Knowledge resource preferences of family physicians." Journal of Family Practice, 30 (1990), 353359.

24

Integrative Model for the Evaluation of Information

Services

those which disregard the questions of barriers.36 This research needs to be incorporated into an integrated model of evaluation. 2.3 Models for evaluating digital libraries fit the integrative model of evaluation Most formal evaluation to date has focused on the mission, purpose and attributes of the information service, as shown in the upper left-hand quadrant of the model outlined in Figure IV- elements which can be analysed using measures of inputs, outputs, or by comparing ratios of inputs to outputs to investigate how efficiently resources are managed for a given output.

Purpose and type of service - What to whom, why & how Organisational culture User Groups

Component services offered -Timeliness - Relevance - Helpfulness etc

Homogenous or varied groups Sil igte or multiple roles

Ressources - Financial - Staffing - Physical

Individual U s e r s

Personal information seeking styles - Needs/expectations - Cutura! influences - Barrier,

Information Products

^

H i g h I law p r o d u c t i o n c o s t s

Stakeholders

Internet Staff - Parent Organisation

H i g h I l o w quality H i g h I low value

External Availability

Consortium partner* - Friends groups - Community

Audience

Figure 4: Evaluation of an Information Service that Focuses on the Mission, and Planning Activities of the Organisation Attributes of information products can similarly be identified using straightforward analysis of documentation, and activities. Such measures fit well within the constraints of methods which have been used in the past—measures of extensiveness, ef-

36

Wyatt, Jeremy. "Clinical innovation methods to support appropriate practice." In: Zollner, Η. Proceedings of the WHO Expert Workshop on Appropriateness in Health Care Services, Koblenz, Germany 23-25 March 2000.

25

Rowena Cullen ficiency, cost, quality and effectiveness.37 The attributes on the right-hand side of the model have been less well addressed by research. While the work of Childers and Van House38, and Cullen and Calvert 39 has examined expectations of stakeholders, this has not been converted into measures routinely used to assess library effectiveness. Even less work has been done on the application of information seeking styles and preferences of users, as groups and individuals, as a basis for evaluation. To investigate these elements further, and to further probe the 'fit' between user expectations and information service and product attributes, we will need models of evalua-tion which go far beyond measures of satisfaction used to date. Even the Servqual model 40 ' 41 cannot analyse with any sophistication the ability of information services to overcome barriers, and meet user needs, and deliver measurable, quantifiable outcomes. Figure V shows how from a slightly different perspective, the needs and views of user groups, and customer service, can become the focus of evaluation. The 'fit' between user expectations of service, and the perceptions of information service managers (a SERVQUAL gap 1) is one evaluative model represented by the link between Customers/users and the Mission, purpose and attributes of the information service in this version of the model. Similarly the 'fit' between the selection of information products (print or electronic) and users' needs is represented by the link between Customers, and Products. The 'fit' between customers' needs and wants, and stakeholders' perceptions of the information service, which underlies the multiple constituencies approach to evaluation is also represented in this model. Thus, the integrated model itself is seen to be a fluid construct, which can reflect whichever perspective is selected for evaluation 3. Evaluation Methods which better reflect the Integrative Model These complex and fluid models present a real challenge in terms of an appropriate methodology and a 'research construct' to operationalise the evaluation. There are, however, methods emerging that address all these dimensions, and interestingly, they are being applied to new models of the library, or information service —the customised, audience-orientated 'born digital' library. For example, to evaluate the Perseus Digital Library, Marchionini goes beyond the traditional measures of physical libraries (circulation, collection size, patron visits, reference questions answered, patron satisfaction, and financial stability) in order to reflect the reality of the new digital environment. Marchionini outlines a triangulated approach to the evaluation of the Perseus Digital Library (PDL) in order to "make inferences about and develop arguments about PDL meaning and impact . . . based on the belief that evaluation is a research

37

38

39 40

41

Hernon, Peter and Ellen Altman . Service Quality in Academic Libraries. Norwood, N.J.: Ablex, 1996 , 19. Childers, Thomas and Nancy van House. "Dimensions of Public Library Effectiveness." Library & Information Research, 11,3 (1989), 273-301. Cullen, and Calvert. 1993. Hernon, Peter and Ellen Altman. Assessing service quality: Satisfying the expectations of library customers. Chicago: American Library Association, 1996. Hernon, Peter and Danuta A. Nitecki. "Service Quality: a Concept not Fully Explored." Library Trends, 49,4 (2001), 687-708.

26

Integrative Model for the Evaluation of Information

Services

process that aims to understand the meaning of some phenomenon situated in a context, and the changes that take place as the phenomenon and the context interact". 42

Customers I Users Homogenous or varied groups Single or multiple rotes Individual Users Mission and Purpose J

tMT

-

Personal information seeking styles

J Purpose and type of servkr VWmS to whom whyS how Organisational culture Component services offered • Timeliness

I

" • H H B I - Needs/expectations - Cuturai influences - Barriers

Relevance

Helpfulness etc I Ressource«

- Financial Staffing Physical

Stakeholders

Information Products Internal

High / low production costs

- Staff Parent Organisation

High I low quality High I low value

External Availability Consortium partners Fi ends groups Community

Audience

Figure 5: Evaluation of an Information Service that Focuses on the Customers of the Organisation Marchionini's evaluation focuses on four entities, learners, teachers, the technical system, and the content, that can be equated to the elements of an information service identified in Figure III, and uses a variety of methods to evaluate each entity (data collection methods include observations, interviews, surveys, document and learning analysis, observation and transaction log analysis). This multifaceted evaluation thus addresses, inputs, processes, and outputs as well as outcomes—both staff research, and student learning outcomes, are evaluated. Aspects of information needs of various user groups, varying roles of users, cognitive preferences, and barriers to use are assessed along with information service attributes (primarily technologically based) affecting the relevance, timeliness, and ease of use of the system. The match of audience requirements to content, and attributes of content, or products, are also examined. Current endeavours in this and other systems to examine outcomes necessarily raise questions about cost and benefits, and Return on Investment. Again, the integra42

M a r c h i o n i n i , Gary. " E v a l u a t i n g D i g i t a l Libraries: a L o n g i t u d i n a l a n d M u l t i f a c e t e d V i e w . "

Library Trends, 49(2) 2000, 304-333.

27

Rowena Cullen tive model can help with this increasingly important aspect of evaluation, by addressing relationships across the four key elements of the information service, and developing new ratios—not the old ones based on efficiency, but new measures, initially of effectiveness, and eventually of cost-effectiveness, because the element of user needs, and application of information is brought into the formula. We would then move cost benefit-analysis from the Process section of Fig I to Outcomes, since it is another important measure of the outcome of information service delivery, which, for its calculation, depends on impact measures. Hill et al show how this can work in practice. Using similar research methods as Marchionini, Hill et al investigate the expected Return on Investment (ROI) for participants and staff of the Alexander Digital Library. Applying the same mix of quantitative and ethnographic research, interrogating users about time saved, and investigating levels of productivity, Hill et al develop a model that can produce valuable well-rounded evidence of the effectiveness of the system. "Given that these studies reveal situated uses of a system under development as well as user interaction in the library environment that the system is seeking to augment or emulate, the ROI is judged to be high." 43 Thus these more recent evaluative models have independently reached a similar stand-point—taking into account users, content, and system or service attributes, and relating these to the issue of accountability to stakeholders, including but not focusing exclusively on Return on Investment, as a primary outcome measure. The influence of integrative models such as those outlined here is likely to become stronger as libraries and information services become more complex digital environments, and as the selfreporting functionality of digital information services tells us more about users' responses to these new systems. At the same time, these new models also place great emphasis on better communication with users, and better understanding of users' in7formation needs, and preferences; integrative models of evaluation will be the new evaluation paradigm of information services.

43

Hill, L. L„ Dolin, R„ Frew, J., Kemp, R.B., Larsgaard, M„ Montello, D.R., Rae, Μ-Α., & Simpson, J. User evaluation: summary of the methodologies and results for the Alexandria Digital Library, University of California at Santa Barbara, University of California.

28

Peter Brophy

The Quality of Libraries

1. Introduction In a world where the Internet, with its Yahoos and Googles, seems to provide more than enough information for anyone to absorb, there are those who question the future of library services as we have known them. For this reason if for no other, the ability to provide convincing evidence of the benefits of library services is a critical issue, and it has been much discussed in our professional literature. Such debates reflect a concern that the services for which we as librarians are responsible should be the best that it is possible to deliver, and should be clearly focused on the needs of the users we serve. This concern finds expression in efforts to establish methodologies which will enable the quality of individual services to be assessed, and which will assist in their continuous improvement. As well as focusing on the library as a service, it is common to hear professionals speak of "quality assured" information resources, most often though not exclusively in the context of online services. This is a way of reflecting that the content of our libraries is not a random sub-set of the universe of published information but is a carefully selected representation of the best of humankind's recorded knowledge. Over many years librarians have developed sophisticated and complex selection mechanisms, taking note of such matters as peer review (largely for journal articles), publisher reputation (for monographs) and so on, and relating these to the known interests of the user community, present and future. "Quality" is thus a prominent concept in professional debates. The aim of this article is to reflect on the work that has been undertaken in this field and to offer some observations on possible future directions. It explores different understandings of "quality" and relates these to the mainstream of management approaches. There are three foci for this discussion: the use of performance indicators, which also provides an historical perspective; the "quality attributes" approach pioneered by Garvin; and methodologies which seek to explore the gaps between user expectations and perceived performance, based upon SERVQUAL. Finally, a number of techniques which aim to provide a synthesis of methodologies, in order to provide a rounded picture of library performance, are noted. Throughout there is a concern to focus on the question of impact, to ask, "Do libraries change lives?" 2. Quality In the general management literature, the classic definitions of "quality" emerged from the work of a number of researchers and writers, known collectively as the "quality gurus" (see Brophy and Coulling (1996, Chapter 2)). They developed a series of statements about and definitions of the concept of quality, of which the most

30

The Quality of Libraries commonly quoted are "quality is fitness for purpose" and "quality is conformance to requirements". These insights created something of a sea-change in attitudes to managing quality, sometimes assisted by an arresting or even controversial style of presentation. So, for example, Philip Crosby was fond of stating that "there is no such thing as a quality problem". What he meant was that "so-called quality problems are always created by bad management, and can be rectified by management. They do not exist by themselves. Attend to the management issue and the so-called "quality problem" will be resolved." (Brophy and Coulling, 1996) Later commentators made similar points, yet always with a focus on the needs of customers. So Peters and Waterman (1982), to take a typical example, wrote "Remain close to your customers. Find out what the customer wants and likes and concentrate on providing that". It is interesting to consider these ideas alongside the now classic "Laws of Library Science" formulated by S.R. Ranganathan and published in 1931: • • • • •

Books are for use Every reader his book Every book its reader Save the time of the reader A library is a growing organism. (Ranganathan, 1931)

What is of importance, of course, is the emphasis that these "Laws" gave to the perspective of the user of library services and materials, and it is here that the quality management literature and that of librarianship meet and intersect. As noted elsewhere, "while some of these statements have dated - the last in particular must be questionable (unless it is taken to refer to the World Wide Web!) - it is interesting to note that the emphasis in these "Laws" is very much on the library user rather than the collection itself and very much about access." (Brophy, 2001) It was in the 1990s that interest in quality management among librarians reached its peak. In a number of countries, researchers and practitioners undertook a variety of investigations and studies to relate the more general quality management approaches and practices to libraries. This interest was epitomised by a Total Quality Management conference held as part of the Library Technology Fair at Hatfield, U.K., in 1993. Authors from the UK Institute of Management published a guide to ISO 9000, the international quality management standard, in the same year (Ellis and Norton, 1994). In the USA Hernon and Altman (1998) and others explored the application of quality management approaches. Many other examples could be cited. The concept of quality thus came to considerable prominence in the thinking of librarians, and this interest has continued. Before considering these developments further, however, it is useful to look at other ways in which the performance of libraries has been assessed and to consider how this relates to the management of quality. 3. Performance Measurement The measurement of the performance of systems goes back at least as far as the invention of manufacturing technologies at the start of the Industrial Revolution 31

Peter Brophy indeed it could be argued that it goes back to the earliest human societies, where no doubt the performance of individual hunter-gatherers was watched closely by the group! However, systematic approaches are most usefully traced to the development and acceptance of Frederick W. Taylor's management theories (and specifically the 1909 publication of his Principles of Scientific Management), which led to an emphasis on inspection and control. Although he built on the work of earlier theoreticians and practitioners, it was Taylor's emphasis on practical application, and on the need for managers to design work, that led to widespread acceptance of his methods. He was able to point to considerable successes in the ways he employed his theories. For example, he systematically examined the process of shovelling coal at the Bethlehem steelworks and was able to redesign the shovels used for each grade. As a result the workforce of "coal shovellers" was cut from 500 to 140. Other managers and owners were quick to take notice of his revolutionary methods and to apply them to their own industries. It was only gradually that the limitations of this highly mechanistic approach, which in effect treated human beings as machines, came to be recognised. In 1931 - the same year that Ranganathan's "Laws" appeared - W.A. Shewhart published Economic Control of Quality of Manufactured Products (Shewhart, 1931). This contribution marked a shift towards the use of statistical methods, and from these beginnings statistical quality control developed. While the emphasis still remained on the productivity of the individual worker, more attention started to be paid to scientific approaches to management. A number of different schools of thought developed, most notably operational (or operations) research which encouraged the formation of multi-disciplinary teams to bring a range of techniques to bear on complex organisational problems. Linear, and later dynamic, programming was introduced to enable complex management and organisational problems to be tackled. For a time, organization & methods (O&M) and work study became fashionable. More recently, human relations approaches have become more prominent, recognising that full participation by all employees, in decision making as well as in production, produces significant benefits. Currently the emphasis is on holistic approaches which involve a focus on benefits as well as costs and on the interests of all stakeholders, both internal and external. However, running through all of these approaches are two threads: an enduring search for "quality", which we consider below, and a need for robust and rigorous performance indicators. The nature of these indicators can be seen to mirror the dominant management theories of the day. Libraries have followed these trends closely. Thus the earliest indicators used by libraries tended to refer to little more than the inputs to the system and a few rather crude outputs. For example, library annual reports would record the amounts spent, the numbers of books added to stock, the number of employees, and so on - all input measures. They would also record the number of books issued and perhaps the number of visitors to the library straightforward outputs. Calculating ratios and trends over time provided useful information for policy makers and managers, while the use of such statistics at the micro level provided data on the basis of which such matters as staffing levels in different departments could be assessed.

32

The Quality of Libraries Increasingly sophisticated indicators and frameworks were developed by, inter alia, Van House et al. (1990) and King Research (1990), while in 1995 the European Commission funded a toolkit of library performance indicators and techniques (Ward et al., 1995) and the UK's SCONUL produced it's The Effective Academic Library (Higher Education Funding Council for England, 1995). Similarly, other professional bodies such as the Association of Research Libraries (ARL) and the International Federation of Library Associations and Institutions (IFLA) established working groups to develop systematic approaches, while the International Standards Organisation was persuaded to support the development of an international standard for library performance indicators (ISO 11620, 1998). Indicators for electronic libraries were developed in the late 1990s, notably in the cross-European EQUINOX project ( - see also Brophy and Clarke (2001)) and in the USA by McClure and Bertot (e.g. Bertot, 2001). As part of this work, a major concern which has emerged in the general management literature in recent years started to be mirrored in librarianship. This is that overemphasis on inputs and outputs, and thus on efficiency, leads to loss of broader vision and the ignoring of opportunities both to innovate or simply to serve customers better. The argument, put forward strongly by such management "gurus" as Minzberg (1982) and Peters and Waterman (1982), is that there is a pressing need to move the focus onto outcomes and impacts - what is the effect on the customer of all these outputs? Does the customer receive and perceive real benefit from the service? What is the impact of the service? In recent years such questions have been addressed by an increasing number of authors. Saracevic and Kantor (1997) provided a framework for studying the value that users experience as a result of library and information service use. More recently, the ARL has undertaken a "New Measures Initiative" (Association of Research Libraries, 2000; DeWitt, 2001). Hernon (2002), criticising current measures of outcomes and impact, writes that the key question is, "how did they (the library users) change their behaviour as a result of the service?". Brophy (2002) has suggested that impact itself needs to be thought of as a multilayered concept - and that not all impacts are positive: "One way to approach the concept is to think of "levels" of impact. For example a library service may result in any of the following impacts: •

•

• •

Hostility: a user may be so disappointed with the service that he or she decides that it is a total waste of money. Perhaps the result is a letter of condemnation to an influential third party such as a councillor. Hopefully, such impacts are very rare. Dismissive: the user is not actively hostile, but simply feels that the service is not worthwhile. It is a waste of personal effort to get involved, even if no attempt is made to undermine the service. None: the user has neither positive nor negative feelings or views about the service. It is almost as if it did not exist. Awareness raised: here the service has just about had a positive impact, but simply in terms of the user being made aware of something which he/she

33

Peter Brophy

•

• • •

•

was not aware of before. They know the service exists, do not dismiss it out of hand and might turn to it in the future if they feel a need. Better Informed: as a result of coming into contact with the service the user has better information than before. This information may have been memorised or recorded for future use. Improved knowledge: the information obtained has been considered and the user is now more knowledgeable about the subject. Changed perception: the knowledge gained has resulted in a change to the way that the user looks at a subject. Real learning has taken place. Changed world view: here the user has been transformed by the service. His or her view of the world has shifted significantly, and constructive learning has taken place which will have long term effects. Changed action: the new world view has led to the user acting in a way he or she would not have done before. Learning has turned into action, so that the encounter with the service has changed not just that user, but - in some way the broader world.

The key question, of course, is how we might measure levels of impact experienced by users systematically and reliably. We return to this later. 4. Quality Attributes The quality management movement would seem to have much in common with this shift in usage of performance indicators towards measures of outcome and impact. The classic definitions of quality, concerned with "fitness for the user's purpose", clearly relate to the benefits that customers perceive as arising from a service or product which they purchase or use. Recognising that these classic definitions are somewhat simplistic, a more sophisticated approach was developed by David Garvin in the USA in the 1980s. He recognised that "quality is a complex and multifaceted concept" and suggested that there are eight critical dimensions or attributes that can be used as a framework for determining the overall quality of a product or service (Garvin, 1987 - see below). Garvin suggested that many of the problems of defining and recognising quality arise because the concept can be approached from many different perspectives. He suggested that at least five views can be identified in the literature and in practice: • • • • •

the transcendental view: quality can be recognised, but cannot be defined; the customer view: quality as fitness for the customer's purposes or conformance to the customer's requirements; the manufacturer view: quality as conformance to specification; the product view: quality is related to inherent characteristics of the product; the value-based view: quality is dependent on what a customer is willing to pay for it. (Garvin, 1984)

34

The Quality of Libraries While Garvin's quality attributes approach was originally intended mainly for manufacturing industries it has subsequently been adapted for use in libraries and information services by Marchand (1990) and by Brophy and Griffiths (Brophy (1998); Griffiths and Brophy (2002), Griffiths (2003)). The latter team suggested that library and information services might be assessed on the basis of ten quality attributes. This approach may also be contrasted with the suggestion of Abels, White and Hahn (1997) concerning the quality attributes applicable to web pages (see also Madu and Madu, 2002). Table 1 below contrasts Garvin's original attributes with these later adaptations. GARVIN

BROPHY and GRIFFITHS

ABELS et al.

Performance, the primary purpose of the product or service and how well it is achieving that primary purpose.

Performance, concerned with establishing confirmation that a library service meets its most basic purpose, such as making key information sources available on demand.

Performance based on use, including ease of use, and content.

Features, secondary characteristics which add to the service or product without being of its essence.

Features: aspects of the service which appeal to users but are beyond the essential core performance attributes, such as alerting services.

Features such as links to other sites which might better answer a particular question.

Reliability, the consistency of the product or service's performance in use.

Reliability, which for information services would include availability of the service. Such problems as broken Web links, lack of reliability and slowness in speed of response would be measured as part of this attribute.

Reliability, including both availability and currency/accuracy of information provided.

35

Peter Brophy GARVIN

BROPHY and GRIFFITHS

ABELS et al.

Conformance, whether or not the product or service meets the agreed standard, which may be internally or externally generated.

Conformance: whether the service meets the agreed standard, including conformance questions around the utilisation of standards and protocols such as XML, RDF, Dublin Core, OAI, Z39.50 etc.

(Not defined)

Durability, the amount of use the product or service can provide before it deteriorates to a point where it needs replacement.

Durability, related to the sustainability of the information or library service over a period of time.

(Not defined)

Currency of information, that is, how up to date the information provided is when it is retrieved.

(Treated as part of "Reliability")

Serviceability, how easy it is to repair a product or correct a service when it goes wrong, including the level of inconvenience experienced by the customer.

Serviceability, which may translate to the level of help available to users during, for example, information retrieval, or otherwise at the point of need. The availability of instructions and prompts throughout an online service, context sensitive help and the usefulness of that help could be measured in order to assess performance under this attribute.

Serviceability concerned with the handling of complaints and conflicts, with the aim of creating a happy and satisfied customer.

Aesthetics, the appearance of the product or service.

Aesthetics and Image, both of the physical library and of web-based services based upon it.

Aesthetics, concerned with visual attractiveness

36

The Quality of Libraries GARVIN

BROPHY and GRIFFITHS

ABELS et al.

Perceived quality, in essence the reputation of the product or service among the population, especially those with whom the potential customer comes into contact.

Perceived Oualitv: the user's view of the service as a whole and the information retrieved from it. It may be useful to measure perceptions both before and after a service is used.

Reputation, related to past experiences of the site.

Usability, which is particularly relevant to electronic services and includes issues of accessibility.

Structure, which is concerned with how information is structured within the web site's presentation. Storage capability, which is concerned with whether all required information can be stored in order to answer queries which may, for example, require an historical analysis. Security and system integrity, including the handling of payment (e.g. credit card) data. Trust, whether users are wiling to disclose personal information. Closely linked to "Security and system integrity". Responsiveness, which includes courtesy and willingness to be flexible (for example with a cancelled order).

37

Peter Brophy GARVIN

BROPHY and GRIFFITHS

ABELS et al.

Product/service differentiation and customization, which asks what is unique about this particular web site, not least to differentiate it from its competitors. Web store policies, which relates to the customerorientation of policies and might involve a comparison with a high street store. Assurance, concerned with the creation of good customer experiences through the knowledgeability and courtesy of staff. Empathv, which mav be expressed through the availability of individualised personal attention. Table 1: Garvin's Quality Attributes and their adaptation for library and web services 5. User Expectations and User Experience Three American researchers, Zeithaml, Parasuraman and Berry (1990) worked with customer focus groups in each of four sectors (retail banking, credit cards, securities brokerage and product repair and maintenance) using market research based methodologies to establish the criteria used by customers in assessing the quality of services. Their work is interesting precisely because it deliberately focussed on how customers perceive quality in services rather than in products, and thus has considerable relevance to libraries. It is worth noting, incidentally, that this team of researchers observed that customers find it more difficult to assess the quality of services than the quality of products. By analysing the data from their focus groups, Zeithaml, Parasuraman and Berry were able to identify ten "dimensions" of service quality, common across all of the services examined, and closely related to Garvin's 38

The Quality of Libraries analysis and the later adaptations described above. These dimensions were (adapted from Brophy, 1996): 1.

Tangibles: is the service an attractive place to visit? Are the staff appropriately dressed? Do they use modern, up to date equipment?

2.

Reliability: is my telephone call returned when the receptionist said it would be? Is my bank statement free of errors? Does the washing machine work when it has been repaired?

3.

Responsiveness: when a problem occurs, is it quickly put right? Is the company willing to arrange to repair the washing machine at a time to suit me?

4.

Competence: do front-line staff give the impression of knowing what they are doing? Similarly, does a repairer appear to know how to diagnose a fault and carry out a repair with confidence?

5.

Courtesy: are staff pleasant, even when asked difficult (or what may appear to be ridiculous) questions? Does the repairer wipe his or her shoes rather than trample mud all over my hallway carpet? Do staff manage not to appear busy even when they are - so that I gain their attention when I need it?

6.

Credibility: Does the service enjoy a good reputation - do other people speak well of it? Do I get a credible and worthwhile guarantee with a repair, such that I can have confidence that any problems will be put right quickly and without further expense?

7.

Security: Is it safe to use the service? For example, is my credit card safe from unauthorised use? Do I have confidence that the repair was properly carried out to an acceptable standard?

8.

Access: If I have a problem, can I get access to a senior member of staff to help me resolve the cause? Does the company answer the telephone when I ring? Is it easy to find the repair company's premises?

9.

Communication: Is the service explained clearly and the options outlined comprehensively? Do the staff avoid using unnecessary jargon? Do they listen to me? If something unexpected occurs and the repair company cannot keep the appointment that they have made, do they contact me in good time to rearrange it?

10. Understanding the customer: If I am a regular customer, does someone on the staff recognise me? Do they try to understand my individual needs? Do they try to arrange the repair visit to meet my convenience rather than their own?

39

Peter Brophy Zeithami, Parasuraman and Berry went on to refine their analysis and published a set of five key issues, which have become known as the "Rater" set from the initial letters of each of the headings: 1.

Reliability

2.

Assurance

3.

Tangibles

4.

Empathy

5.

Responsiveness.

Perhaps the most critical insight of these researchers was that methodologically quality may be measured by assessing the differences or "gaps" between customer expectations and customer experience. They developed this observation into a widely adopted methodology called SERVQUAL (Parasuraman, Berry and Zeithaml, 1988). In recent years the Association of Research Libraries (ARL) has adapted this approach for use in libraries. LibQUAL+ is described by ARL as "a suite of services that libraries use to solicit, track, understand, and act upon users' opinions of service quality" (see ). Firmly based on the SERVQUAL methodology described above, LibQUAL+ has been adopted by a considerable number of academic libraries in the USA and further afield, including in the UK and France. The methodology has been described by Cook (2001). Blixrud (2002) reports that over 78,000 individuals were surveyed by LibQUAL+ partners during 2002, producing initial findings that the dimensions of service that make up a library user's perception of service quality include: • • • •

"Service affect; i.e., responsiveness, assurance, empathy, and reliability - the human dimensions of library service Library as place; i.e., campus center of intellectual life, but may not be a concern if the physical facilities are adequate Personal control; i.e., ability to navigate both the information universe in general and the web in particular Information access; i.e., ubiquity of access meaning information delivered in the format, location, and time of choice and comprehensive collections."

These generalised conclusions will be particularly valuable in moving forward the development of methodologies to capture users' views of the quality of the library services they use. 6. Synthesis It is readily apparent that the major challenge for library managers is to draw together these various strands of work into robust, economic and above all practical frameworks. Happily, there is evidence that this is happening. Librarians are showing 40

The Quality of Libraries considerable interest in such techniques as Benchmarking, the Balanced Scorecard and Storyboards. We will examine each briefly. 6.1 Benchmarking The idea behind benchmarking is to formalise comparisons with other organisations which have something in common with one's own. Benchmarking is concerned with developing systematic and structured approaches to finding and implementing best practice. It links the identification of the best way of doing things in the sector (or even outwith the sector if useful examples can be found) with a determination to improve one's own organization and "to be the best". Benchmarking can be internal or external. That is, it can be done within the organization, maybe across divisions or sections or maybe just by comparing new and old ways of doing things. Or it can be done externally, by comparing the organisation with others. Some sectors, including libraries in some countries, have developed their own "benchmarking clubs", where organisations agree to work together to share benchmarking methods and data. While benchmarking may be more or less formal, it is often associated with iterative processes that draw in a range of both quantitative and qualitative methodologies. Rather than focusing on a particular technique or on the whole organisation, it seeks to use appropriate methodologies to explore particular issues in depth - before moving on to the next issue. It therefore encourages managers to range across the available techniques to select those which best suit a particular purpose, to undertake that analysis, determine and implement action, review - and move on. There is now considerable experience of benchmarking in the library sector. Creaser (2003) notes that the following benefits were observed from one academic library benchmarking exercise in the UK: • • • • • • • • • • •

"Establishment of best practice Process improvements (large and small) Continued evaluation of customer opinion and needs Identification and validation of clear trends Networking between groups of staff involved in similar operations (quoted by many as "invaluable") Exchange of views (and the generation of new ideas!) Staff development including learning new analysis methods such as process improvements and activity based costing (ABC). Greater local ownership of processes and tasks Changing perspectives about individual roles and also overall purpose (both for library and academic staff and students) University recognition by establishing a solid reputation for the active use and Implementation of benchmarking performance measurement methods."

41

Peter Brophy 6.2 The Balanced Scorecard The Balanced Scorecard derives from the work of Kaplan and Norton (1992) in the USA but has been developed for use in the library sector by Poll in Germany (Poll, 2001), in Australia by Cribb and Hogan (2003), among others, and by a number of practitioners in the USA (e.g. Self (2003)), where the Library of Congress even has a video presentation available (see ). As Poll remarks, "service quality has many aspects - the Balanced Scorecard attempts to integrate them." The particular value of the Balanced Scorecard is that it ensures that the library is assessed from a range of different standpoints: that of the user; that of resource utilisation; that of internal processes; and that of learning and growth of the organisation. These four perspectives are centred around clarity of vision, of values and of strategy. In the Balanced Scorecard methodology, there is particular emphasis on the "double feedback loop". This means that not only does feedback relate to the organization's outputs, but it also relates to the outcomes and impacts. The Balanced Scorecard helps an organisation to check whether improvements in one area may have been achieved at the expense of another, and so keeps the whole organisation in balance. Its underlying philosophy is very different to that of earlier control-based systems. As Kaplan and Norton (1992) noted: "...because traditional measurement systems have sprung from the finance function, the systems have a control bias. That is, traditional performance measurement systems specify the particular actions they want employees to take and then measure to see whether the employees have in fact taken those actions. In that way, the systems try to control behaviour. Such measurement systems fit with the engineering mentality of the Industrial Age. The Balanced Scorecard, on the other hand ... puts strategy and vision, not control, at the centre. It establishes goals but assumes that people will adopt whatever behaviour and take whatever actions are necessary to arrive at those goals." 6.3 Storyboards Although not widely used in the library and information sector, storyboards are an attempt to capture the qualitative evidence for service performance. In essence they attempt to capture the "story", very often of an individual user, which illuminates the effects that the service has had. They thus bring us back to the question of impact: it may be impossible to quantify the impacts of libraries in a scientific way, but what we can do is to record incidents in people's lives which, taken together, indicate the kinds of impact which are being felt. Two short examples from a study of The People's Network in the UK illustrate the approach (Brophy, 2003): "I obtained a place at College on their basic ECDL (European Computer Driving Licence) course and used the library to practice some of the things I learned at the college. If it hadn't been for the 42

The Quality of Libraries initial use of the library computer I don't think I would have considered the college course ...I now have a new job in which I need computer skills. So from playing about with the library computer I now have a successful career, all because computers were installed in local libraries." "(One user) is 72 and until recently has never clicked a mouse, seen a floppy disk or knew what the Internet was. Each week she brings her folder, floppy disk and walking stick to the library to enhance her growing knowledge and skill using the computer. The Computer Buddies Scheme has turned her quiet life around and after 4 months she is now adept at using the computer. " Storyboards focus very much on outcomes and help to illuminate what can sometimes be rather arid statistical descriptions. They cannot, however, be used on their own. 7. Conclusions Clearly, the assessment of library quality has come a long way in recent years. On the one hand we now have robust sets of performance indicators which provide the basic "picture" of library performance. Beyond that, researchers and practitioners have developed ways to explore the customer experience, to find out whether users and non-users have positive or negative perceptions of library services, and to use these insights to better manage services. The application of performance measurement and quality management approaches to libraries has clearly paid many dividends. But perhaps we are now seeing a movement beyond individual techniques in an attempt to draw on the best of what is now many years' experience and to synthesise the different approaches. As this becomes more commonplace, we will undoubtedly find ourselves looking again at the question of impact. Do our libraries actually do any good? We now have an increasing amount of evidence to support this assertion (see, for example, Poll (2003)). The question which remains is, how can we build on the achievements of those who have for years laid the foundations for, in Richard Orr's famous phrase, measuring the "goodness" of library services? (Orr, 1973) Inspired by the successes of the past, we must continue to ask, "Do libraries change lives?" and we must continue to seek the evidence that, yes, indeed they do. References Abels, E.G., White, M.D. and Hahn, Κ. "Identifying user-based criteria for Web pages." Internet Research: Electronic Networking Applications and Policy, 7,4 (1997), 252-62. Association of Research Libraries. ARL New , 2000.

43

Measures

Initiative.

Peter Brophy Bertot, J. "Measuring service quality in the networked environment: approaches and considerations." Library Trends, 49,4 (2001), 758-775. Blixrud, J.C. "Evaluating library service quality: use of LibQUAL+™." IATUL Proceedings (new series) 12, 23rd Annual IATUL Meeting, June 2-6, 2002, Kansas City, KS, USA. Brophy, P. and Coulling, K. R. Quality Management for Information and Library Managers, Aldershot: Gower, 1996. Brophy, P. "It may be electronic but is it any good? Measuring the performance of electronic services." Robots to Knowbots: the wider automation agenda. Proceedings of the Victorian Association for Library Automation 9th Biennial Conference, January 28-30 1998. Melbourne, Australia: VALA, 1998, 217-230. Brophy, P. The Library in the Twenty-First Century. London: Library Association Publishing, 2001. Brophy, P. The evaluation of public library online services: measuring impact. London: Resource, 2002. Brophy, P. The People's Network: a turning point for public libraries: first findings. London: Resource, 2003. Brophy, P. and Clarke, Z. EQUINOX: Library Performance Measurement and Quality Management System: Telematics for Libraries Programme: Project LB5634/A EQUINOX 24471/0: Deliverable D2.5: Edited Final Report. 2001

Cook, C. (ed.) 'The maturation of assessment in academic libraries: the role of LibQUAL+™." Performance Measurement and Metrics, 3,2 (2001), 34-112. Creaser, C. (ed.) As others see us: benchmarking in practice. LISU Occasional Paper 33. Loughborough: Library & Information Statistics Unit, 2003. Cribb, G. and Hogan, C. Balanced scorecard: linking strategic planning measurement and communication, 2003

to

DeWitt, D. (ed.) Evaluating the Twenty-First Century Library: The Association of Research Libraries New Measures Initiative, 1997-2001. New York: Ha worth Press. 2001.

44

The Quality of Libraries Ellis, D. and Norton, B. Implementing BS5750 /ISO 9000 in libraries. London: Aslib, 1993 Garvin, D. A. "What Does "Product Quality" Really Mean?" Sloan Review, 1984, 25-45.

Management

Garvin, D. A. "Competing on the Eight Dimensions of Quality." Harvard Business Review, November-December, 1987, 108-109. Griffiths, J.R. Evaluation of the JISC Information Environment: student perceptions of services. Information Research, 2003. Griffiths, J.R. and Brophy, P. "Student searching behaviour in the JISC Information Environment." Ariadne, 33, 2002.

Heath, F. et al. "Emerging Tools for Evaluating Digital Library Services: Conceptual Adaptations of LibQUAL+™ and CAPM." Journal of Digital Information, 4,2 (2003). Hernon, P. "Quality: new directions in the research." The Journal of Academic Librarianship, 28,4 (2002), 224-231. Hernon, P. and Altman, E. Assessing service quality: satisfying the expectations of library customers. Chicago: American Library Association, 1998. Higher Education Funding Council for England. The effective academic library: a framework for evaluating the performance of UK academic libraries: a consultative report to HEFC(E), SHEFC, HEFC(W) and DENI by the Joint Funding Council's Ad Hoc Group on performance indicators for libraries. Bristol: HEFC(E), 1995. Kaplan, R. S. and Norton, D. P. "The Balanced Scorecard: measures that drive performance." Harvard Business Review, 1992, 70-79. International Organization for Standardisation (ISO). ISO 11620: Information and documentation - library performance indicators. ISO, 1988. King Research Ltd. Keys to success: performance indicators for public London: HMSO, 1990.

libraries.

Marchand, D. "Managing information quality." In: Information quality: definitions and dimensions (ed. I. Wormell) London: Taylor Graham, 1990, 7-17. Minzberg, Η. Structures in fives: designing effective organizations. Englewood Cliffs: Prentice-Hall, 1983.

45

Peter Brophy Orr, R.H. "Measuring the goodness of library services: a general framework for considering quantitative measures." Journal of Documentation, 29,3 (1973), 313-332. Parasuraman, Α., Berry, L. L. and Zeithami, V. A. "SERVQUAL: A multiple-scale for measuring customer perceptions of service quality." Journal of Retailing, 64 (1988), 12-40. Peters, T. and Waterman, R. J. In search of excellence: lessons from America's best run companies. New York: Warner Bros., 1982. Poll, R. "Performance, processes and costs: managing service quality with the Balanced Scorecard." Library Trends, 49,4 (2001), 709-717. Poll, R. "Measuring impact and outcome of libraries." Performance Measurement and Metrics, 4,1 (2003), 5-12. Ranganathan, S. R. Five Laws of Library Science. Madras, India: Madras Library Association, 1931 Self, J. "From values to metrics: implementation of the balanced scorecard at a university library." Performance Measurement and Metrics, 4,2 (2003), 57-63. Shewhart, W.A. Economic control of quality of manufactured product. Macmillan, 1931.

London:

Taylor, F.W. Principles of Scientific Management. New York, Harper Bros., 1909 Van House, Ν. Α., Weil, Β. T. and McClure, C. E. Measuring academic library performance: a practical approach. London and Chicago: American Library Association, 1990 Ward, S. et al. Library performance indicators and library management Luxembourg; European Commission DGXIII-E3. [EUR 16483 EN], 1995.

tools.

Zeithami, V. Α., Parasuraman, A. and Berry, L. L. Delivering quality service: balancing customer perceptions and expectations. London: Collier Macmillan, 1990.

46

Maurice Β. Line

An Agenda for Overdue Change

1. Introduction I am pleased to make this small contribution to the festschrift for Roswitha Poll, who has herself made significant contributions to our discipline. In this essay I summarise ideas and proposals that I have put forward over the years with a view to advancing practice and serving people better. There is little in it that I have not written before; but it has never been assembled together until now. I can guess at the likely reaction to this paper: that the proposals in it are unrealistic and will never be acted upon, not because they cannot be acted upon, but because they are Utopian - they would cost too much and there are too many vested interests in the way. To which I would reply that ideas that were once believed impossible have frequently become reality, and that in any case it is useful to set out ideals as goals to be worked towards, even if they can never be entirely attained. As our poet Robert Browning said, "A man's reach must exceed his grasp". The objectives of the IFLA/UNESCO programmes of Universal Bibliographic Control and Universal Availability of Publications were never fully attainable; but they inspired many countries to make dramatic improvements in control and availability, and helped to change thinking. 2. Thinking in context and for the future This leads me into the first section. The thinking of librarians is commonly constrained by the day-to-day reality that is in front of them. They tend not to see what they are doing in its context, and when they do the context is limited to their immediate environment. Many academic librarians would prefer to keep on doing what they are doing without interference from their parent institutions, but they cannot easily do that. They can and do often carry on without awareness of wider trends in the academic world. In all the large and increasing volume of literature on library and information science, how many papers are there on developments in higher education and the implications for the provision of information, in libraries or by other means? Yet higher education seems likely to experience massive changes, the chief of which is a switch from teaching to learning as the central feature. The implications of this for libraries, or whatever their successors come to be called, are profound, since if the emphasis is on learning the provision of all kinds of learning materials is crucial. Teachers may become largely mentors to self-learning students. Also, the importance of individual universities as institutions is already declining, as it becomes possible to take courses put on by several universities at once. Students may choose which course they want, and their home university may play a minority part in provision. How fast

47

Maurice Β. Line these changes occur, and how far they go, is open to discussion; my point is that librarians do not seem to be discussing the issues. Similar points might be made about public libraries and librarians. Almost every country is undergoing huge socio-economic changes, as a result partly of extensive immigration, which looks certain to increase, whatever governments do to try to stop it, but mainly as a result of worldwide political change, of which mass migration is only one part. Similar changes are taking place within countries, as the more developed countries lose more and more of their manufacturing to the developing world, leading to large internal population movements; as people live longer, and the dependent sector of the population begins to outnumber the productive sector; as higher percentages of the population are better educated and more wealthy; as more and more people have computers and access to the Internet; as demands for regional devolution eventually become irresistible; and so on, and so on. All of these trends have implications for public libraries. So why are the voices of public librarians silent on these matters? The reasons are the same as those that affect lawyers and theologians, and doubtless many other professions. It is much easier to spend mental effort on dis-cussing the minutiae of the law, or precise details of scriptural interpretation, than it is to consider the end purposes of law and theology: respectively justice and an under-standing of the divine. Likewise, the repeated examination of detailed cataloguing rules, and more recently ways of constructing online catalogues, make easier subjects to think and write about than the bigger issues of which examples were given above. I call this the lamp-post syndrome; absorption in better ways of painting lamp-posts than in improving the light the lamps shed. I am not questioning the necessity for cataloguing rules or for lamp-post paint; what does distress me is the attention given to them while the big issues are neglected, and the obsession with means rather than ends. So my plea is for greatly improved vision: wide-angle lenses to see the broader context in which libraries exist; and telephoto lenses to see further ahead. The further ahead one looks, the mistier the view, and mistakes will be made; but the very act of looking ahead, if it becomes a regular discipline, is valuable in itself. 3. Bibliographic control and access Great advances have been made in the last 25 years in bibliographic control. Almost all of these are due to the use of computers, which have made it far easier to compile and combine records and make them more accessible. As a result, a higher and higher proportion of the documents published all over the world and contained in libraries are recorded in more or less standard form and are available online to more and more people. So what is wrong? Are we not on the right track to continued improvement? Well, yes, if improvement is defined narrowly. For the advances have nearly all been in increased coverage of monographs as single items. Let us take a first look at what national bibliographies do not cover: • Individual essays in collected volumes • Chapters in one- or two-author books, which often deal with a variety of topics that the subject entries allocated to the books do not cover • Most scientific and technical reports. 48

An Agenda for Overdue Change It is easy to see how this has happened. Librarians tend to think of books as unitary physical objects rather than containers of information. Each object (=book) has to be clearly designated for administrative purposes, just as different types and sizes of screw have to be described in a screw factory. Between one and three subject headings are added to each description almost as an afterthought. So what the poor user gets is crude and inadequate subject access to that portion of the library's stock that consists of books. Certainly "main entries" are needed. But I imagine that what most users want when searching for themselves is good subject access. I cannot imagine any of them being interested in national bibliographies as such. I for one am not in the least interested in what is published in the UK when I am in a library (though I may be so in a bookshop). What I want is a list of material on subjects that I am interested in, in languages I can read, whether the material is a book, a journal article, a report or whatever. The only real useful purpose of national bibliographies is as files of records that contribute to a comprehensive worldwide collection of records. This means that ideally I should have integrated access to book catalogues, journal indexes and other bibliographic files as if they were one. The files themselves would obviously not be integrated, but there are front-end systems that could make them searchable as one; some such already exist. To return to subject access, to the contents of books. 1 have always been puzzled by the fact that an article on one small aspect of chemistry may have three or four index entries in Chemical Abstracts, while a 400-page book which deals with many topics has no more (and usually fewer) than this in national bibliographies. This is clearly absurd. How could it be remedied without great expense? First, chapter headings could be included in the basic records, to be searched by keyword. Admittedly, some headings yield little information, but at the least they would give some additional subject access to many books. Ideally, however, access to the indexes of the books is needed. This could be provided by putting the indexes in supporting files, which could be called up as desired. The search procedure would be as follows: •

• •

User conducts a first search, using the classification or keywords in titles, and finds several books that look as if they hold information on the topic of interest. User conducts second search on chapter headings of the selected books; this refines the search, and reduces the number of likely books. User calls up the indexes of these books, one by one, and looks for relevant entries.

The indexes would be stored as digitised files, or possibly on microfilm. No attempt would be made to merge indexes; the variations in indexing style of books would make the resulting files impossible to use, as well as enormous. The technology for such a system is already here; the investment would obviously be high, but is it really out of reach? One barrier to the above system is that the normal approach to the files would be in English. This would not be a difficulty with the final stage of searching, book indexes, since users would not consult indexes to books in languages they could not 49

Maurice Β. Line handle. It would however create problems for keyword indexes. Fortunately for the English-speaker, but less fortunately for everyone else, most of the large journal indexes are American or British. The only solution is to make more progress with multi-lingual thesauri and automatic translation. This reached a reasonable level some years ago but seems to have stalled, at least until very recently. It constitutes another item on my agenda. Whatever level it reaches, it will never be perfect, but it would be a lot better than what we have now, and it is stupid not to seek improvement because perfection is impossible; we do not run our lives on such a principle, so why should we apply it to indexing? (I am reminded of the story of the mathematician and the engineer in one comer of a room. They were told they could approach a pretty girl in the far corner, but each step must be half the length of the previous one. The mathematician, knowing he would never get there; did not move. The engineer, however, reckoned that he could get near enough for all practical purposes.) I suggested above that national bibliographies were of little use in themselves. I must now modify this statement, and say that we need a different kind of national bibliography, one that is truly national, in that it contains not material that happens to be published in the country but material that is relevant to the country: works by, for example, foreign scientists on topics of universal interest would be excluded, and instead there would be items that related to the history, literature, artefacts etc. of the particular country, wherever they were published. For this to be possible, co-operative effort by national bibliographic services would be necessary. Each country would be responsible for identifying items relevant to other countries and would send the equivalent records to the appropriate bibliographic service. The process should be easy, though the time and effort involved would be considerable. Some large countries would probably have to be excluded - or rather, countries about which a very great deal has been written: for example, France, Germany, Russia, the UK, the USA, etc. A pilot system could be tried out among a few countries in the first place, to identify the costs and problems. 4. Access to cultural artefacts Many museums are creating digital artefacts of objects in their collections, and most significant musical compositions already exist in digital form. Some of this material is accessible online, but separately from bibliographic files. It cannot be very difficult to link these files, so that, for example, a researcher into Mozart can quickly call up relevant compositions and portraits of his contemporaries. Experimental systems of this kind should soon be turned into operational systems that are open to any user. Nor need "culture" be restricted to artistic objects; an engineer should be able to call up structural models, and a sociologist survey data, relating them directly to bibliographic files. 5. Document provision and access Document access and supply constitute an activity that has seen great growth and even greater change over the last two decades. This activity is commonly known as "interlibrary lending" in the English-speaking world, although most items are now sent as copies or electronically, and many items are not supplied by libraries (or even 50

An Agenda for Overdue Change necessarily ordered by them); a much more correct umbrella term would be Remote Document Supply (RDS). Requesting is increasingly directly linked to bibliographic references, and users can obtain many articles direct from suppliers. So where are the problems? First, while the supply of journal articles has been transformed, obtaining books on loan - and I really mean "loan" here - is not much easier today than it was twenty years ago. The only real changes are that there are more union catalogues online, they cover more libraries, and requesting is mostly online. Locating copies of books within most countries is therefore a little faster and there is more chance of success - at least, there would be if budgetary reductions had not resulted in fewer books being acquired by libraries. I have no figures to prove this, but it would surprise me if the proportion of the world's book production that is acquired in total by the libraries of most developed countries (and every underdeveloped country) has not shrunk; the shrinkage will have been slow, but cumulatively it must be substantial. In the UK, coverage of foreign-language books has always been a relatively weak area, and it must have become weaker. We clearly need quick and easy access to resources in libraries in other countries. In Europe as a whole, a very high percentage of all wanted books must be available in one library or another. All that is needed is more comprehensive union catalogues in electronic form for each country in Europe, which can be searched as if they were one; catalogues in the user's own country would be searched first, because of the cost of sending books across international borders. In fact, that is not quite all that is needed. Many if not most libraries lend abroad on condition that the items are not used outside the borrowing library - presumably on the assumption that users in other countries are more likely to lose or abuse books, an assumption that has no evidence to support it as far as I know. In the case of rare or valuable books, this restriction is understandable, but that applies as much within countries as between them. It is true that postal services are not wholly reliable, but that is beyond the control of libraries. There is a further weakness that applies to all kinds of material requested. Libraries frequently give lower priority to users who happen not to live locally, even though these users are already suffering some delay by not having what they want in their own libraries. This habit has proved surprisingly hard to break; many libraries seem to leave requests for other libraries for several days until they mature, as if they were apples waiting to ripen. Nor is responsibility for unnecessary delay confined to supplying libraries; some libraries spend two or three days checking or otherwise processing requests from their own users. There are other categories of books that are very hard to supply. These include books that are recently published, and books in very heavy demand. Both categories are likely to be held in other libraries, but they are usually in demand there too. It is a waste of time trying to obtain such books by conventional loan methods. Purchase is preferable to lengthy waiting, and possibly little more expensive if all costs are taken into account. In the case of books wanted by large numbers of students for short periods, e-books may in due course provide a solution - but that again is not RDS. Another prospect that seems likely to become more common, and that might be used to satisfy a request from a specific user, is the book that is not economically 51

Maurice Β. Line viable as a normal print production but can be downloaded from a remote electronic version and bound at the receiving end. Such downloads will not be cheap, and would presumably be added to stock. (If users want to have such books to keep, they should have them downloaded through a bookshop.) Libraries should keep a watchful eye open for developments here so that they can be prepared. Finally, for the really valuable item - which might even be a manuscript - there is the possibility of reading it remotely by means of television and a remote page-turner. This would be an alternative to making a microfilm. Only significant research libraries would be equipped for this. These last two possibilities would extend RDS into areas where it does not at present operate. They are not urgent, but are included here as steps that could be taken to make RDS almost comprehensive in coverage. 6. Preserving form and content To the danger of natural disasters has been added in the last few years the risk of terrorism or other violent human action. The National and University Library in Sarajevo and the National Library in Baghdad are notable recent examples. Major libraries make attractive targets for terrorists, as their damage or destruction damages part of a nation's heritage and history. How to ensure that this does not happen must be a priority question. There is no simple or cheap answer. Digitisation is proceeding fast in many countries, possibly as fast as it can in view of the cost and logistics of the process. The cost is a continuing one, as the electronic versions have to be "refreshed" regularly. For this and other reasons, microfilm is still preferred by many conservation experts; after all, microfilm can always be fed into the computer, so that the flexibility of an electronic text is available. The question then is what measures can be taken to try to safeguard the national heritage. A distinction must be made between documents that are mainly of value or interest in themselves as physical objects - such as medieval illuminated manuscripts, original editions of famous printed works - and those whose main value lies in the information content. For some items this distinction will be hard to make, but for most it should be easy: the decisive question is "If the original were destroyed, would the loss be incalculable, assuming the information in it were preserved?" The first category of items that are intrinsically valuable, for which the answer is "Yes", should surely be stored in very secure buildings, as bomb-proof and secret as possible, with occasional outings for public exhibitions; facsimiles can be made of the most important or beautiful for consultation purposes. As for the other category, items of interest for their content, digital or microfilm versions, carefully stored but needing little space, would constitute an adequate back-up, and the original items can be housed in the ordinary library stock. Heritage items are scattered among various libraries, and digitisation, filming and secure storage are expensive. A programme such as the above needs therefore to be pursued at a national level, with shared facilities.

52

An Agenda for Overdue Change 7. Library buildings There are now numerous examples of good library buildings of all kinds, public, academic and national. A "good" building is one that is attractive, easy to use and physically accessible by the handicapped, is neither too hot or too cold, has comfortable chairs and desks, and provides a range of facilities such as a large number of computer workstations, scanners and user-operated photocopiers. Further, it should have areas for discussion as well as quiet areas, and these should offer refreshments. In many if not most countries public libraries are experiencing a decline in use as a result of the new ability of users to have direct online access to information resources, and they have been compared unfavourably with modern bookshops. It has been shown that this decline can be put in reverse by new buildings with the characteristics listed above. Alongside good new buildings can be seen new buildings that seem designed to reproduce old weaknesses as well as incorporating new ones that accord with concepts of ultra-modern architecture; they are often the products of arrogant architects or even megalomaniac politicians who believe that libraries should be monuments to themselves rather than buildings for use by human beings seeking information. One well-known major (and extremely costly) library manages to achieve a remarkable combination of ugliness and dysfunctionality. Change on this front may not be easy to achieve, since the views of those in authority are often given precedence over those of librarians. A well publicised travelling exhibition of dysfunctional libraries, alongside libraries that really work, might shame the perpetrators into reconsidering their ways. So might protests on the part of the public at any new building that does not provide what they want and need. A final thought on library buildings. Why are there not more examples of combined libraries and bookshops? They would operate separately, but would have shared facilities such as a coffee shop and recreational areas. It is known that most book borrowers are also book buyers, so both parties should benefit. 8. Conclusion Some of the proposals above are obvious, some less so. Some are radical, some require incremental change, which is in some cases already in progress. They are merely outlined here; each one could from the subject of a book, or at least a long journal article. What is clear is that individually they would, apart from the first proposal (thinking in context), take time, effort and money to implement, and that together they constitute a formidable agenda. To me they seem very much worth aiming at.

53

Geoffrey Ford and Suhair Abdinnour

Art or Science? Two Approaches to Materials Fund Allocation in Academic Libraries

1. Introduction One of the major tasks of all library directors is to allocate resources for the acquisition of library materials - books, serials, databases and so on. In academic environments, the degree of freedom that is allowed to the directors in this task is dependent on the skill with which they balance the competing claims from those most powerful of vested interests - the academic staff. It is a widely held view that a system based on a formula is the "fairest" approach. Discussion among librarians over many years has led to the conclusion that the academic staff are convinced: a. b.

that an ideal formula exists; that librarians know what this formula is;

c.

that librarians keep the formula secret for their own nefarious purposes.

Librarians know: a. b. c.

d.

that is no ideal formula; that there resource allocation is a political process; that it is nearly always possible to find a formula that theoretically satisfies local political pressures, and may be successfully implemented for a few years; that, in practice, the strict application of such a formula over a longer period of time will throw up anomalies so that it will nearly always prove unacceptable in some way.

A more extended treatment of the search for formulae, with examples of theoretical and practical applications of formulae may be found in Ford (1999). An informal survey of library directors in British university libraries in the summer of 2003 once again confirmed that some are using formulae, some are looking for formulae, and some have given up using formulae. Opinions ranged from We've developed a formula that works for us reasonably well to

54

Art or Science? My experience, gained the painful way over some years, is that formula based approaches simply do not work unless so many opt outs and exceptions are allowed that the formula becomes meaningless. It should be clear that the basic question posed by this paper is not answerable in simple terms. There is no room for an extended philosophical dissertation on the implied dichotomy as between art and science, "soft" and "hard" approaches, the use of judgement or numbers, even subjective versus objective. A case study of one British university illustrates two contrasting approaches to materials fund allocation. As it happens, these were tried in successive years, and this illustrates the need for library directors to be able to react to collective "mood swings" among their academic colleagues. 2. Background to the Case Study In the university in question, there were eight academic divisions, known as faculties. These were Arts, Education, Law, Social Sciences, Engineering, Mathematics, Medicine and Science. The first four constituted the "arts side", the latter the "science side". This informal division into two groups was intermittently important in the annual round of budget discussions and negotiations both within the university at large, and more specifically within the library. For many years the process of allocating funds had been automatic - each year, after top slicing funds to buy general works, and major interdisciplinary serials and databases, the remaining funds available were allocated to each faculty, usually being increased from the previous year by an amount equivalent to the expected rate of inflation. Separate funds were maintained for books and serials for each faculty, so the inevitable consequence, well known to librarians, was a drift of funds from the arts side to the science side, for two reasons: science relies more on serials, and scientific serials increase in price more than arts ones. The imbalance between the library funding for the various faculties was highlighted by a member of the Library Committee, which was formally responsible for fund allocation. The Committee included a representative of each faculty, and to some of these members of academic staff was delegated the task of finding an appropriate method of making allocations that would be seen to be more equitable. 3. The Artistic

Approach

3.1 Inventing the Rules Like many similar working parties before (and since), and with the aid of library staff who ferreted out the data, the group speculated on the relevance of many factors that might be considered to affect the allocation process; and how, if consensus could be achieved on these, they could or should be weighted to produce a formula. It was no surprise that these academic staff were keen to see the effect that incorporating any new variable would have on the pattern of allocations before they would commit themselves to admitting it to the set to be considered, and much time was spent on trial calculations. Eventually it was agreed that a small number only of factors could be considered. The factors chosen for the final short list were: 55

Geoffrey Ford and Suhair Abdinnour Fl: the size of the community served F2: the importance of the library for the study of the subject F3: the published output F4: price F5: breadth of subject F6: the importance of the subject to the university. Discussion then centred on how these could be made operational and then combined in an acceptable formula. The size of the community served by each faculty (Fl) was easily measured, but there were unresolved differences of opinion as to whether academic staff and research students should receive a higher weighting than undergraduates. It proved impossible to find any way of measuring the importance of the library to the subject (F2), nor of the breadth of the subject (F5). Although some data were available for the published output of books and serials (F3), these were incomplete or non-existent for foreign countries. Similar considerations applied to prices (F4). Further discussion on factors F1-F5 may be found in 4.3.1, below. The final factor, the importance of the subject to the university (F6), was potentially explosive. Although there was general agreement that this was an important consideration, the relative importance of subjects might change over time, or be driven by external pressures from the national government, or by benefactors. Any public statement of even a simple ranking would evoke such a storm of protest that any benefits deriving from a new budget allocation process would be discounted. Note that, in principle, this method can be used for any number of budgetary units and factors. The discussions were by no means systematic, but eventually the librarians servicing the group (the Director, and the Head of Collection Management) suggested that rather than try to find a formula, the group should use a simple matrix to view the factors. The final version of this matrix had eight rows (one for each faculty) and six columns (one for each factor), with a row and a column for totals. The matrix was in

Faculty

Size

Library importance

Items published

Price

Arts Education Law Social Sciences Engineering Maths Medicine Science Totals

Breadth

Subject importance

Totals

4800

Figure 1: The Matrix for the Budget Allocation Game

56

Art or Science ? effect the board on which the budgetary allocation "game" could be "played" (figure 1). The matrix thus had 48 cells. Each "player" was given 4800 "points" to distribute across the matrix. Each point had the same monetary value - the total materials budget available divided by 4800. It was decided that the players would be allowed to weight each factor and faculty in any way that they chose: at the extremes it would be possible to allocate all the points to one cell, or to distribute them evenly across all cells. 3.2 Playing the Game As remarked, each faculty was represented on the Library Committee by a member of academic staff, and these representatives were chosen as the players. The Head of Collection Management (Ford) had the agreeable task of explaining the rules, and collating the results. The representatives were furnished with such data as were available: numbers of staff and students in various categories, numbers and average prices of academic books published in the UK, numbers and average prices of serials published world wide. It will be seen that the complete freedom to weight the factors in any way, combined with the absence of any numeric measures for factors F2, F5 and F6, and incomplete data to provide measures for F3 and F4, ensured that the allocation process at this stage was almost completely subjective. The individual weights (numbers of points allocated to each cell) varied considerably, but what was remarkable was, that when summed across the rows, the allocations made by each representative for each faculty were very similar, so that it was reasonable to strike an average and present this as the consensus of the group. Compared with the existing allocations, the results overall suggested a transfer of funds away from science based subjects; five faculties would have experienced a net gain, and one (Education) would have been embarrassed by the resulting largesse - it would have been difficult to spend the allocation. 3.3 Results The artistic method did not become operational. Although there were five "winners", as remarked, one faculty could not have spent its allocation. It would have been possible to impose new rules for a further iteration - for example, that no faculty should receive an allocation greater than a certain amount. No such action was taken. The method just did not find political acceptability. The ultimate decision makers had participated in developing the method, and had agreed the fundamental criteria (the six factors), but they could not live with the results. Perhaps the use of the word "game" (although strictly accurate) was the ultimate nemesis of the method. 4. The Scientific Approach 4.1 A Change of Tack The former approach was criticised, not least by the scientists in the university, as not being scientific. So, in the following year, operational research techniques were used to develop a systematic method for allocating library materials funds. The study was carried out by Abdinnour as a M.Sc. project (Abdinnour, 1987), with Ford acting as 57

Geoffrey Ford and Suhair Abdinnour supervisor, library contact and facilitator. Technically speaking, the exercise was to build a mathematical model. It was important to continue the involvement of the Library Committee to ensure that the concerns of the academic staff were represented in the analysis, even if the results proved politically sensitive. As before, the purpose of the exercise was to arrive at allocations for the eight academic faculties. It was decided that a linear programming model would be appropriate. 4.2 Linear Programming The linear programming model consists of a linear expression, known as the objective function, that is to be maximised, subject to a number of linear constraints. In this study the objective function to be maximised was the benefit derived from the allocated funds. A major part of the exercise was the identification of the benefit factors that could be used to derive the objective function (4.3, below). There were four different kinds of constraints. The first of these was a global fund constraint - clearly, the total mount of money allocated could not exceed the available budget. Then there were lower and upper bound constraints: these would ensure that at least some money was allocated to each faculty, and that no faculty got too big an increase over the previous year. The third set took into account the supply side: this was to ensure that the number of books and serials purchased did not exceed the number published. Finally, there was a grouped constraint. For a number of years, the price inflation of serials had ensured that an increasing proportion of the total budget had crept towards the "science" side of the university. To ameliorate this trend, lower and upper bounds were set on the proportion of the total budget that should be allocated to the "arts" side. These constraints were informed by discussion with members of the Library Committee, and by the experience of the artistic method in the previous year. Although any linear programming model is strictly mathematical in operation, human judgement is necessary in its formulation, in this case in the setting of lower and upper bounds. It was decided, in consultation with representatives of the academic staff, that no faculty should be allocated less than 90% of the allocation in the previous year, nor more than 50% more. Politically, a cut in allocation of more than 10% would not have been acceptable. Similarly, the grouped constraint was the result of a political issue. The suggestion was made that the "arts" side of the university should be allocated between 25% and 50% of the total. 4.3 The Objective Function As already stated, it was decided that the most sensible objective was to maximise the benefit (to the university) that could be obtained from the allocation of funds. The first step was conventional, in that it involved identifying the factors that might be relevant in fund allocation. 4.3.1 Factors Affecting Allocation Some of these were suggested by the Library Committee, others were identified from the extensive literature. These factors were made operational as variables, to which numerical values could be assigned. Data were collected or estimated for two

58

Art or Science? successive years for the analysis. In each case it was important that data could be analysed by the eight faculties, could be independently verified, and could be replicated relatively easily. This led to the rejection of a number of possible factors that could not be made operational in any objective way. - for example, "hard" subjects as opposed to "soft"; the need to provide for subjects taught on more than one site; and so on. In principle, so long as data are available, the method can be used for any number of variables (we used 23), and any number of budgetary units (we used the 8 faculties, but it would have been possible to use individual departments). 4.3.1.1 Clientele There were seven variables in this category, which collectively corresponded to Fl in the artistic approach. The first four represent student enrolment: VI) V2) V3) V4)

number of undergraduate students number of postgraduate students following taught courses number of postgraduate research students the total number of students (the sum of the first three).

This variable is one of the commonest used in allocation formulae, since it is an indicator of potential demand for library materials. The division of the student body into three categories is conventional, and there is evidence that levels of use do vary between these categories. Student numbers are commonly quantified as full time equivalents (FTE), with standard formulae for calculating part time students as fractions of full time. The source of data was the Academic Registry of the university, and therefore non-controversial. The remaining three variables relating to clientele are considered under 4.3.1.4. 4.3.1.2 Publication Output Corresponding to F3 in the artistic approach, it was important to include variables for this factor, since there is much variation between subjects in the numbers of books and serials published. There were insufficient data available to satisfy the purists, but two external data sources were used as indicative. V5) number of academic books published in the UK (data published by Loughborough University/CLAIM, the predecessors to the Library & Information Statistics Unit) V6) number of academic serials published world wide (data obtained from databases of serials agents). 4.3.1.3 Prices One measure of "fairness" is the degree to which faculties enjoy equivalence of purchasing power, as there is wide variation in the prices of books and serials in different subjects. These variables corresponded to F4 in the artistic approach.

59

Geoffrey Ford and Suhair Abdinnour V7) average price of books (data published by Loughborough University /CLAIM, the predecessors to the Library & Information Statistics Unit) V8) average price of serials (data obtained from serials agents). The analysis of data from the library's own records would have ensured that the full range of subjects, and of countries of publication, was covered; but this analysis would have reflected only the items that the library could afford to buy, and may therefore have underestimated the true pricing situation. 4.3.1.4 Purpose of Use The purposes for which libraries are used varies. While all students can be considered to be in learning mode, academic staff have a variety of roles, which may affect potential use. These were made operational as three variables: V9) number of staff engaged in both teaching and research V10) number of staff engaged in research only VI1) total number of staff (the sum of V9 and V10). These categories are standard ones, and data were obtained from the university's personnel office. An additional refinement would be to add a variable for "Teaching only" staff, but these were few in number at the time of the case study. 4.3.1.5 Range of Subjects It can be argued that the number of the numbers of staff and students are not so important as the range of subjects studied. One measure of this is the number of programmes offered, but there were operational difficulties in counting these at the time of this case study. As a surrogate, the number of departments in each faculty was used. This was a first approximation to the factor F5 in the artistic approach. VI2) number of departments. 4.3.1.6 Demand for Materials The measurement of demand is fraught with difficulties, and measures of use are frequently used instead. At the time of the study there were no readily available figures for in-library use that differentiated between subjects, nor for use of databases, so circulation had to be used as a surrogate. Citations in research theses and use of interlibrary loans were counted to give some indication of demand for research materials. VI3) number of items circulated (from library records) VI4) number of references cited in research theses (sample survey of 4 years' theses) VI5) number of interlibrary loans (from library records).

60

Art or Science? 4.3.1.7 History A number of factors reflect the historical position. Opinions vary as to the relevance of these factors to current problems of allocation, but it was important to include them in the analysis - if only to be able to eliminate them. VI6) size of existing collection, for comparison with publishing output, and to reflect current status of the library for each faculty (from library records) VI7) percentage of items cited less than five years old (data from Price (1970) admittedly rather out of date but the only source at the time that provided the necessary operational break down by faculty) VI8) amount of money allocated in year (from library records). The last of these was potentially controversial, since it reflected past decisions, that may have been purely subjective. However, at the analytical stage it was important to include it. 4.3.1.8 Miscellaneous VI9) relative importance of library provision to the study of a subject. This was the only subjective factor included in the analysis. It was derived from the results of the previous year's "artistic" exercise in resource allocation, since that did represent the collective opinions of academic staff representatives. It is perhaps ironic that for this factor, F2, no measure was agreed by the academic staff, but their individual decisions on the number of points that should be allocated in the game actually provided a measure. V20) price inflation rate of books and serials. A weighted average for the price inflation of books and serials actually purchased by the library over the preceding five years. Important because the rate of inflation varies considerably between subjects. V21) time spent by academic staff on teaching undergraduates V22) time spent by academic staff on teaching graduate students V23) total time spent by academic staff on teaching (V21+V22). Since use of libraries varies between different categories of students, it was thought that it might be interesting to investigate how staff time was allocated. The data came from a rather old source (CVCP(1969)), but for the purposes of analysis and modelling it sufficed. 4.3.2 Identifying Relationships It seemed likely that there would be considerable correlation between these variables. The next stage of the analysis was to simplify the problem by identifying significant relationships. As a first step, each variable was correlated with each of the others. This process generated a formidable array of numbers, both the correlations and the levels of statistical significance associated with each correlation. The mathematical techniques used were standard ones, but for the purposes of this paper we need to note only that:

61

Geoffrey Ford and Suhair Abdinnour • • •

perfect correlation between two variables yields the coefficient of 1.0; where there is no correlation the coefficient is 0.0; a perfect inverse correlation has a coefficient -1.0.

There are two levels of significance that are commonly used, 1% and 5%. At the 5% level, there is a 1 in 20 chance that the correlation occurred by chance. At the 1% level, there is a 1 in 100 chance that the correlation occurred by chance. In this case study, a correlation coefficient of 0.83 or greater is significant at the 1% level; a coefficient in the range 0.71-0.82 is significant at the 5% level. As stated above, the analysis was conducted on data for two successive years. Table 1 shows some of the interesting correlations that were significant at the 1% level in both years, with the coefficient of correlation for the second year, with interpretations. Variable Number of serials published (V6) Citations in theses (V14)

Correlates with Number of undergraduate students (VI)

Coefficient 0.87

Possible Interpretation Appears to be a chance correlation

Money allocated (VI8)

0.88

Size of collection (V16) Circulation rate (V13)

Number of books published (V5)

0.99

Number of books published (V5)

0.99

Use of interlibrary loan (V15)

Number of staff engaged in teaching & research (V9)

0.98

Higher allocations yield larger collections to refer to and thus be cited? Library reflects level of academic interest, and hence publication rate? The more books are published, the more are in the collection, and hence the more that they are available for use An obvious link

Table 1: Correlations Significant at the 1% Level There were rather more correlations significant at the 5% level in both years, and a few are shown in Table 2. There were several negative correlations, but most of the values were low, and could be treated as insignificant. The only significant negative coefficients (at the 5% level) were all connected with the rate of obsolescence. It is difficult to interpret these, but a possible explanation is that the measure used is not very helpful - the data were rather old. The point of studying these correlations was to gain insights, and much further work could be done to identify which, if any, are typical of all similar institutions. The general significance of the correlations in this study are that two variables that are highly correlated with each other are both telling the same story, so that one can be

62

Art or Science? discarded. Variables with low correlation coefficients are separately important, and should both be accounted for in the model. Variable Number of books published (V5) Average price of serial (V8)

Correlates with Relative importance of library to the study of a subject (V19) Money allocated (VI8)

Coefficient 0.82

Possible Interpretation Suggests that the relative importance of a library increases with the availability of books

0.79

Circulation rate (VI3)

Relative importance of library to the study of a subject (VI9) Number of undergraduate students (VI)

0.82

Serials price inflation is generally higher than for books, and there is a strong tendency for the proportion of the total budget allocated to serials to increase over time This indicates that users' behaviour is affected by their value system

Use of interlibrary loan (VI5)

0.73

Number of undergraduates is related to numbers of teaching staff, and inter library loan traffic increases with numbers of teaching staff

Table 2: Correlations Significant at the 5% Level 4.3.3 Principal Components Analysis From the mass of data originally collected, a number of correlations were thus derived. The purpose of the next stage was to simplify the data without significant loss of information. This was done by transforming the correlated variables to a new set of uncorrelated variables called Principal Components. These are linear combinations of the original variables, calculated using a standard mathematical technique. These are calculated in decreasing order of importance, so that the first component accounts for as much as possible of the variance in the original data. As there are 23 variables, the maximum number of components possible is also 23, and between them these account for 100% of the variance. The objective of the analysis is to see whether a few of the most important components can between them account for the bulk of the variance, and thus simplify the model for operational use. In this study, the first three components accounted for about 80% of the total, the first four accounted for 88 %, the first five for 93%. For the purposes of illustration it was decided to use just three principal components. In effect each component is a formula in which each variable is weighted by a loading derived from the mathematical analysis. Each principal component thus has a numerical value. We can further transform the data by examining the correlation of each variable with the components. 63

Geoffrey Ford and Suhair Abdinnour The principle components were used to derive the benefit factor (see below, 4.3.4). For simplicity, each component was represented by a single variable, although in principle, any number in combination could have been used. The first component was highly correlated with the total number of students (VI), so that variable was chosen to represent it. Similarly, size of existing collection (VI6) and number of citations in theses (V14) were chosen to represent the other two principal components. While there was nothing surprising about the emergence of these three variables as being significant in the budget allocation process, the mathematical analysis demonstrated that these three were effectively independent of each other. There is some satisfaction to be derived from the fact that the three variables identified could be said respectively to reflect the teaching side, the existing situation, and the research needs. 4.3.4 Benefit Factor The three important variables can be said to add to the benefit of each faculty when they are used in the allocation process. More students equals more money to buy books; greater number of citations in theses equals more money to support the research activity. It is possible to criticise the use of the size of existing collection as a variable in the allocation process - it could be argued that past decisions should not have such an influence. But it will be recalled that this variable was highly correlated with the number of books published (see Table 1), so that publishing output could equally well have been chosen to represent the second component. In this case, size of existing collection was chosen because the data were easy to collect, and the variable included the number of serials as well as books. In order to derive a benefit factor, a small sample of academic and library staff were invited to rank the three factors in order of importance. The average of these rankings gave the weight that each variable would be given, specifically: Number of students: W1 = 0.365 Size of collection: W2 = 0.326 Number of citations: W3 = 0.309. The calculation is then:

Benefit factor for each faculty

_ ~

W2 χ size of collection for faculty Total size of collection

W1 χ number of students in faculty Total number of students

The resulting benefit factors are shown in the Table 3. Faculty Arts Education Law

Benefit Factor 0.197 0.047 0.063 64

W3 χ number of citations in theses from faculty Total number of citations

Art or Science? Faculty Social Sciences Engineering Mathematics Medicine Science

Benefit Factor 0.140 0.136 0.041 0.112 0.262

Table 3: Benefit Factors These benefit factors could then be used in the linear programming model for budget allocation. 4.4 Model Results The constraints having been set (4.2, above), and the benefit factors identified, the linear programming model thus formulated yielded a set of suggested allocations (Table 4). The allocations were calculated in terms of actual cash, but are here transformed into percentages of the total budget available. Faculty Arts Education Law Social Sciences Engineering Mathematics Medicine Science Totals

Previous Year's Allocation (% of total) 13.4 2.9 5.2 10.0

Model Allocation (% of total) 13.6 2.7 5.0 13.6

12.2 5.3 18.1 32.9 100.0

12.0 5.7 16.5 31.0 100.0

% Change in cash from previous year 12.45 4.28 7.33 50.63 8.63 18.18 1.07 12.45 10.95

Table 4: Allocations from Linear Programming Model It will be observed that, although the Science Faculty had the highest benefit factor, it was the Social Sciences Faculty that achieved the highest percentage increase; and that, while the overall total budget had increased by more than 10%, five faculties did not achieve this level of increase in the model. This suggests that the original allocations had a degree of imbalance that the model could begin to address. 4.5 Sensitivity Analysis This analysis was designed to explore the robustness of the model allocation. An interesting test was to change the weights of the objective function. For example, we can interchange the values for the first two variables, so that the weights become

65

Geoffrey Ford and Suhair Abdinnour Number of students: Size of collection: Number of citations:

W1 = 0.326 W2 = 0.365 W3 = 0.309.

Although this generated a new objective function, there was no change to the allocations generated by the model. Another way of testing the sensitivity is to change the constraints. For example, removing the constraint on the number of books that can be bought by the Science Faculty has a marginal effect only. However, removing the same constraint from all eight faculties, had a dramatic effect. Most of the allocations hit the lower bound (i.e. 90% of the previous year's allocation) with nearly all the surplus thus generated going to the Science Faculty. 4.6 Improvements to the Model It will be obvious from the description of the process that many improvements could be made. In particular, the estimation of the constraints. For example, the constraint of maximum numbers of serials that could be purchased. These numbers were rather large, but there seemed to be no easy way of getting more realistic figures. The advent of electronic publishing, and the corresponding introduction of bundling of access arrangements has overtaken the more traditional acquisition process, and libraries now need to top slice larger proportions of the total budget. Running separate models, with separate benefit factors and constraints, for the books and serials budgets would be another improvement. 4.7 Discussion To summarize, the scientific approach involved the formulation of a model, the building of the model, analysis, and testing and interpretation. We believe the model is easy to understand, and is certainly flexible. (Of course, if an optimal solution is to be sought annually, then the data would need to be updated annually, but the three final variables are all relatively easy to update.) There are five features of the scientific method described here that are of particular interest from a librarian's point of view. First, the correlations between variables: not just the expected ones, but also the unexpected ones; further work with data from other libraries could be of interest. Second, the variables that represent the three principle components: the three variables all ring true as being important in the allocation process. On reflection, the "number of books published in the previous 12 months" is preferable to the "size of library's existing collections", because the object of this exercise was to find an allocation method that would take account of the current needs of the university, rather than one that reflected past history. With this change, the variables would then represent the size of the user community, the supply of new publications, and the use of the library for research. (In passing, we note that "circulation of books" is also highly correlated with "size of collection": if circulation is used to represent the second component, then the results of this stage of the modelling would be identical with those reached by factor analysis conducted by McGrath, Huntsinger and Barber (1969)). 66

Art or Science? A third feature of interest is the relative weights assigned to the three variables. It might have been expected that more weight would have been given to "number of students". In the application of the model, the weighting of these three variables would be an important political consideration. Fourthly, the results of the linear programming model: although they differed in detail, they reflected the trend established by the "artistic" approach described earlier. That is, that money should be transferred from the "science side" to the "arts side". Finally, and in some ways the most important feature: the process could be made operational fairly easily. The principal components analysis yielded variables that have face validity, the data required are "hard" numerical data, and there is scope for incorporating political judgements. 4.8 Implementation The next stage in the process was to gain acceptance and implementation. Since some of the decision makers were involved in the model building from the start, the final solution was well understood. Thus one might have expected that it would have been accepted and implemented. This proved no more easy than with the artistic approach, and the Library Committee proved unwilling to agree on a method for weighting the variables for the benefit factor. This would have been necessary to make operational the process described in 4.3.4. In the event the Committee decided to retreat from the process, and to act on the advice of the library Director, in collaboration with a few "wise persons". These were able, of course, to use any method they chose. 5. Conclusion Art or science? The choice is for the reader, but it is worth noting that each approach has elements of both. The "artistic" method relied on some small degree of quantification of subjective judgements, while the "scientific" method involved making subjective choices in setting parameters for the objective data. The failure of the academic staff representatives on the Library Committee to agree on the methods of making political judgements more transparent is noteworthy. Perhaps we may conclude (almost) where we began: the library director must be versatile, nimble and open-minded. References Abdinnour, S.F., Allocation of library funds to the different faculties at Southampton University. M.Sc. dissertation, Southampton University, 1987. CVCP Report of an enquiry into the use of academic staff time. London: Committee of Vice Chancellors and Principals of the Universities of the United Kingdom, 1969. Ford, G. "Finance and budgeting." In: Jenkins, C. and Morley, M. (editors). Collection management in academic libraries, 2nd edition. Aldershot: Gower, 1999, 3970.

67

Geoffrey Ford and Suhair Abdinnour McGrath, W.E., Huntsinger, R.C. and Barber, G.R. "An allocation formula derived from factor analysis of academic departments." College and Research Libraries, 44(2) 1969, 116-127. Price, D.J.S. "Citation measures of hard science, soft science, technology, and nonscience." In: Nelson, C.E. and Pollock, D.K. Communication among scientists and engineers. Lexington: Heath Lexington Books, 1970, 3-22.

68

Sigrid Reinitzer und Roswitha Karpf

Wissenschaftliche Bibliotheken im Kontakt zu Schulen und Schulbibliotheken

Es ist mir eine besondere Freude, für eine langjährige Kollegin einen Beitrag zu ihrer Festschrift verfassen zu dürfen. Natürlich sollte es ein Thema sein, das speziell für diese hochgeschätzte und bewunderte Kollegin von Interesse ist, aber es sollte wiederum kein Thema sein, bei dem sie die größte Kompetenz besitzt, wodurch sie vielleicht sogar gelangweilt würde. So entschied ich mich für ein bibliothekarisches Thema, dem die wissenschaftlichen Bibliotheken bisher nicht ihre erste Aufmerksamkeit schenkten, das aber in einer Festschrift, für eine Kollegin mit einem so weit gespannten Interessenbereich nicht fehlen sollte. Es wird in diesem Beitrag um die Zusammenarbeit von wissenschaftlichen Bibliotheken mit Schulen und Schulbibliotheken gehen. Gerade in einer Zeit, in der die allgemeine Lesekompetenz gefördert werden soll, muss der Jugend möglichst frühzeitig die Nutzung einer Bibliothek ermöglicht werden. Diese Überlegungen stellen sich für alle Bibliothekstypen, für die wissenschaftlichen Bibliotheken ebenso wie für die öffentlichen Büchereien, die Schulbibliotheken, aber auch für die Bibliotheksverbände der einzelnen Länder und für die Weltorganisation der bibliothekarischen Institutionen, die IFLA (International Federation of Library Associations and Institutions). Schulbibliotheksmanifest Die IFLA hat das Manifest für die Schulbibliotheken erarbeitet, im November 1999 wurde es im Rahmen der Generalkonferenz der UNESCO approbiert und 2000 von der IFLA herausgegeben. Das Manifest ist unter folgender URL zu finden: chttp:// www.ifla.org/VII/sll/pubs/mani-g.htm>. „Lehren und Lernen mit der Schulbibliothek. Das Schulbibliotheksmanifest der UNESCO". Darin wird als zentraler Wille aufgezeigt: „Die Schulbibliothek stellt Informationen und Ideen zur Verfügung, die grundlegend für ein erfolgreiches Arbeiten in der heutigen informations- und wissensbasierten Gesellschaft sind. Die Schulbibliothek vermittelt den Schülern die Fähigkeit zum lebenslangen Lernen, entwickelt ihre Phantasie und befähigt sie so zu einem Leben als verantwortungsbewusste Bürger." In diesem Manifest wird auch darauf verwiesen, dass die Zusammenarbeit mit anderen Bibliothekaren und Informationsfachleuten der örtlichen Gemeinschaft gefördert werden soll.

69

Sigrid Reinitzer und Roswitha Karpf In den letzten Jahren bemühten sich Universitäten immer stärker das Image des Elfenbeinturms abzuschütteln und auf die sogenannte „breite Öffentlichkeit" zuzugehen. So bestand auch für die Grazer Universitätsbibliothek das Ziel, weitere Benutzerkreise für die wissenschaftliche Bibliothek zu interessieren. Umgekehrt hatten auch die Schulen mit ihren Schulbibliotheken die Aufgabe die Schülerinnen und Schüler auf die Zukunft vorzubereiten und das lebenslange Lernen frühzeitig als wichtiges Ziel aufzuzeigen. So ergab sich der Wunsch aus Sicht der Leiterin der Universitätsbibliothek zugleich mit dem Wunsch einer Germanistik- und Geschichtsprofessorin eines Grazer Gymnasiums diesen Weg gemeinsam zu gehen. Die zentrale Frage war: Wie werden aus Schülerinnen während der Schulzeit kompetente Nutzerinnen der wissenschaftlichen Bibliotheken? Dazu bedarf es sowohl der Fixierung des Stellenwertes von Bibliotheken innerhalb eines Bildungskonzeptes als auch des Engagements von Bibliothekarinnen und Lehrerinnen. Die Nutzung von Bibliotheken ist im österreichischen Schulsystem gut verankert. Zum einen durch die zahlreichen Schulbibliotheken, zum anderen durch den Lehrplan insbesondere im Fach Deutsch, der den Besuch - und gemeint ist wohl die Nutzung von wissenschaftlichen Bibliotheken vorsieht. Österreichische

Schulbibliotheken

Vor über 20 Jahren wurde mit einer Offensive zur Errichtung von Schulbibliotheken begonnen und das Modell „Schulbibliothek an höheren Schulen unter Mitarbeit von Schülern" entwickelt. Die Schulbibliotheken werden von zu Bibliothekarinnen ausgebildeten Lehrerinnen betreut, wodurch die bestmögliche Integration der Bibliothek in Unterricht und Schulleben garantiert ist. Im gymnasialen Bereich verfügt Österreich über ein dichtes Netz von Schulbibliotheken, deren Größe, Öffnungszeiten und Dotationen durch einen ministeriellen Erlass über das ganze Bundesgebiet einheitlich geregelt sind.1 Pflichtschulen und berufsbildende höhere Schulen holen in den letzten Jahren stark auf. Zurzeit gibt es in Österreich rund 2000 Schulbibliotheken mit insgesamt 6 Millionen Medien, das entspricht etwa dem Gesamtbestand der Österreichischen Nationalbibliothek. Je nach Schultyp werden die Schulbibliotheken von Bund, Ländern oder Gemeinden finanziert, darüber hinaus versuchen aber die Schulbibliothekarinnen häufig, die begrenzten finanziellen und personellen Ressourcen durch Sponsoring und Anwerbung ehrenamtlicher Mitarbeiterinnen, z.B. Schülerinnen, Kolleginnen und Eltern zu optimieren. Die Schulbibliothek bildet die Drehscheibe des schulischen Lernens. Von Anfang an hatte sie über die unterrichtsbegleitende Funktion hinaus aber auch die Aufgabe, den Schülerinnen Möglichkeit zu weiterführendem Lesen zu geben und so ihre lebenslangen Leseinteressen und Lesegewohnheiten zu fördern. Deshalb ist bei Bestandsaufbau und Bestandspflege auf Ausgewogenheit Bedacht zu nehmen, dass neben den für den Unterricht benötigten Medien auch „Lesefutter" zur Verfügung steht. Lesen soll nämlich durchaus Freude und Spaß bereiten, der Entspannung und Unterhaltung dienen.2

2

Modell „Schulbibliothek an höheren Schulen unter Mitarbeit von Schülern", chttp:// www.schulbibliothek.at/sb-aktuell/service/f_gesetze.htm> am 15. 8. 2003. Vgl. am 30. 8. 2003.

70

Wissenschaftliche Bibliotheken im Kontakt zu Schulen und Schulbibliotheken Ausdrücklich wird im „Grundsatzerlass zum Unterrichtsprinzip Leseerziehung" 3 von 1999 auf die Bedeutung und die Funktion des Lesens angesichts neuer Informations· und Kommunikationstechnologien hingewiesen. Der Erlass erklärt den Erwerb von Lesekompetenz zum Unterrichtsprinzip in allen Unterrichtsfächern und gibt den Lehrerinnen Richtlinien und Anregungen, um • • • • •

die Freude am und die Bereitschaft zum Lesen zu fördern Lesen als selbstbestimmte, lebensbegleitende Tätigkeit zu vermitteln die Kooperation mit (öffentlichen) Bibliotheken zu intensivieren einen kompetenten Umgang mit Textträgern unter Einbeziehung neuer Informations- und Kommunikationstechnologien zu vermitteln und an der Entwicklung einer schulischen Lesekultur als wichtigem Beitrag zur Schulqualität mitzuwirken.

Wie wichtig die Fähigkeit zu sinnerfassendem Lesen für den Einzelnen ist, zeigt das folgende Zitat aus dem Grundsatzerlass: „Lesen fördert den Erwerb und die Verwendung von Sprache in ihrer Funktion als Medium des Denkens, des Informationsaustausches und der Gestaltung von Beziehungen. Dadurch hat Lesen zentrale Bedeutung für die individuelle Entwicklung im kognitiven, emotionalen, sozialen, kreativen und pragmatischen Bereich und schafft Grundlagen für selbst bestimmtes und selbst organisiertes Denken, Bewerten und Handeln im privaten, beruflichen und öffentlichen Leben." Damit möglichst viele Kinder die notwendige Lesekompetenz entwickeln, wurde als Antwort auf die Ergebnisse der PISA-Studie in Österreich im Jahr 2002 die Aktion LESEFIT4 gestartet, ein Projekt des österreichischen Buchklubs, vom Bundesministerium für Bildung, Wissenschaft und Kultur initiiert. Die Motivation zum Lesen soll mit dieser Aktion gefördert und die Lesefreude geweckt werden. Man hofft die Zahl der Kinder mit Leseschwächen zu halbieren. Eingebunden in das 2002/2003 sehr erfolgreich verlaufene Projekt waren Lehrerinnen, Eltern, Schulbibliotheken und der österreichische Buchklub. Die breit angelegte Öffentlichkeitskampagne im Rahmen von LESEFIT, die Aktion LESEzeichen, wurde mit Aktivitäten zur schulischen Lesekultur verknüpft und stieß auf eine unglaublich hohe Resonanz. Ihr Ziel war es, ein sichtbares Zeichen für das Lesen und die Lesefreude zu setzen. Eltern lasen mit ihren Kindern ein Jahr lang Bücher und die Kinder bastelten dazu fast 110.000 Lesezeichen. Die große Abschlussfeier der Aktion fand mit 300 Schülerinnen aus allen Bundesländern in der österreichischen Nationalbibliothek statt. „Fit beim Lesen - fit für das Leben", und „Lesen können heißt lernen können" mit diesen Schlagzeilen wird in Österreich für das Lesen und die Schulbibliotheken geworben. Studien haben ergeben, dass es einen engen Zusammenhang zwischen Schulbibliothek und Lernerfolg gibt, so dass die Frage „Leisten Schulbibliotheken 3 4

am 30. 8. 2003. am 30. 8. 2003.

71

Sigrid Reinitzer und Roswitha

Karpf

einen Beitrag zum Lernerfolg?" mit „Ja" beantwortet werden kann, wenn diese Schulbibliotheken gut ausgestattet sind, von gut ausgebildeten und engagierten Schulbibliothekarinnen geführt werden und der Erwerb von Lese- und Medienkompetenz wichtiges Ziel des UnteiTichts ist.5 Den neuen Möglichkeiten, Erfordernissen und Technologien des Informationszeitalters Rechnung tragend, hat das Bundesministerium für Bildung, Wissenschaft und Kultur zur Jahrtausendwende eine Arbeitsgruppe erfahrener Schulbibliothekarinnen unter wissenschaftlicher Betreuung mit der Weiterentwicklung des österreichischen Schulbibliothekmodells zur „Multimedialen Schulbibliothek an höheren Schulen unter Mitwirkung von Schülerinnen und Schülern" beauftragt.6 Die multimediale Schulbibliothek mit der entsprechenden Infrastruktur (Ausstattung der Bibliotheken mit PCs, OPAC, Anbindung ans Internet, Angebot der unterschiedlichsten Print- und digitalen Medien usw.) unterstützt moderne Unterrichtsformen wie offenes Lernen und projektorientierten Unterricht. Schlüsselqualifikationen wie Medienkompetenz, Informationskompetenz, Recherchefertigkeiten etc. sind erklärtes Lernziel des Unterrichts in der Schulbibliothek. Als zentrale Beratungsstelle zum Schulbibliothekswesen in Österreich steht der „Bibliothekenservice für Schulen" des Bundesministeriums für Bildung, Wissenschaft und Kultur zu Verfügung. Besonders hilfreich erweist sich die Homepage , die im Jahr 2003 mit dem Comenius-Euromedia-Siegel ausgezeichnet wurde. Hervorgehoben werden in der Begründung der Jury das pädagogisch-didaktische Engagement und die Tatsache, dass zahlreiche für den Unterricht und den Heimgebrauch verwendbare Handreichungen von der Homepage herunter ladbar sind. Die Homepage wendet sich in ihren Untermenüs an Lehrerinnen, Schülerinnen und Eltern. Sie enthält u.a. Tipps zur Errichtung und zum Bestandsaufbau der Bibliothek, Lesetipps, zahlreiche Handreichungen für den Unterricht, alle relevanten Gesetze und Verordnungen sowie aktuelle Neuigkeiten und Zugang zu einer Vielzahl an Online-Medien. Monatlich werden rund 3.800 Zugriffe verzeichnet, die länger als drei Minuten dauern. Wesentliche Teile der Homepage sind nach den WAI-Richtlinien7 gestaltet und sollen Menschen mit eingeschränkter Seh- und Hörfähigkeit ebenfalls einfachen Zugang zu Informationen rund um die Schulbibliothek ermöglichen.8 Bereits in der Anfangsphase von hat die VÖB (=Vereinigung Österreichischer Bibliothekarinnen und Bibliothekare) in Kooperation

6

7 8

Vgl. Johanna Hladej. „Leisten Schulbibliotheken einen Beitrag zum Lernerfolg?" In: Büchereiperspektiven, 1/03, S. 46-47. Weitere Informationen: am 15. 8. 2003. Entwicklung eines Modells für die „Multimediale Schulbibliothek" an höheren Schulen unter Mitwirkung von Schülerinnen und Schülern. 2001. am 15. 8. 2003. Die multimediale Schulbibliothek, Phase 2. Begleitende Forschung zum Projekt. 2002. Entwicklung eines Modells für die „Multimediale Schulbibliothek" an höheren Schulen unter Mitwirkung von Schülerinnen und Schülern. Phase 2. Zwischenbericht. 2002. Web Accessibility Initiative: .

vom 15.7.2003, am 15.8. 2003.

72

Wissenschaftliche Bibliotheken im Kontakt zu Schulen und Schulbibliotheken mit den österreichischen Schulbibliothekarinnen für diese Homepage eine Kurzdarstellung der österreichischen wissenschaftlichen Bibliotheken erstellt, die komprimiert die für Lehrerinnen und Schülerinnen wichtigsten Informationen über die einzelnen Bibliotheken, die entsprechenden Kontaktstellen und die für Schulen angebotenen Services auflistet: . Die österreichischen Schulbibliothekarinnen kommunizieren über eine eigene Mailing-Liste, die Arbeitsgemeinschaft Schulbibliotheken Vorarlberg gibt seit 1993 die Zeitschrift „Schulbibliothek" heraus, die Schulbibliothekarinnen und an Projekten mit Bibliotheken interessierte Lehrerinnen anspricht und so eine wertvolle Information darstellt. Außerdem bieten die „Büchereiperspektiven", die Mitteilungen des Büchereiverbandes Österreichs, eine Plattform für Publikationen aus dem Bereich des Schulbibliothekswesens. Ein wesentlicher Faktor für die Qualitätssicherung der schulbibliothekarischen Arbeit sind die in allen Bundesländern angebotenen Ausbildungs- und Fortbildungslehrgänge. Die Wichtigkeit der Kooperation aller Bibliothekstypen wurde von allen Seiten erkannt. Zum Beispiel sieht der Studienplan des sechssemestrigen, 270 Stunden umfassenden Akademielehrganges „Ausbildung von Schulbibliothekaren zur Leitung einer multimedialen Schulbibliothek unter Mitarbeit von Schülern an höheren Schulen", Wien 2002, Exkursionen in öffentliche und wissenschaftliche Bibliotheken vor sowie die Entwicklung von Modellen für Zusammenarbeit, Erfahrungsaustausch und Möglichkeiten einer Vernetzung.9 Kooperationen In den letzten Jahren wurden Kooperationen zwischen den für die Lehrerfortbildung zuständigen Pädagogischen Instituten, dem Bibliothekenservice für Schulen, den öffentlichen und wissenschaftlichen Bibliotheken sowie der Vereinigung österreichischer Bibliothekarinnen und Bibliothekare forciert. Die Universität Graz veranstaltete in den Jahren 1997-2001 jeweils Österreichische Buchmessen unter Einbindung von Wissenschaftern, Fachleuten aus Wirtschaft und Politik sowie des Verlags- und Bibliothekswesens. In den Jahren 2000 und 2001 gab es gemeinsame Präsentationen aller Grazer Bibliotheken sowie von Grazer Schulbibliotheken. Hierbei wurde die Gestaltung der einzelnen Homepages vorgestellt und gemeinsame Fortbildungsveranstaltungen wurden durchgeführt. Auch die beiden letzten Österreichischen Bibliothekartage 2000 in Wien und 2002 in Klagenfurt boten Gelegenheit zu umfassenden internationalen Kontaktgesprächen und Fortbildungsveranstaltungen für Bibliothekarinnen der wissenschaftlichen Bibliotheken und der Schulbibliotheken. Beim Österreichischen Bibliothekartage 2000 in Wien widmeten sich drei Vorträge im Rahmen des Themenkreises „Kultur- und Informationszentren für Öffentlichkeit, Bildungsstätten für lebensbegleitendes Lernen" den Schulbibliotheken:10 • Johanna Hladej, Bibliothekenservice für Schulen: Schulbibliotheken in Österreich. Eine Bestandsaufnahme.

10

am 30. 8. 2003. am 30. 8. 2003.

73

Sigrid Reinitzer und Roswitha Karpf • •

Roswitha Karpf, Graz: Doppelfunktion Lehrerin - Schulbibliothekarln. Werner Schöggl, Wien: Die multimediale Schulbibliothek - das Modell der Zukunft?

Beim 27. Österreichischen Bibliothekartag 2002 in Klagenfurt war ein ganzer Tag den Schulbibliotheken gewidmet.11 Der Themenkreis „Schulbibliotheken und Kooperationen von wissenschaftlichen Bibliotheken mit Schulen" befasste sich mit folgenden vier Schwerpunkten: 1) Schulbibliotheken im internationalen Vergleich, vorgestellt wurden von • Johanna Hladej, Bibliothekenservice für Schulen: Situation der Schulbibliotheken in Österreich. • Markus Fritz, Amt für Bibliothekswesen Bozen: Die Schulbibliothek als Lernort: Bausteine des Bibliotheksunterrichts, Rahmenbedingungen und Perspektiven. • Peter Heinz Rothmann, Bayerisches Staatsministerium für Unterricht und Kultus: Schulbibliotheken in Bayern: Leseförderung in der multimedialen Schulbibliothek. • Willi Treichler, Landesbibliothek Bern: Schulbibliotheken in der Schweiz zwischen Wandschrank und Internet. • Brigitte Kühne, Direktorin der Universitätsbibliothek Växjö: Vom Kindergarten zur Universität: Bibliotheken im Lifelong-Learning-Prozess. Entwicklungstendenzen in Schweden. 2) Neue Medien: Urheberrecht, Internet und e-Learning • Isolde Müller, UB Graz: Das geistige Eigentum als Schleuderware? • Wolfgang Modera, AEC Linz: eLearning-Bildungsportal - Verwertungspolitik und Verwertungsstrategien von NM(Neuen Medien)-Produkten. 3) Kooperationen mit öffentlichen und wissenschaftlichen Bibliotheken • Werner Molitschnig, Förderungsstelle für Erwachsenenbildung Kärnten: Bewährte Synergien durch Kooperationen von Schul- und öffentlichen Bibliotheken. • Ute Bergner, UB Graz und Roswitha Karpf, Graz: Modell einer Kooperation zwischen wissenschaftlichen Bibliotheken und Schulen. 4) Die multimediale Schulbibliothek in Österreich • Werner Schöggl, Wien: Die multimediale Schulbibliothek - Ort des Dialogs. • Margit Macho, Klagenfurt: Managementfunktionen in der multimedialen Schulbibliothek. • Bruno Sygmund, Klosterneuburg: Die multimediale Schulbibliothek - das pädagogische Zentrum des Unterrichts. • Wendelin Hujber: Bestandsentwicklung in der multimedialen Schulbibliothek. • Stefan Hofer: Geänderte Anforderungen verlangen geänderte Ausbildung was muss ein (Schul)Bibliothekar heute können? 11

am 30. 8. 2003. 74

Wissenschaftliche Bibliotheken im Kontakt zu Schulen und Schulbibliotheken •

Michael Sporer, Innsbruck: Die Rolle der Schulbibliothek im Rahmen des IT-Schwerpunktes am BG/BRG Sillgasse-Innsbruck.

Eine weitere Möglichkeit der Kooperation mit Schulen und Universitäten bieten die 50 Österreichbibliotheken in 23 Ländern an. Diese wurden vom Bundesministerium für auswärtige Angelegenheiten seit 1985 vornehmlich in den ost- und südosteuropäischen Staaten eingerichtet.12 An vielen Schulen und Universitäten der Bibliotheksstandorte werden österreichische Bücher von den Lehrenden in den Lehrbetrieb einbezogen; die Bibliotheksbestände werden ebenso von Schülerinnen und von Studierenden genutzt. Lehrplan und Bibliotheksnutzung In dem seit dem Schuljahr 2000 gültigen österreichischen Lehrplan für die Unterstufe (= Sekundarstufe I) der allgemeinbildenden höheren Schulen (=AHS, die österreichische Form des Gymnasiums) ist ausdrücklich die Nutzung von Bibliotheken verankert: ,Mit Ordnungs- und Suchhilfen vertraut werden; Bibliotheken, Medien bzw. andere Informationssysteme zur Erarbeitung von Themen nützen."n Wenn dieses Lernziel auch nur im Deutsch-Lehrplan genannt ist, so zeigt schon die Formulierung, dass es adäquat für jeden anderen Unterrichtsgegenstand gelten kann und soll. Die multimediale Schulbibliothek ermöglicht es den Schülerinnen, nicht nur die erforderliche Lesekompetenz zu erwerben, sondern bereits in der Unterstufe des Gymnasiums entsprechende Recherchestrategien und den gezielten Umgang mit Sachliteratur zu erlernen. Der zitierte Passus aus dem Lehrplan bezieht sich aber selbstverständlich auch auf die Nutzung öffentlicher Bibliotheken, umso mehr, da die Errichtung einer Schulbibliothek in Österreich nicht verpflichtend ist. In der Oberstufe (=Sekundarstufe II) sollen die Schülerinnen in logischer Konsequenz zu selbständigen Nutzerinnen von wissenschaftlichen Bibliotheken ausgebildet werden, was im aktuellen Entwurf für die ab 2004 geltenden Lehrpläne anklingt: „Der Deutschunterricht [...] vermittelt Methoden und Kompetenzen wissenschaftlichen Arbeitens. Diese Fähigkeiten gewährleisten Studierfähigkeit und stellen die Grundlagen für lebenslanges Lernen dar [...] Der Deutschunterricht vermittelt Mediennutzungskompetenz, d.h. die Fähigkeit, sich der Medien zielgerichtet und funktional zu bedienen. [Die Schülerinnen sollen] Bibliotheken, Universitäten und Museen besuchen ",14 12 13 14

. Deutsch 1.-4. Klasse, Lehrplan 2000 für die AHS, S. 31-34. Entwurf Deutsch-Lehrplan für die Oberstufe, Stand 29.4.2003, deutsch.at/lehrplan/index.htm> am 15. 8. 2003.

75