Measuring Behavior [Second ed.] 0675080789

Citation preview

Measurmg

Behavior Second Edition .

John O. Cooper No.

Digitized by the Internet Archive in 2024

https://archive.org/details/measuringobehavio0D000coop

The Charles E. Merrill Series on

Behavioral Techniques in the Classroom Thomas M. Stephens, Consulting Editor

John O. Cooper

Measuring Behavior (Second Edition)

John O. Cooper & Denzil Edge

Parenting: Strategies and Educational Methods

John P. Glavin

Behavioral Strategies for Classroom Management

Stainback, Payne, Stainback, & Payne

Establishing a Token Economy in the Classroom

Thomas M. Stephens

Directive Teaching (Second Edition )

Thomas M. Stephens

Implementing Behavioral Approaches in Elementary and Secondary Schools

Thomas M. Stephens

Teaching Children Basic Skills

Thomas M. Stephens

Teaching Skills to Children With Learning and Behavior Disorders

1]

ae SPECIAL EDUCATION AND REHABITILITATIO} N

=

30STON

Measuring

}

aoe

UULLEGE

Behavior

Second Edition

John O. Cooper The Ohio State University

Charles E. Merrill Publishing Company A Bell & Howell Company Sydney London Toronto Columbus

wen

Published by Charles E. Merrill Publishing Company A Bell & Howell Company Columbus, Ohio 43216

Copyright © 1981 by Bell & Howell Company. All rights reserved. No part of this book may be reproduced in any form, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.

A previous edition was published under the title Measurement and Analysis of Behavioral Techniques. Copyright © 1974 by Bell & Howell Company.

Library of Congress Catalog Card Number:

ISBN:

80-81787

0-675-08078-9

Printed in the United States of America

{23

4.5

6° 7 98" 9510

GF S860 85

S4anssees2 een

To Bunny, Chris, Sharon, Greg, and Brian

Ma

a

Contents Foreword Preface

ix xi

Introduction

Part 1

Part 2

Measurement

6

1

Direct Measurement of Permanent Products

2

Observational Recording

Graphic Presentations

Techniques for Behavior Analysis

96 98

Appendix

A _ Instrumentation for observational Recording 186

Appendix

B_

Glossary Index

202 207

70

72

Analytic Teaching 4

8

32

Reporting Effects of Instruction 3

Part 3

1

Preparing for Printing by Twila Johnson

196

7é

i

Foreword

As Professor Cooper correctly notes in the preface to this book, recent progress in using applied behavioral technology in classrooms has been considerable. Those advances have been incorporated in this edition. Because of his methodical presentation, this text is an indispensable source for those school practitioners who take seriously a data-based approach to their instruction. An improved technology of instruction should be of concern to all school personnel. Economic considerations are the first grounds for such concern. Spiraling instructional costs demand that we have sound methods for identifying effective classroom treatments. No longer can we afford the luxury of instructional fads and wishful thinking about how to accelerate academic learning. Professionally, we never could afford such wasteful practices. But now, and in the long-range future, we must be accountable for how our limited resources are expended. Professional considerations are also important. Sound teaching can make a difference in children’s learning. Teachers now have available to them technology which can modify and reduce the frequency of maladaptive, inappropriate, and incorrect responses. They also have the ix

x

Preface

power to help students maintain what they have learned. Today, those teachers who are prepared to use the technology presented here will be equipped to teach in ways that were unknown just a few decades ago. Rate of learning is increasingly important in a world where each day an increasing amount of knowledge must be acquired. Social and academic concepts and skills are typically acquired more rapidly through applied behavior analysis than through more conventional instructional approaches. Each

text in the Merrill Series on Behavioral

Techniques for the

Classroom is devoted to a particular aspect of schooling or behavioral methodology. In this book, tactics and procedures are presented for measuring and analyzing school-related behaviors. John Cooper has made an important contribution to teacher training and, thereby, to improving classroom instruction of children and young people. Readers will find in this book specific tactics for observing, measuring, and analyzing classroom behavior. In the past, teachers have expected others to conduct evaluations of their students’ performances. Even when others performed these services, results were often unusable because by the time evaluations were completed, the evaluated students were often no longer under the teacher’s instructional care. Consequently, instructional modifications for those particular students often did not occur. The approaches presented in this text

will enable

teachers

to conduct

observations,

evaluate

re-

sults, and—when necessary—modify their instruction. No longer must teachers wonder, without ever knowing for sure, whether their instruction is effective. This book will give you sound ideas about evaluating instruction in practical ways. It will inform you, but by itself it cannot change your behavior. For that, you must practice the suggestions it contains. Thomas M. Stephens Consulting Editor

Preface

This text presents tactics for behavioral analysis. Generally the book addresses all school personnel (e.g., principals, school psychologists, counselors, therapists). However, it is designed specifically for persons preparing to become regular classroom teachers and special education teachers. Measurement and analysis are the key to successful teaching. The application of laws and principles of behavior in themselves generates no assurance of behavior change. Many programs in education may be employing specific technologies because of fad or custom rather than verified effect. The application of a technology of teaching should be based on continuous scientific verification at the classroom level by classroom teachers. This book is designed to serve as a basic textbook in measurement and analysis for directive teachers and also for seniors and beginning graduate students in education and psychology. Since the publication of the first edition (Measurement and Analysis of Behavior

Techniques),

much

progress

has been made

in applying

behavioral analysis techniques in the classroom. This revision was undertaken to incorporate these changes. Among the recent developments Xi

xii

Preface

included in this revision are: the delayed multiple baseline design, the multiple baseline probe design, the changing criterion design, and the multielement baseline design. In addition, the measurement chapters have been updated to include a discussion of trials to criterion, measurement of the complexity of task items, pre- and posttests, and probes. The topic of measurement and analysis is covered in three parts. Part 1 outlines measurement techniques for the directive teacher. Two types of measurement are presented—direct measurement of permanent products and observational recordings. Part 2 is concerned with graphic presentations for teacher use. Part 3 presents some basic procedures for analyzing teacher effectiveness. I am grateful for the support and assistance of many colleagues, teachers, and students in completing this book. Special appreciation is extended to Thomas M. Stephens, friend and colleague. John O. Cooper

Introduction

A Ph.D. aspirant was once requested to express to his oral examination committee his views on how children would be educated in the year 2000. The student described several innovative approaches to education that he predicted would be common by the year 2000. After the candidate finished his discourse, an elder member of the examination committee stated that all instructional approaches espoused by the student were employed in his university elementary laboratory school during the 1930s. Why, he asked the student, do we have to wait until the year 2000 for these innovations to become common in educational settings if they were already applied forty years ago? In light of the committee members’ remarks, the student concluded that educational practices in the year 2000 would probably differ little from those employed in 1970. Risley (1969) made a similar observation. “Procedures which we are now using and empirically demonstrating to be effective were used over a century ago, but have been discarded in the interim.” Risley supports his statement by comparing the following:

1. Rousseau’s hypothetical account of rearing Emile with the work 1

yD

Introduction

of Hart, Allen, Buell, Harris, Risley, and Mees in 1964;

and Wolf

in 1964

and Wolf,

2. Itard’s procedures for teaching the Wild Boy of Aveyron with the work of Lovaas in 1966 and 1967;

3. Seguin’s school procedures for the education of retarded children with that of Sherman in 1965, Baer, Peterson, and Sherman in 1967, and Meyerson, Kerr, and Michael in 1967.

Many other examples in the literature show how procedures for educating children have changed very little from generation to generation. Of

course,

we

have

introduced

audiovisual

instrumentation,

better-

quality printing, modern facilities, and so forth, but the actual instructional interaction between teacher and student has changed little since the time of Socrates. The history of education has not demonstrated a cumulative development in educational approaches. Probably this is a result of basing many educational changes on historical accidents, untested theories, and the

opinions of influential individuals. Yet one might wonder, surely some educational change has been based upon sound experimentation and empirical evaluation? Just look at the number of yearly M.A. and Ph.D. theses written in colleges of education that end with the statement: The implications for the classroom teacher are—one,

two, three, and four.

Has this had no effect on changing our educational endeavors? Look at the number of journals of education reporting research. With the abundance

of research in education, has there not been a cumulative

development of instructional approaches? Why have we predominately used historical accidents, untested theories, and opinions as bases for educational change?

To find answers to these questions, we must look to the statistical models used in educational measurement and evaluation. The measurement and evaluation tactics used most in education are concerned with groups of students and how one group compares with another or with the relative standing of a student in a particular group. This concern with group analysis is a paradox because a basic tenet of American public education has been the importance of the individual student. Yet most of our measurement and evaluation procedures are designed for

groups of students. This approach obscures the unique qualities of the individual and provides the teacher with no relevant instructional information. As Kunzelmann stated, “Useful information to educators is that which demonstrates differences within an individual. Building toward an educational system which is sensitive to individual differences allows for the isolation of environmental influences or instructional opinions which contribute to and thus change classroom performance”

Introduction

3

(p. 17). However, the reader should not infer that group measurement and analysis procedures are not relevant to certain school endeavors. They allow school officials to draw general conclusions, to summarize data in convenient form, to make predictions on how effective a procedure will be for a given group of students, to anaylze the effectiveness of educational institutions, and to provide sources of information for public relations. All of the above outcomes are important for administrative decision making, yet they provide no relevant individual instruc-

tional information for teachers. This book addresses the subject of a technology of teaching individual students. It is not new procedures or approaches to education that are setting the occasion for this development but rather tactics for measurement and analysis of individual behavior. Designs for the scientific analysis of individual students have only recently been developed. Mussen, Conger, and Kagen (1963) concurred by stating that we have developed methods for scientifically analyzing group data but have not yet generated tactics for analysis of individual cases. Specifically, the procedures presented in this book will provide the classroom teacher with tools to answer questions such as the following: Was I effective in teaching John two-column addition? Would he have acquired this skill without my intervention? Could another teaching procedure be more effective? What materials set the occasion for increased rates of correct student performance?

What effect do different seating arrangements have on academic performance? Is role playing producing changes in the social behavior of my students? The emphasis throughout this text is on the application of measurement and analysis tactics in classroom situations. First, a description of the tactic is presented. Second, an actual applied example of a classroom application is given. Finally, a discussion of the considerations of application follows the applied example. This form of organization is used in the hope that information as to how a measurement or analysis technique is used by teachers will enable the reader to know the purpose of the technique and the circumstances in which it should be used. All techniques presented in this book are described with narrative examples and, whenever possible, with graphic illustrations. Student evaluation exercises are presented in the body or at the end of each chapter. All evaluation exercises should be answered correctly before progressing further in the text. Suggested applied assignments for practice in using the techniques follow the evaluation exercises. In most

4

Introduction

cases, the reader’s knowledge and skill in using the tactics under discussion will be increased if the applied assignments are performed. A final applied assignment is given at the end of the book which allows the reader to integrate most of the applications that have been presented. The reader should plan to spend approximately eight weeks to complete this assignment. It should be noted that in several places in chapter 2 the reader is referred to appendix A. This appendix contains a description of instru-

mentation used in observation recording. Appendix A is a crucial part of chapter 2 and should not be slighted.

References Baer, D. M., Peterson, R. F.,

& Sherman, J. A. The development of imitation

by reinforcing behavioral similarity to a model. Journal of the Experimental Analysis of Behavior 1967, 10, 405-16. Hart, B. M., Allen, K. E., Buell, J. S., Harris, F. R.,

& Wolf, M. M. Effects

of social reinforcement on operant crying. Journal of Experimental Child Psychology 1964, 1, 145-53. Itard, J.-M. G. The wild boy of Aveyron. New York: Appleton-CenturyCrofts, 1962. Kunzelmann, H. P. Precision

teaching.

Seattle:

Special Child

Publications,

1970. Lovaas,

O. I., Bererich, J. P., Perloff, B. F., & Schaeffer, B. Acquisition of

imitative ‘speech by schizophrenic children. Science 1966, 151, 705-7. Lovaas, O. I., Freitas, L., Nelson, K., & Whalen, C. The establishment of imitation and its use for the development of complex behavior in schizophrenic children. Behavior Research and Therapy, 1967, 5, 17181. Meyerson, L., Kerr, N., Michael, J. Behavior modification in rehabilitation. In S. W. Bijou & D. M. Baer (Eds.), Child development: Readings in experimental analysis. New York: Appleton-Century-Crofts, 1967. Risley, T. R. Behavior modification: An experimental-therapeutic endeavor. Paper prepared for Banff International Conference on Behavior Modification, April 1969. Rousseau, J. J. Emile. London: J. M. Dent and Sons, Ltd., 1948. Seguin, E. Traitement morale, hygiene et education des idiots. Paris: J. B. Battiere, 1846.

Sherman, J. A. The use of reinforcement and imitation to reinstate verbal behavior in mute psychotics. Journal of Abnormal Psychology, 1965, 70, 155-64. Wolf, M. M., Risley, T., & Mees, H. Application of operant conditioning

procedures

to the behavior

problems

Research and Therapy, 1964, 1, 305-12.

of an autistic child. Behavior

Part

1

Measurement

Until you’ve measured it you don’t know what you're talking about. —Lord Kelvin

Parents, school officials, and state departments of education are

requiring classroom teachers to demonstrate the effectiveness of their instruction. In order for teachers to be accountable, it is necessary for them to possess evaluation skills. Evaluation is based on measurement, which is the way to determine student growth. The following case about Miss Lucas and Gary illustrates one way measurement can be used in the classroom. When Gary entered first grade, his teacher, Miss Lucas, observed that his class assignments were highly accurate but almost always incomplete or late. His peers apparently had no problem in meeting most task requirements on time. Miss Lucas recognized that completing classroom work on time was very important for Gary’s future academic success. Therefore, Miss Lucas wanted an objective, quantitative record of the number of assignments completed on time, since subjective impressions of student work may not be accurate. She knew that without reliable measurement she might subjectively feel that Gary had improved even though his performance remained unchanged. So Miss Lucas began daily data collection by tallying the number of assignments made and the number of assignments Gary completed on time. At the end of one week of measurements, objective statements concerning Gary’s work could be made. Miss Lucas gave twenty assignments, out of which Gary completed eleven on time. His performance was not as bad as Miss Lucas had anticipated, but it was obviously a problem behavior. With this baseline information, Miss Lucas planned an educational strategy to help Gary complete his class assignments on time and had data to make objective comparisons concerning Gary’s improvement during the year. By measuring specific student performance at certain points in time, we can determine which academic and social responses have been learned and which behaviors still need to be learned. Also,

through measurement tactics it is possible to analyze the effectiveness of our teaching. Two ways of measuring student performance are (1) obtaining and examining permanent products and (2) observational recording. Chapter 1 presents techniques for measuring permanent products. This chapter includes a discussion of frequency, rate, and percentage measures, trials to criterion, complexity of task items, pre- and

posttests, and probes. Chapter 2 presents tactics for observational recording, event recording, duration recording, interval recording and momentary time sampling.

Direct Measurement of Permanent Products

Direct measurement

of permanent products is perhaps the most com-

mon measurement tactic used by classroom teachers today. This method of measurement has a long history. When teachers grade a written examination or written responses in a workbook, they are measuring permanent products. Certain types of behaviors result in products that can be measured following student responses. Other examples of permanent products include audio- or videotapes of student performance, written arithmetic computation, written spelling words, written alphabet letters, colorings, completed puzzles, strung beads, and stacked blocks.

Figure 1-1 shows a classroom example of a permanent product. The permanent products in figure 1-1 are the written responses to computational problems. Teachers usually translate permanent products into numerical terms of (1) frequency of occurrence (John worked six of his problems correctly), (2) rate of occurrence (John read 100 words per minute), or (3) percentage of occurrence (John worked 90% of his problems correctly).

Direct Measurement

of Permanent Products

9

Mixed Addition and Subtraction Combinations.

Add: 9

6

ee

ee

3)

5]/

10

7

8

9

Eee

4

5

1S

Bt

6

9

5

7

5

8

%

I!

val

IZ

3y/

3

9

5

3

Tl

NS

2

a

6

4

8

7

5]

Te

4)

7 6

5 ii

in get os

i

5

ee

6I

re

3

3

5

5

ID

wi

|

4)

6 3

6 a8

9 =

7 8

3 5

9 6

re

Th

5]

5.

Sao

9 eo

13 8

5 4

14 T

12 8

Me g Fee

5

TD

Subtract:

13 Le

8 1

11 5

ie

ba

12

11

14

7 8

V 3

4 1

/ 14

9

Ree oO

ee

15

Zz 10 Le

14 &

/ B

13

ew We

Begin Lesson

FIGURE

6

1-1.

3

eS DD.

0 0

10 5

O

5

6

7

9

oe

0

8

5 pe)

aie

oO

“f

2

7

ee 6

a

3

2

10

oO

q

End Lesson

Classroom Example of a Permanent Product

r with frequency, Most teachers and prospective teachers are familia s may not use teacher times many Yet rate, and percentage measures. r 1 looks at Chapte age. advant tional those measures to the best instruc

10

Chapter 1

1. How to use frequency, rate, and percentage measures in classroom situations 2. Which parameters should be used in selecting a measurement tactic for permanent products

Frequency of Response Frequency of response is the number of times a specific behavior occurs

in some

period of time. Tallying, or counting, responses

is possible

when the behavior is readily observable and when it is discrete; that is,

when it can be separated from other responses. Applied Example McKenzie,

Egner,

Knight,

Perelman,-Schneider,

and

Garvin

(1970)

employed frequency counts to measure correct addition problems in a first-grade class. Each school day, 20 students were assigned worksheets containing 50 addition problems of two single digits. The same 50 problems were used daily..A bell timer was set for 2 minutes for each work session. Children were instructed to begin computation on cue and to stop at the sound of the bell. Figure 1-2 shows one student’s frequency of correct responses during 8 days of baseline. Note that during baseline the frequency of correct addition responses ranged from 18 to 32. Considerations

The McKenzie study provides an excellent example of the appropriate use of frequency measures. The number of addition problems (opportunity for response) was always 50 per day. Time was constant at 2 minutes per session. With these two variables controlled, the teacher was able to compare student performance over time (e.g., sessions, days, weeks). For example, the frequency of the student’s correct addition responses

(figure 1-2)

was approximately

18, 18, 31, 29, 29, 30, 32

and 27 for 8 consecutive days. These data demonstrated no improvement in addition skills during the last 7 days.

Frequency of occurrence should be used as a measure of behavior only when opportunity for response and time are constant. If frequency

measures are utilized without reference to opportunity for response and time variables, student performance is left to subjective interpretation. The problem of using frequency measures without controlling opportunity for response and time is illustrated in the following case study.

Direct Measurement of Permanent Products

11

50

$ Cc

°

> oO io

40

6 S) x)

30

8

—& =} Z

20

10

Days

FIGURE 1-2. Frequencies of Correct Addition Responses by One Student in Daily 2-Minute Periods During 8 Days of Baseline. (Adapted from “Training Consulting Teachers to Assist Elementary Teachers in the Management and Education of Handicapped Children” by H. S. McKenzie et al. Exceptional Children, 1970, 37, 141. Copyright by the Council for Exceptional Children. ) Miss Johnson was a teacher of primary-age learning disability children. John, one

of her students, tended to reverse the letters b, d, p,

and q. Miss Johnson was interested in knowing if her instruction was improving John’s reversal problem. Therefore, on Monday, Wednesday, and Friday of each week she made frequency counts of the number of reversed b, d, p, and q letters occurring in John’s written assignments. At the end of 3 weeks, Miss Johnson had recorded the occurrences of

reversals shown in table 1-1. From the data of table 1-1, could Miss Johnson conclude that John’s reversal problem had improved? He obviously was emitting fewer reversed

b, d, p, and q letters during each consecutive week.

However,

from the recorded data, she could not know if the problem behavior had improved because she did not take into consideration John’s opportunities to make b, d, p, or q letters. For example, on Wednesday, week one, John emitted 20 reversals. Yet he may have had 100 opportunities to write b, d, p, and gq. On Wednesday,

week two, John made only 2

reversed letters. But perhaps there were only 3 possible opportunities

12

Chapter 1

for reversals. Were the 2 reversals in week two an improvement over the 20 reversals in week one? In order to answer this question, we must look at rate of response. TABLE

1-1.

Number

of Reversed Letters

Days

Weeks

1

2

3

Monday Wednesday Friday

15 20 9

9 2 10

5 7 6

TOTAL

44

PAA

18

Rate of Response Rate of response is the frequency of occurrence of a behavior during a certain unit of time. Rate is calculated by dividing the total number of correct or error responses by the amount of time spent producing those frequency responses (rate = ). Rate is usually expressed in responses time per minute. Note that frequency is represented by the number of responses while time is the amount of minutes in which all of the responses occurred. Applied Example Hopkins, Schutte, and Garton

(1971) reported rate measures for print-

ing and writing responses of 14 first-grade and 10 second-grade students. The first-grade students copied printed assignments that were typically composed of phonetic drills of descriptions of current events. These first-grade assignments averaged 194 letters. Second-grade children used cursive writing to copy excerpts from stories or poems such as Hiawatha. The average length of the second-grade assignments was 259 letters. To derive rate of occurrence, Hopkins et al. used measures of duration and frequency of occurrence. Duration was defined as the elapsed time from the teacher’s instruction to begin copying the assignment until the student brought his paper to her desk. Frequency was the actual number of letters printed or written by the students. Figure 1-3 presents mean baseline rates of occurrence of printed letters per minute printed by first-grade children. Each data point represents the mean, averaged over all children for that day. The horizontal

Direct Measurement

of Permanent Products

13

dashed line represents the average of the daily means over all days. Note in this figure that the first-grade children printed, on the average, approximately 6 letters per minute.

13 12 2 = = =

5 a Ss a % é oO

2

11

10 9

8

=]

Zi i=} ioe}

oO

=

Days

FIGURE 1-3. Mean Number of Letters Printed per Minute by a Firstgrade Group. (Adapted from “The Effects of Access to a Playroom on the Rate and Quality of Printing and Writing of First- and Second-grade Students” by B. L. Hopkins, R. C. Schutte, and K. L. Garton, Journal of Applied Behavioral Analysis, 1971, 4(2), 81. Copyright 1971 by

the Society for Experimental Analysis of Behavior, Inc.) Figure 1-4 gives the mean baseline rates of occurrence of cursive letters per minute written by second-grade children. In figure 1-4, the second-grade children wrote, on the average, approximately seven letters per minute. Again, each data point represents the mean, averaged over all children for that day. Considerations

Rate of occurrence is considered the basic datum of a science of education (Skinner, 1966). Rate is a preferred measure when evaluating permanent products such as written material. The amount of teacher

14

Chapter 1

time needed in monitoring rate measures can be a problem for classroom teachers. One way to reduce monitoring time is to have students time their own behavior by asking them to note the time they begin and end a particular task. Instructions for recording may be printed on assigned material or given verbally. For example: Time started—10:00; Time stopped—10:25; Total time—25. The teacher or the students can calculate rate of response. Occasionally, students will not have acquired skills in “telling time.’”” When this occurs, the teacher can draw

two

clocks

without

hands

on

the

students’

materials.

(Some

teachers have a rubber stamp of a clock face without hands for this

17 16 15 >)

14

_

&

=

13

2A

12

s i 6

2

SI Z

10

=}

i=)

a =

Days

FIGURE 1-4. Mean Number of Letters Written per Minute by a Second-grade Group. (Adapted from “The Effects of Access to a Playroom on the Rate and Quality of Printing and Writing of Firstand Second-grade Students” by B. L. Hopkins, R. C. Schutte, and K. L. Garton, Journal of Applied Behavioral Analysis, 1971, 4(2), 83. ance 1971 by the Society for Experimental Analysis of Behavior, ne.

Direct Measurement

of Permanent Products

lS}

purpose.) Students could then be taught to copy the position of the big and little hands on the classroom clock when they start and when they stop. For example: 1. Worksheet before lesson

Start

Stop

Rate measures are sentive to efficiency of student performances. As Kunzelmann (1970) states, ‘““The student who worked 10 minutes and got 80 problems correct with a correct rate of 8 problems per minute on the 100-item test is much more proficient than the student who took 20 minutes, also got 80 problems correct, but who had a correct rate

of 4 problems a minute” lems

were

the same

(p. 31). (This statement assumes that prob-

or of similar difficulty.)

Basically,

then, rate is

important if data concerning level of student performance is to be meaningful. Also, of all measures of permanent products, rate is most sensitive to the effects of teaching tactics on student responses because it will reveal very small increments of behavior change. Remember, Miss Johnson asked the question, “Were the 2 letter reversals in week two an improvement over the 20 reversals in week one?” Miss Johnson cannot answer this question because time and opportunity for response were not recorded. Let’s suppose that Miss Johnson took into consideration (1) the opportunity for writing letters b, d, p, and q; (2) the amount

of time John spent on the written assign-

16

Chapter 1

ment; and (3) the number of nonreversed b, d, p, and q letters as well

as the number of reversed letters. With these three types of information, Miss Johnson can easily demonstrate whether John’s reversal problem did or did not change. Concerning opportunity for response, John’s written assignment was comparable in number of words on the 9 measurement days reported earlier. The smallest number of b, d, p, and q letters occurring in the

assignments was 30 and the maximum was 41. The 9-day average number of b, d, p, and q letters was 33. Therefore, there was only a small variance in the opportunity for John to write his letters. At no time did he reach the limit on opportunity for response. John’s reversed and nonreversed letters and time worked on the assignment are presented in table 1-2. TABLE

1-2.

Frequency of Reversed and Nonreversed Letters

Days

Reversed Week

Monday Wednesday Friday

Monday

2 1 1

3 min. 5 min. 2 min.

9

5

2 min.

2 10

9 (5)

2 min. 5 min.

15 20 23

5 min. 6 min. 4 min.

Week Monday Wednesday Friday

Time

1

15 20 9 Week

Wednesday Friday

Nonreversed

5 7 6

2

3

Since the amount of time that John worked on his assignment varied from day to day, Miss Johnson converted John’s frequency of letter writing to rate of letter writing per minute. Miss Johnson was correct in converting to rate of response. If she had not, a day-by-day comparison of John’s letter responses could not be made. By analogy, a high school English teacher might say that last month Bill read at a rate of 250 words per minute, but this month he is up to a rate of 400 words per minute. Bill’s reading performances could not be compared unless the teacher had converted words read to words read per unit of time.

Direct Measurement of Permanent Products

17

Miss Johnson converted to rate of response by dividing the frequency of response by time. For example on Monday, week one, John reversed 15 letters in 3 minutes. This is a rate of 5.0 reversed letters per minute (15/3 = 5). The number of John’s nonreversed

b, d, p, and q letters

on Monday, week one, was 2. This is a rate of 0.7 nonreversed letters per minute (2/3 = 0.67 = 0.7). John’s rates per minute of reversed and nonreversed b, d, p, and q letters are shown in table 1-3. TABLE

1-3.

Rate of Reversed

Day

Reversed Week

Monday Tuesday Wednesday

Monday Tuesday Wednesday

Nonreversed

0:7* O2* 0.5

2 5.0 1.0 2.0

Week

Letters

1 5.0 4.0 5.0

Week

Monday Tuesday Wednesday

and Nonreversed

3.0 5.0 3.0

3 1.0 1.0 220

3.0 3.0 6.0

Note. The rate of 0.7 (7/10 of a response per minute) could be read as 7 nonreversed responses in 10 minutes. Similarly a rate of 0.2 (2/10 of a response per minute) would be the same as 2 responses in 10 minutes.

Miss Johnson can now state with some assurance that John’s reversal problem is improving. The rate of nonreversed b, d, p, and q letters is

accelerating and the rate of reversed letters is decelerating in occurrence.

Percentage of Response Percentage of response is a ratio that expresses the amount of a behavior as a certain number of responses per every 100 responses. Percentage is usually obtained by dividing the total opportunities for response (e.g., 25 test items) into the number of correct or incorrect responses (e.g., 17 correct answers) and multiplying that result by 100 (17/25 = .68 X

100 = 68%).

18

Chapter 1

Applied Example

Hopkins et al. (1971) were concerned with the relationships between correct and incorrect letter responses. Each student’s letter responses were scored by drawing circles around incorrect responses. Criteria for incorrect responses were as follows: . All omissions of assigned letters . Substitutions of letters in place of those assigned . Reversals, e.g., printing or writing A instead of K . Omitting any part of a letter, e.g., failing to cross a ¢t . A failure of any vertical stroke to be within i5 degrees of the APWNP perpendicular to the baseline (printing) or failure of written letters to be consistently slanted about 60 degrees from the baseline 6. Short lowercase letters being less than half as tall as the distance between the baseline and the centerline or being more than 50% taller than the distance between the baseline and the centerline 7. Tall lowercase or uppercase letters being less than 75% as tall

as the distance between the baseline and top line or being more than 25% taller than this distance 8. Lowercase letters, which-should extend below the baseline, e.g., y, failing to extend below the baseline at least 50% of the distance between the base- and centerlines or extending further than 100% of this distance (p. 79)

The total number of errors was tallied and written at the top of each student’s copy paper. Hopkins et al. reported the mean proportion of errors each day. This was computed by dividing the total number of errors by the total number of letter responses. Percentage was then obtained by multiplying the quotient by 100. Figures 1-5 and 1-6 show the percentage of errors during baseline for the first and second grades. Considerations

Ideally, if the total number of responses is less than 100, percentages should not be computed (Guilford, 1965). In the Hopkins et al. (1971)

study, the first-grade class averaged 194 letter responses; the secondgrade averaged 259 responses. Guilford states that if a lower limit must be set, it is unwise to compute percentages when the divisor is less than 20. The addition of only one frequency in the numerator will produce a corresponding change of at least 5 percent when the divisor is less than 20. For example, suppose that a student answered 11 questions

Direct Measurement of Permanent Products

19

correctly on a 13-item test, he would have scored 86 percent correct (11/13 X 100 = 86% ). However, had the student answered one addi-

tional

item

correctly,

his percentage

would

have

been

92 percent

(12/13 XK 100 = 92%), an increase of 6 percent! When this situation occurs, it is recommended that a ratio of total correct and total oppor-

tunities (12/13) be used, rather than percentage. 22 Baseline

Errors of Percentage

Days

FIGURE I-5. Percentage of Errors in Printing by a First-grade Group. (Adapted from “The Effects of Access to a Playroom on the Rate and Quality of Printing and Writing of First- and Second-grade Students” by B. L. Hopkins, R. C. Schutte, and K. L. Garton, Journal of Applied Behavioral Analysis,

1971, 4(2),

81. Copyright

1971

by

the Society for Experimental Analysis of Behavior, Inc.) One other caution should be noted in using percentages. Frequently teachers incorrectly report a mean or average percentage that was calculated from other percent scores. For example, suppose Jill’s percentage scores were as follows: 72%, 88%, 80%, 100%, and 90%. It would be incorrect to figure Jill’s average percentage correct score

for the five tests as follows:

20

Chapter 1

72% 88% 80% 100% +-90% 430%

+ 5 = 86%

To calculate the correct mean percentage for the five tests, we must use frequency scores. If the total of possible correct responses for Jill’s five tests was 150, and she responded correctly to 132 test items, her mean or average percentage correct would be 88%. number of correct responses

-

=

total possible correct -

132 150

= 0.88 K 100 = 88%

average

Baseline

6

5

uO Re) o

=

q

3

2

iar 1 0

2

4

6

8

Days

FIGURE 1-6. Mean Percentage of Errors in Writing by a Secondgrade Group. (Adapted from “The Effects of Access to a Playroom on the Rate and Quality of Printing and Writing of First- and Second-grade Students” by B. L. Hopkins, R. C. Schutte, and K. L. Garton, Journal

of Applied Behavioral Analysis, 1971, 4(2), 83. Copyright the Society for Experimental Analysis of Behavior, Inc.)

1971 by

Direct Measurement of Permanent Products

21

There are two advantages to using percentages. First, the use of percentages is an efficient way to deal with large numbers of responses, and second, most people easily understand percentages. But there are also disadvantages. First, percentages do not say anything about efficiency; percentage may increase while error rate remains the same or increases. Second, a percentage figure may not be as easy to understand as a simple tally of correct or error responses.

Variations on Frequency, Rate, and Percentage Measures Techniques selected for measuring academic performance must be relevant to the purpose of collecting data. For example, student responses to items on a formal achievement test may not provide sufficient information for planning individual programs of instruction. No matter what technique is selected, measurement must provide data on student responses to the actual materials used during the instructional process,

and it must be continual; that is, frequent samples of

performance must be obtained. The basic procedures for direct and daily measurement which have been described are frequency, rate, and percentage. These basic procedures can be applied in a variety of ways. Four widely used variations have been described by Cooper and Johnson (1979): (1) trials to criterion, (2) complexity of task items, (3) pre- and posttests, and

(4) probes. Trials to Criterion

Trials to criterion measures the number of times response opportunities are presented before the student achieves a preestablished level of competence or proficiency. For example, the measurement could be reported as 10 trials to achieve 100% accuracy (percentage) or 10 trials to achieve 60 words per minute (rate). In all cases the criterion set will probably require a measure of proficiency during each trial. Measuring trials to criterion is appropriate when making an evaluative comparison of two or more kinds of materials or instructional methods. For example, a teacher might report the number of times the word that was presented orally before the student read it inde-

pendently and compare that number to the number of times the word that was presented in written form to the student for tracing before he read it independently.

22

Chapter I

Applied Example

A common technique used in developing math skills is to have students say or write answers to a set number of problems until a specified rate correct has been demonstrated. For example, each day, the students give the answers to 50 multiplication problems. The teacher’s criterion is for the students to complete the 50 multiplication facts in one minute at 100% accuracy. The student responses in this illustration can be measured appropriately as number of trials or opportunities needed to meet the criterion of answering 50 multiplication problems in one minute at 100% accuracy. For instance, the teacher could report that Mary reached criterion in 11 trials and that Bill required 20 trials to reach criterion. Considerations

Measuring trials to criterion can give a teacher useful information for planning the amount of time that should be allotted for instruction. If trials to criterion are counted for several words taught to a student in different ways, the teacher.can compare performance to determine if the student masters words better with one of the teaching methods than she does with another.

Complexity of Task Items The response units suggested for previously discussed measures have been whole units (e.g., whole words read or spelled, finished mathematics problems). Measuring responses in whole units, however, does not take into account the complexity of each possible response. A method for measuring academic responses which would take complexity into account would be to count the operations necessary to achieve a correct response. These operations may be reported as frequency, percentage, or rate. For example, in measuring spelling performances, rather than count a complete word as correct or incorrect the teacher might want to consider the number of letters that are correct within each word. Applied Example Helwig

(1973)

number

of operations

reported

mathematics

response

sufficient to produce

rate

calculated

the answer.

from

The student

Direct Measurement of Permanent Products

was given 20 multiplication and division problems to session selected at random from a set of 120 problems. time for each session was recorded by the teacher. The were of the type a X b = c anda + b = c where the

23

work at each The length of 120 problems student had to

find a, b, or c. The student was asked to find one of the factors—the

product, dividend, divisor, or quotient. The problems required from one to five operations to find the answer. For example, to find the answer to the problem 55 X 5 = ? requires the student to perform four operations:

First operation

Second operation

55

2 oh)

__5 (Multiply the ones: ZOE

> = 25)

Third operation

__5 (Carry the two tens from 5

the first operation)

Fourth operation

38) __5 (Multiply the tens:

BS) __5 (Add the two tens from

259.2

275

.2¢50 = 250)

the second operation)

A correct answer would be scored as four correct responses. When there was more than one possible way to find the answer, a mean number of operations was figured for that problem. (For example, in the problem 4 X c = 164, the answer by multiplication requires two operations, by division four. The mean number of operations would be three.) Each set of 20 problems was checked and the number of correct responses and errors was recorded and reported as rate of response.

Pre- and Posttests One tactic used by teachers to measure students’ acquisition of academic skills is pre- and posttesting. The pretest is used primarily to provide an assessment of what the student can do before instruction. The posttest is given following instruction. Pre- and posttest measurement can be used in two ways. First, the teacher can report the student’s posttest score as a final grade. Second, the teacher can report the student’s score as instructional gain from pretesting to posttesting. To illustrate, say that a student scored two correct out of 20 on the pretest, and 10 correct out of 20 on the posttest. Therefore, the teacher would report a gain score (increase) of 8, or 40%

(8 + 20 = .40, or 40% ).

24

Chapter 1

Applied Example

Johnson and Bailey (1974) employed pre-, mid-, and posttesting in conjunction with direct and continuous measurement of basic arithmetic skills of kindergarten children. A 122-item skill-based arithmetic test was constructed to be used as a pre- and posttest measure and also as a midtest at two-week intervals. The following procedure was employed: Approximately 3 minutes at the end of each session were used for “review test”; i.e., the tutor simply went over all of the tasks worked on that day and recorded the student’s responses. The tutor then scored the tests and graphed the data, giving the student immediate feedback to his daily progress. When a student received a perfect score, on any task for three consecutive review tests, teaching of the task was discontinued and a new task was introduced. (p. 226) Considerations

A major advantage of pretesting is that it provides the teacher with a rough estimate of student skills in the instructional area. The pretest allows the teacher to modify his or her teaching topics to cover only skills that the student has not previously acquired. Also using of preand posttests requires very little of the teacher’s time. An obvious limitation of pre- and posttests is that both measures must be applied only to items that represent exactly the same unit of study and that do not represent more complex task requirements. Because only two measures are compared, the teacher must be sure that the pre- and posttests are equivalent in requiring the same kind, complexity, and manner of response. This limitation ceases to be a problem if the teacher combines pre- and posttest measures with some other form of continuous measurement.

Probes An important question of instructional effectiveness is whether the teaching of certain specific items has a generalized effect and improves students’ accuracy on similar untaught items. For example, if the teacher specifically teaches some words with a long vowel-consonantsilent e pattern, will the student correctly read other untaught words which follow the same pattern? To answer this question the teacher could periodically present untaught words, or probes, as well as the words being taught, and then record correct or incorrect responses to probes. Probe items must be related in form or function to items taught.

Direct Measurement of Permanent Products

2)

Applied Example

Clark, Boyd, and Macrae (1975) were concerned with teaching six disadvantaged youths to complete biographic information on job application forms. After training, students were tested on the three application forms on which they received training. Students were also tested on a fourth application form which had not been used in training (a probe form). Test results were reported in percentage of correct responses on all four application forms. The authors stated, “The tests show that after an item had been taught, it was correctly used in completing application forms on which the youths had been trained and forms on which they had never been trained” (p. 67). Considerations

The advantages of probe measures are similar to the advantages of pretests. They provide a rough assessment of student skills and require very little teacher time. Also, probes tell the teacher whether instruction on selected items has affected the student’s accuracy on similar items that were not taught. If accuracy on untaught responses increases, the teacher can say that the student has mastered the concept and that additional instruction is unnecessary. However, if accuracy on untaught items does not increase, the probe measures can be used as measures of baseline skill prior to teaching those items. Table 1—4 provides a summary of the measurement techniques discussed in chapter 1.

Evaluation

-

You may wish to evaluate how well you comprehend measurement of permanent products by responding to the following test items. Cover the answer column with a sheet of paper; then write your answers in the blank spaces provided. The correct answers can then be checked. written arithmetic computation coloring written spelling words strung beads

1. List four examples of permanent products. a. b. Cc; d.

after (subsequent to)

2. Permanent

products

can

be

the student’s

measured behavior.

26

Chapter I

TABLE

1-4.

Measurement Techniques for Permanent Products

Definition

Application

Data

Evaluation

FREQUENCY Simple count of correct or error responses occurring in some period of time

1.

2. 3.

Easily understood when communicating with parents and children Simple procedure

For day-to-day comparisons of data, total number of opportunities for response should not vary with

Generally

assignments

does

interfere 4.

not

with

on-

Reflects

going teaching Requires minimum

ficiency

academic

ef-

academic

ef-

amount of _ teacher time 5. When used with permanent products,

source of errors may be

analyzed

structional

for

in-

planning

RATE Frequency of a behavior during a certain unit of time. To compute,

1.

divide

the total number of correct error responses by the amount of time spent working on the material.

2.

Most sensitive measure of academic behavior, i.e., most like-

Measures

ly to changes

comparable if difference in length of sessions is extreme; e.g., rate computed from a 1-minute sample may be higher than that computed from a 30-minute sample for some students and some academic re-

show.

small

Excellent as criterion for mastery of skills

Some

parents

and

children may have difficultly understand-

4. 5.

Reflects ficiency

ing rate Approximate

for measuring all academic behaviors May require more teacher time to measure and compute

are

not

sponses

Student or

(not teachers

materials)

control possible

should

maximum rate of re-

sponse Accurately

student

reflects

efficiency

if

complexity of tasks does not change be-

tween

measures

PERCENTAGE ~

Amount

of behavior ex-

pressed as a certain number of responses in every hundred. To compute,

divide the number of correct or error. responses by total number of opportunities for repone then multiply by

Easily

understood

when communicating with parents and children

Student performance at sessions offering

unequal

response

portunities compared

can

opbe

Direct Measurement of Permanent Products

TRIALS Number of times response Opportunities are presented before student achieves a preestablished criterion

1.

3.

TO CRITERION

Easily understood when communicating with parents and children

2. An

appropriate

mea-

sure for making an evaluative comparison of two or more kinds of materials or instructional methods May give’ teachers useful information for planning time to be given to instruction

COMPLEXITY Measurement of the number of Operations sufficient for a correct response. When there is more than one algorithm for an answer, a mean

OF TASK

Can show if criterion as rate

efficiency is stated

In all cases the criterion set will prob-

ably

require

sure

for

a mea-

each

trial;

however, only the number of trials to achieve the criterion is usually reported.

ITEMS

Use when complexity of response varies greatly within one measurement episode or when a fine analysis of response accuracy is desired

number of operations can

Qi

Number of operations may be used to compute frequency,

percentage, or rate

be reported

PRE- AND Measurement of behavior before and after instruction

POSTTESTS

1.

Provides a minimal estimate of student skills in an _ instructional area 2. Requires a minimum of teacher time 3. Pretest used to provide an assessment of what student can do before instruction

4.

Measurement at intervals during instruction is critical if time between preand posttests is weeks or months

Teacher may report posttest score as final grade

5.

Teacher may report instructional gain from pre- to posttesting

PROBES Infrequent measurement of responses that have no current instruction or consequation

1.

Use to assess whether

the

teaching

of spe-

cific items has a generalized effect on similar untaught items 2. Requires minimum of teacher time rough 3. Provides a estimate of student skills 4. Can be used to measure. baseline _ skill prior to instruction

Probe items must be related in form or function to items taught

28

Chapter 1

. Responses

frequency

that generate permanent

prod-

ucts can be translated into numerical terms

rate percent

of

;

OF

refers to the number of times a specific behavior occurred in some period of time.

Frequency

. Frequency of occurrence should be used as a measure of behavior only when and

opportunity for response; time

are

constant.

. If frequency measures are utilized without

subjective interpretation

reference to opportunity for response and time variables, student performance is left to

is defined as the freof occurrence during a unit of

Rate

quency time. frequency time

. Rate

is calculated

frequency

. Rate =

by dividing the

by

time minute

10. Rate is usually expressed in responses per

rate of occurrence

te The basic datum of a science of education is

Percentage

12

20

is a ratio that reports data as a given amount in every hundred responses. . It is unwise to compute percentages when the divisor is less than

Trials to criterion

14.

operations

LS: A method for measuring the complexity of academic responses would be to count the necessary for a cor-

Probes

16.

measures the number of times response opportunities are presented before the student achieves a preestablished level of competence or proficiency.

rect response.

instruction

tell the teacher whether on selected items has affected

a student’s accuracy on similar items that were not taught.

Direct Measurement of Permanent Products

29

Exercise 1-1 1.

In figure 1-1, what is the frequency of reversed two-digit number sums

2.

In figure 1-1, what is the frequency of nonreversed two-digit number sums of 12 or above? In figure 1-1, what is the rate per minute of all addition and sub-

of 12 or above (e.g., student wrote 21 instead of 12)?

3.

traction

4. 5.

errors?

(Note,

do not count unanswered

problems

as errors.)

In figure 1-1, what is the rate per minute of all correct addition and subtraction answers? In figure 1-1, what percentage of two-digit number sums of 12 or above are reversed?

Answers for Exercise 1-1

Re NO

17 reversals 4 nonreversed

3.

17 addition errors

sums

+ 1 subtraction error

20 minutes addition problems

4.

15 correct

+

Sy

20 minutes = 1.6 correct answers per minute 17 reversed sums 21 two-digit sums of 12 or above

= .9 errors per minute

17 correct subtractions

x 100 = 18%

References Clark, H. B., Boyd, S. B., & Macrae,

J. A. A classroom

program

teaching

disadvantaged youths to write biographic information. Journal of Applied Behavior Analysis, 1975, 8, 67-75. Cooper, J. O., & Johnson, J. Direct and continuous measurement of academic

behavior. Directive Teacher, 1979, 1, 10-11; 21. Guilford, J. P. Fundamental statistics in psychology York: McGraw-Hill, 1965. Helwig, J. Effects of manipulating

and education.

New

an antecedent event on mathematics response rate. Unpublished paper, Ohio State University, 1973. Hopkins, B. L., Schutte, R. C., & Garton, K. L. Effects of access to a playroom on the rate and quality of printing and writing of first and second-grade students. Journal of Applied Behavior Analysis, 1971, AND). Ts Gil, CB Johnson, M., & Bailey, J. S. Cross-age tutoring: Fifth graders as arithmetic tutors for kindergarten children. Journal of Applied Behavior Analysis, 1974, 7, 223-32.

30

Chapter 1

Kunzelman,

H. P. Precision

teaching.

Seattle:

Special Child

Publications,

1970. McKenzie, H. S., Egner, A. N., Knight, M. R., Perelman, P. F., Schneider, B. M., & Garvin, J. S., Training consulting teachers to assist elementary teachers in the management and education of handicapped children. Exceptional Children 1970, 37, 137-43. Skinner, B. F. Operant behavior. In W. K. Honig (Ed.), Operant behavior: Areas of research and application. New York: Appleton-Century-Crofts,

1966.

Observational Recording

Measurement of permanent products is the foundation for most classroom evaluation. Indeed this measurement tactic should always be a major source of data for teachers. Yet, as instructional technology develops through behavior analysis, many teachers now see that observational recording is an equally important measurement tactic. Classroom teachers are concerned with many socially significant behaviors that do not result in permanent products. These behaviors must be observed and recorded as they occur. When teachers look at behavior and produce records of that behavior as it occurs, they are engaging in observational recording.

By way of illustration, let us look at a problem behavior in Mr. Clamer’s fifth-grade class. Mary Jane’s behavior has been a source of disruption in Mr. Clamer’s class for the past three weeks. Whenever she is called on to participate in class discussion—asked a question or given a directive—she defiantly responds by shaking her head no, much to the amusement

of her classmates.

Obviously,

when Mary Jane shakes

her head no, she does not leave a permanent Sy

product that can be

Observational Recording

3368)

measured one week after it occurred. Observational recording must be used to measure this behavior. : At first, Mr. Clamer ignored Mary Jane’s gestures of no, hoping that they would cease. But eventually he decided that these gestures were increasing and he planned an intervention tactic. He made a daily tally of the number of occurrences of no gestures for 5 consecutive school days before introducing intervention. With these data, Mr. Clamer could determine whether Mary Jane made a no gesture with her head less often during intervention than she had before intervention. Other examples of socially significant school behaviors that require observational recording include verbal responses to teacher questions, cooperation, increases in sentence speaking, teasing, guffaw laughing, aggressive hitting, and noise levels in the lunchroom. Chapter 2 addresses the topic of observational recording. Hall (1971) outlined five major observational recording techniques for teachers: (1) continuous recording, (2) event recording, (3) duration recording,

(4) interval recording, and (5) momentary time sampling. Each of these observational recording techniques will be presented. After studying chapter 2 you should be able to: 1. Give applied examples of each observational recording technique 2. Discuss how appropriate each technique is for the classroom teacher 3. Select appropriate tactics for observational recording

Interobserver Agreement Mr. Jones taught tenth-grade Spanish. He and Ms. Roberts, the school psychologist, were concerned with Larry’s boredom in school and his lack of enthusiasm for the subject matter of his courses. Their goal was to help Larry develop an appreciation for the subject matter presented in his classes and more positive feelings toward school. A teaching intervention was started. The intervention tactic consisted of a maximum of teacher attention contingent upon improvements in Larry’s behavior. Mr. Jones had heard of the effects of social reinforcement and was convinced that this technique would help Larry. Ms. Roberts felt that social reinforcement would not hurt anything but had reservations concerning its effectiveness in this case. During the next two weeks, Ms. Roberts attended Larry’s Spanish class and made notes on all

behavioral

improvements.

Also,

whenever

Mr.

Jones

observed

Larry showing appreciation for subject matter or more positive feelings

34

Chapter 2

toward school, he tallied the number of occurrences in a daily progress report. : At the end of two weeks the teacher and school psychologist met to discuss Larry’s progress. Mr. Jones reported marked improvement. Ms. Roberts,

on the other hand, stated that her observations

showed

ab-

solutely no change in appreciation for subject matter or more positive feelings toward school. When Mr. Jones and Ms. Roberts realized they could not agree upon Larry’s behavior, they asked the following question: What exactly was Larry doing that caused them to say that he needed to develop an appreciation for subject matter and positive feelings toward school? The most obvious problem was the fact that Larry was not completing his assignments. Mr. Jones and Ms. Roberts decided, therefore, to count the number of assignments Larry completed in one week. When they conferred the second time, Mr. Jones reported that Larry completed 12 of 13 assignments and Ms. Roberts reported 13 completed assignments. By counting the number of completed assignments, the teacher and psychologist were able to communicate what they had observed. As this illustration demonstrates, behaviors must be defined specifically. They must also be defined objectively; that is, people must agree upon the occurrence or nonoccurrence of behavior. Directive teachers can only work with observable and measurable behaviors. In the case of Larry, Mr. Jones and Ms. Roberts did not at first define an observable

and

measurable

behavior.

However,

the second

behavior

definition (number of completed assignments) proved to be a scientific definition because they were able to report close agreement concerning what was observed. In directive teaching it is necessary to know that the teacher’s observations and recordings are reliably reported. A common way to determine reliability of classroom measurement is through simultaneous measurement by independent observers. Hall (1971) defined interobserver agreement as “the degree to which independent observers agree on what they have observed in the same subject during the same observation session” (p. 6).

Interobserver agreement measures are usually reported as percentage of agreement among two or more independent observers. Percentage of agreement is calculated by dividing the number of agreements by the number of agreements plus disagreements and multiplying by 100.

agreements X 100 = percentage of agreement agreements + disagreements To illustrate, Mrs. Gomez was interested in decreasing the number of times that Jack talked out in class. She invited Mrs. Anderson, the

Observational Recording

355)

school principal, to function as an independent observer of Jack’s talkouts. Before class, Mrs. Gomez gave the principal a definition of talk-outs and told her to make a tally each time she heard Jack talk out between 9:00 and 9:15 A.M. The definition given to the principal was: “Talking out is defined as a vocalization, comment, or vocal noise initiated by the student. It cannot be in response to the teacher or another peer. Each occurrence will be tallied as one talk-out if it is preceded or followed by a breath, time interval or change of topic.” At the end of the 15-minute observation period, Mrs. Anderson had tallied 10 talk-outs. The teacher recorded 12 talk-outs. Mrs. Gomez calculated the reliability of her data by dividing the number of agreements (10) by the number of agreements plus disagreements (10 + 2) and multiplying by 100. Mrs. Gomez’s calculations were:

10 —~ = 0.833 X 100 = 83% 12 Mrs. Gomez’s agreement measure was sufficiently high that she can accept her observation as reliable and proceed with teaching Jack appropriate classroom verbal behavior. As a rule of thumb, an agreement measure should be on the average above 80% before the teacher continues with instruction. Perhaps the most frequent cause of unsatisfactory agreement measures among observers is a lack of precision in defining behavior. Behaviors must be defined in terms of body movements that result in outcomes that can be seen or heard or felt; that is,

defined behaviors should be observable and measurable. If agreement measures are below 80%, the observers should discuss why agreement was not reached concerning the occurrence or nonoccurrence of behavior and arrive at a new collective behavior definition. This process should continue until an acceptable agreement percentage is reached. However, after acceptable agreement measures have been reached with the original observers, an outside person should be asked to be an observer. The: original observers may only have reached agreement between themselves! Asking a third independent person to serve as observer is the best check a teacher can use to tell if the behaviors selected for observation are generally recognizable. An outside observer should be able to tell when Jack is or is not engaging in the behavior that the teacher wants to change. Some key points to remember concerning reliability are: 1. Other students, other teachers, teacher aides, school secretaries,

school psychologists or counselors, consultant teachers, parents, volunteers, and principals can serve as observers for the classroom teacher.

36

Chapter 2

2. Interobserver agreement measures should be made over several sessions prior to and during each phase of instruction. 3. Observers should not be informed of the teacher’s instructional strategies. “An observer may tend to err without being aware of doing so, by recording results in the anticipated direction. Thus, if one was aware that a reinforcer was being applied, he might be more likely to record an increase in behavior. If he knew that the reinforcer had been withdrawn, his tendency might be to record a decrease in the behavior” (Sulzer and Mayer, 1972, p. 269). Exercise 2-1

During your next lecture period, or assembly meeting with a speaker, arrange with another person to tally the number of occurrences of some behavior emitted by the lecturer or speaker. You could count one of many behaviors, such as hand gestures, jokes, steps taken, pushing eye glasses up, rubbing head, number of “uhs,” and so on. Agree upon the behavior to be tallied before the lecture or assembly meeting. Tally the number of times the behavior occurs during the first 15 minutes of the meeting. Calculate your percentage of agreement. Were your data reliable? If not, what should you do?

Continuous Recording Continuous recording often takes the form of “anecdotal reports” or “diary records.” The aim of this recording technique is to record all behavior as if occurs. No specific behaviors are pinpointed for observation to the exclusion of other behaviors. This technique produces a written narrative of individual or group behaviors for a specified time period. The conditions or situations under which the behaviors were emitted are also described. The main idea of continuous recording is to produce as complete a description as possible of student behaviors in specified settings. Wright (1960) reported directives for continuous recording: Begin in reporting each observed sequence with a description of the scene, the actors, and the ongoing action. Report throughout in everyday language. Describe the situation as fully as the behaviors. Thus include “everything the child says and does,” but include also “everything said and done to him.” ... Describe the larger “adaptive actions” of the child, but weave in as well the “hows” of these actions as far as possible. “Nonadaptive aspects of behavior” are important on this account. . . . Do not substitute interpretations that generalize about behavior for descriptions of behaviors, but add such interpretations when they point to possi-

Observational Recording

Sy

bilities of fact that might otherwise be missed. Segregate every interpretation by some such device as indentation or bracketing. Straight reporting must be left to stand out. (pp. 84-85)

Wright also described several possible procedures for continuous recording: Notes on the scene of observation, which obviously are needed for sufficiently detailed and accurate description, can be kept in improvised shorthand. These field notes can be enlarged in writing immediately after each observation period, or they can serve then as a base for a dictated narration of the observed behavior sequence. Also, a co-worker can hear this account through and at once question the observer where it is thin or unclear. The original dictation plus the interrogation and the observer’s responses can be sound recorded. All of the recorded material can then be copied and revised in an improved account. . Observations can be timed to permit various measures of duration. Timing of the field notes at intervals of approximately one minute or even 30 seconds has been found practicable. When long records are made, observers can work in rotation; and the time of

each observing period can be regulated to minimize effects of fatigue upon ability to see and remember an always fast train of events.

(p. 85)

Regardless of the recording instructions or procedures employed, a number of common elements are found in all continuous recording techniques. 1. A time sequence is given. This sequence may be reported as large units of time (e.g., 9:00-9:50 A.M.) or as smaller multiple time samples, so on).

(e.g., 9:00-9:05,

9:05-9:10,

9:10-9:15,

and

2. Target behaviors are unselected; no specific behavior is pinpointed. 3. Data take the form of narrative descriptions of behaviors. These narrative descriptions should include a three-term contingency of (a) events that occur before a behavior is emitted, i.e., antecedent stimuli, (b) the behavior of the child, and (c) the

events or stimuli that occur after the specified behavior. 4. The narrative is plainly worded and easy to read. Applied Example Bijou, Peterson, and Ault (1968) reported an example of continuous recording with a preschool child named Timmy. The setting was a play yard.

38

Chapter 2

Timmy is playing by himself in a sandbox in a play yard in which other children are playing. A teacher stands nearby. Timmy tires of the sandbox and walks over to climb the monkey-bars. Timmy shouts at the teacher, saying, “Mrs. Simpson, watch me.” Timmy climbs to the top of the apparatus and shouts again to the teacher, “Took how high I am. I’m higher than anybody.” The teacher comments on Timmy’s climbing ability with approval. Timmy then climbs down and runs over to a tree, again demanding that the teacher watch him. The teacher, however, ignores Timmy and

walks back into the classroom. Disappointed, Timmy walks toward the sandbox instead of climbing the tree. A little girl cries out in pain as she stumbles and scrapes her knee. Timmy ignores her and continues to walk to the sandbox.

This example of continuous recording is written in a style used by reporters for a newspaper or magazine. Directive teachers, however, need to form a clear impression of temporal relationships among antecedent stimuli, responses, and consequent stimuli. The four-column form shown in figure 2—1 is helpful in delineating time relationships.

DESCRIPTIVE Setting:

OBSERVATIONAL INFORMATION Observation date:

FORM

Student: Antecedent Events

FIGURE

Child Responses |Consequent Events

2-1

The episode of Timmy’s play yard behavior was transcribed by Bijou et al. into a four-column form, and each behavioral and stimulus event was consecutively numbered. The continuous record for his behavior is shown below. Setting. Timmy (T.) is playing alone in a sandbox in a play yard in which there are other children playing. T. is scooping sand into a bucket with shovel, then dumping the sand onto a pile. A teacher,

Observational Recording

39

Mrs. Simpson (S.) stands approximately six feet away but does not attend to T. : Time

Antecedent Event

9:14

Response

Consequent Social Event

1. T. throws bucket and shovel into corner of sandbox. De Stands up: 3. ... walks over to monkey-bars and stops.

4. ... turns toward teacher. Se SAVS eins: Simpson, watch Imes 6. Mrs. S. turns toward T.

6. Mrs. S. turns toward T. 7. T. climbs to top of apparatus. 8. ... looks toward teacher. 9. ... says, “Look

how high I am. I’m higher than anybody.” 9:16

10. Mrs. S. says, “That’s good, Tim. Youre getting quite good at that.” 10. Mrs. S. says, “That’s good, Tim. You're getting quite good at that.” 11. T. climbs down. eee

LUDS OVE tO

tree. 13. ... says, “Watch me climb the tree,

Mrs. Simpson.” 14. Mrs. S. turns and walks toward classroom. 14. Mrs. S. turns and walks toward classroom.

9:18

15. T. stands, looking toward Mrs. S.

16. Girl nearby trips and falls, bumping knee. 17. Girl cries. 18. T. proceeds to sandbox. 19. ... picks up bucket and shovel. 20. ... resumes play with sand.

Note that a response event (e.g., 5, ... says, “Mrs. Simpson, watch me.”) may be followed by a consequent social event (e.g.,

40

Chapter 2

6. Mrs. S. turns toward Timmy.) which may also be the antecedent event for the next response (e.g., 7. T. climbs to top of apparatus.) Note too, that the .. . form retains the temporal relationships in the narration. Note, finally, that only the child’s responses are described. Inferences about feeling, motives, and other presumed internal states are omitted. Even words like “ignores” and “disappointed” do not appear in the table. (pp. 78-79)

Considerations

Directive teachers should employ continuous recordings primarily (1) as an assessment aid in pinpointing specific behaviors that need to be learned,

(2) to identify possible environmental

conditions that set the

occasion for student responses, and (3) to identify possible consequent events that maintain behaviors. Perhaps the major reason for utilizing this technique is that a large range of behaviors can be observed and recorded. Continuous recording is a measurement procedure seldom employed by classroom teachers. The teacher cannot concurrently give instruction and measure behavior with continuous recording techniques. The recording technique is more appropriate for observers other than the directive teacher, e.g., teacher aides, volunteers, school psychologists, consultant teachers, principals, secretaries, students.

Some

have

reported that continuous

observer agreement measures

records

generate

good inter-

(see Wright, 1960, p. 86, for references).

Yet one must question high interobserver agreement (80% or better) with continuous recording. It is impossible to record everything as it occurs in time, and it is highly unlikely that two observers would report exactly the same things in the same way. For example, tables 2-1 and 2-2 present simultaneous observations compiled by two doctoral students trained in observational recording. These two records do not show high interobserver agreement concerning specific behaviors. Yet, even though they differ in the specific temporal sequences, events, and responses recorded, the two reports provide similar overviews.

Evaluation At this point, you may wish to evaluate how well you comprehend the information on observational recording by responding to the following

41 Observational Recording

UH

,ouo SIq} JaMsue

19] [L.9M,, SAPS T9YIVIL

gsuodsol siq

,ouy,,

‘S}oofJol IoyoVL,

ynoqe uoysenb syse JoyoVolL,

siamsue

Yoo] & WIY SoATs IBYIVOL ..ouy,,

suyjods sesueyD

4991109 0} S[[9} IeYyoRIL

JUonbasuoD

g[ilus yyIM puodsal syuepnig s]UsAq

MoIp oy yey

punoie 9jo.119 ][9} 0} PUY SosteI SLY)

gpiio smeip APoInO

gummy Aaa spuodses sty

ginjoid ynoge No SyIe} SIIYD dn way} UMOP SYZOO] ‘svep! SIOYO SYD os Sulop hq spuodses stiq)

“IoMsue 0} me Ul puey YM—yorq 1Y4s11— puoses oUO MOPUIM 9Y} INO SOOT SY

ou

1, U1e

yey,

_

s}oafqo Surmmeu Aq spuodse1 sly 21g

Yoo $,Ua[H 0} syuIod—seyo}VM—SHY Ayoei109 spuodseal stiqD

sosed suin} siiyD sosuodsoy pilyO

yueq Assid ul sdury} punole

J9Yysve TL,

syooy stiyD

senurjuo0d

punoie

QOD MBIP 0} PO} SYOOG savy s}UIPNIS

suorjsonb

UOT}VSIOAUOD

esed suin} sly)

SIoMsUe S,P]IYyd JayJoUe 0} sUd}SIT—b syse IoysVeL,

suonsonb syse 19yoee], yoo siy 3@ SUTYOO] 0} Yoeq sod sty

ug[H ‘Aqny Jo suonsenb poyse Jayoeval,

[eds noA Op MoY SLIYyD sysyv SJUIWUSISSe [LNPIAIPUL SoATs JOyIVIL,

sainjolgd—jyeyo Woz

CO: TI-Ss:O1

own,

“[-¢ ATaVL

quapnis YWA ‘SYD

Zurpeoi dnois [jews—,soesed uinL,,

s]JUsAq WeposdjUy

soy uoneuoyuy feuoHeatosqQ VW JearesqQ Aq pepsoooy wonoeso}Uy jidng-royoeo], Suipre

©

“SUIL

Soop sty}

poysy e@

peor

uonsonb

..~yUeq 01) 11% (€

Joyjoue Aoq

JoyJoue yuspnjs

[194s «d

uolsonb

Og,, NOK oALY ev

paysy

rayovay, sey

I9yOVIT sysoyo

AA,,yey

s$joT,, gas

sasueya oy}

..ON,, UdY}) sppe

..Y9X,, ysIy) 0}

sjuapnjs

puey

(Apyornb

$}991109—A0q Jayjo ON

(s}UoUIWOD

‘puodsal ouop

soulsYIM J9y}O Zurpeos Jo yeyo peorAq

spjoyH dn

ohq

T,,

pasueyo oy}

«Suljeds

S8.J9[—poos asueyo Sul[[eds «SPIOM

JUonbesuo_d s}UsAq

JoaresqQ q

.J21Nd,,pres)

19y9e9]

JayovoL, sey JaMsue

esuodsol

Jayjoue yuspnjs

(APJos

JURM O} SAIS NOA Ue § -9q asned NOA PsyIOM «PIVY

noX,,

sjey,, ssoyy

sulamsue—yeq [[e suorjsonb ioa Jayoes}poyse Aj[euolses90—dnois poyoo] JNO MOPUIM

AI9A, yey)

suniem Joj

sosuodsoy

sullamsue suonjsonb poyse jo

Yoo

JAYS

yieyo

suly[eds

Ioyjoue quopnjs

sdooy

sjutod 0}

$7eqLL,, JOU

siya

siiyD syooy 1

i4O

SUIpeoI ‘a[dI19

smoys Jaded

Jayoee}

SUTIIIS UI

priya

pidng-soyoeayuonsesojuy papiosoy Aq

Chapter 2

Zp

JWopaoojuy s}UdAq

[BUONeAIOSGQ.UOHeUWIOJUTSulpresoy

IaYyDVTSOWOD 0} —;:/,-----"

12345678907!(%)&;:/,”----

Folio Medium Extended Italic

Franklin Gothic

Futura Bold

abcdefghijkimnopqrst uvwxyzeececoaB

ABCDEFGHIJKLMNOPG RSTUVWXYZAGECS abcdefghijkimnopaqrstu vwxyzzeccoB \;:/-.-~-

12345678907!0 2; :4"">~--~e

123456789071) -

ABCDEFGHIJKLMNOP QRSTUVWXYZA-ECS

ABCDEFGHIJKLMNOPQRS TUVWXYZAGCSO abcdefg hijkimnopqrstuvwxyzeeec

0B 12345678902'(&);: /------

Letraset 33 New Bridge Road, Bergenfield, New Jersey 07621

Grotesque 9 Italic

Folio Medium Ext.

ABCDEFGHISK LMNOPQRSTUVWXYZ abcdefghijk!

ABCDEFGHIJKLM NOPORSTUVWXYZ abcdefghijkl

mnopgrstuvwxyz

mnoparstuvwxyZ

1254567890 GPIBESO RS:

1234567890 &IBESOBE

Standard Medium ABCDEFGHIJKL MNOPORSTUVWXYZ

ENGINEERING STANDARD

abcdefghijk| mnoparstuvwxyz 1234567890

ABCDEFGHIJKLM NOPQRSTUVWXYZ 1234567890

2!&G) FIGURE B-6.

&71$(); Available Typefaces by Brand Name 200

Preparing for Printing

numbers many

201

are separate. Check before buying. If your graphs call for

numbers,

or several of one number

(such as 5s and Os), it is

usually better to buy an entire page of numbers. Lines. The axis lines should be the heaviest (about 3 points) and the lines within the graph narrower (one point). Broken and dotted lines to show more than one measure on a single graph are also available. Lines are available from some companies in a roll. Symbols.

Asterisks, stars, squares, etc., to represent data points may

be purchased in varying sizes, usually with several sizes of several symbols on one page. Simply choose a size that conforms with the lettering. To use dry-press materials, place the page facedown on the paper, centering the letter. Then with a pencil or pen scribble over the back of the entire letter. The letter will turn gray when it is released from the plastic page. After all letters are correct and in place, rub them again with the backing sheet for firm adhesion. What to Do for a Mistake. Erase the mistake with clean pencil eraser or scrape it off with a sharp knife. Making the Photograph. Printers usually require glossy prints of graphs to be published. The reduced photographs can be made by most newspaper or job printing establishments. Prices for this service vary; it might be wise to call for estimates. That makes everything ready for publication—Congratulations!

Glossary

Key words and phrases used in this book are defined below. The definition given is that used or implied in the book. For technical words and phrases used in this book but not included in the glossary or for more complete definitions see Owen R. White, A Glossary (Champaign, Illinois, Research Press, 1971).

AB

of Behavioral

Terminology,

Design. A procedure for demonstrating amount of behavior change where behavior is measured (1) before instruction (see baseline) and (2) during instruction. This design cannot show what variables caused the behavior change. Analysis. Identification of the effects of a teaching procedure on student behavior. Analysis is concerned with the amount of behavior in both the presence and absence of instruction. Instructional events are systematically manipulated to demonstrate that the teacher can exercise control over student behavior. The two designs most widely used in behavioral analysis are the reversal and multiple-baseline designs. Assessment. A survey of student behavior to detect those responses and skills that are adequate and those yet to be learned or mastered.

202

Glossary

203

Baseline. The level or amount of behavior occurring prior to the introduction of a teaching procedure. Baseline data are usually expressed in numerical terms as rate, frequency, percent, or duration. Teaching procedures are frequently introduced only after five or more sessions of baseline data have been collected. Behavior. Refers to any act of an individual or group that is observable and

measurable. Behavior can be external or internal. The term is used interchangeably with response. Central Tendency. A statistic intended to typify measurement of group or individual behavior. Central tendency is an approximation of a typical student’s behavior. The mean and median are frequently used as an indices of central tendency. Changing Criterion Design. A design used with objectives that specify a gradual stepwise increase or decrease in academic or social behaviors. The teacher changes the response criterion in a stepwise fashion from the baseline level to a desired terminal level. Continuous Recording. Sometimes called an anecdotal record, this form of recording requires the observer to write down everything he observes as it happens. Cumulative Graph. A graphic presentation of successive summed numbers (rate, frequency, percent, duration) which represent behavioral occurrences. Data Point. A quantitative score of an individual or group behavior as reported on a graph. Delayed Multiple-Baseline Design. A variation on the multiple-baseline design. The first baseline and the first intervention are begun prior to gathering data for the second baseline. Additional baselines are added in this staggered or delayed fashion. Baselines subsequent to the first one could be initiated prior to beginning intervention on a previous baseline. Differential Reinforcement of Other Behavior (DRO). A procedure in which

reinforcement is given contingent on the nonoccurrence of the target behavior. Duration Recording. Recording the extent of time that a behavior continues or lasts, usually reported as percentage of time. Event Recording. A tally or frequency count of discrete events as they occur. Frequency. The number of times a specific behavior occurs in some period of time; frequency counts take the form of a tally. Functional Relationship. In applied behavior analysis, a functional relationship is demonstrated when a behavior varies systematically with the application of an instructional procedure. Sometimes called a cause-and-effect relationship. Incompatible Behavior. Refers to a behavior that is the opposite of or that cannot occur at the same time as some other behavior. Interobserver Agreement Measures. Procedure to verify the reliability of the measurement of behavior. Reliability is usually concerned with the percentage of agreement between two or more independent observers on the occurrence

or

nonoccurrence

of behavior.

Percentage

of interobserver

204

Glossary

agreement is calculated by dividing the number of agreements by the number of agreements plus disagreements and multiplying by 100. Interval Recording. Used to measure the occurrence or nonoccurrence of a behavior within specific time intervals. Time intervals usually range from 6 to 30 seconds depending on behavior to be observed. Behavior is usually recorded only once per interval and reported as percentage of occurrence. Intervention. A systematic manipulation of teaching tactics. Magnitude of Behavior Change. The quantity by which a behavior increases or decreases with respect to some prior amount of behavior. Mean. An arithmetic average of a set of data. Median. The middle score in a ranked distribution. Model. A teacher, person, material, or device which demonstrates a skill or task to be imitated by a subject or student. Momentary Time Sampling. The measurement of the occurrence or non-

occurrence of a behavior immediately following time. Multiple Baseline. A multiple-baseline analysis can more similar behaviors are emitted by the same behavior occurs in different stimulus conditions, havior occurs in more than one subject. When contingencies may be applied to one behavior, stimulus

condition,

a specified interval of be used when two or subject, when the same or when the same bethese conditions exist, then the other; in one

then the next; or with one subject, then sequentially

with other subjects. Functional relationships are established if changes in each behavior correspond to experimental manipulations. Multielement-Baseline Design. A design used to analyze the effects of two or more discrete conditions of instruction. The student’s behavior is measured repeatedly under alternating baseline and instructional conditions. Conditions can be arranged to occur in regular alternation or randomly. An identifying stimulus must be associated with each condition in the design. Multiple-Probe Baseline Design. A variation on the multiple-baseline design. The procedure for the multiple-probe technique includes collecting initial probe data for each baseline tier, recording additional probe data on each baseline tier after criterion is reached on the intervention tier, and recording consecutive sessions of baseline data for each tier just before the intervention is started. Noncontingent Reinforcement. Consequences of behavior usually presented to the student on variable time or fixed time intervals regardless of student behaviors. Noncumulative Graph. Sometimes called a line graph. Student scores are placed on squared graph paper. Points representing amount of behavior (y-axis) are plotted opposite the points along the x-axis representing the times

at or during which

the behavior

occurred

(x-axis).

After

scores

(data points) have been placed, they are connected by a line. Observational Recording. Recording that results when teachers or other individuals look at behavior and record that behavior as it occurs.

Glossary

205

Percentage. A ratio that expresses the amount of behavior as a certain number of responses in every hundred responses. Permanent Product. A product that is tangible and can be measured any time after the students’ behavior has occurred. Written spelling words, written arithmetic computation, and strung beads are examples of permanent products. Pinpointed Behavior. A specific, measurable behavior that is chosen for instruction. Postcheck. Periodic measurement of a behavior after the teacher has formally terminated training for that particular behavior. Pre- and Posttests. The pretest is used primarily to estimate what the student can do before instruction. The posttest is given following instruction. Probe. Infrequent measurement of responses that have no current instruction or consequation. Range Score. The difference between the lowest and highest score in a set of measurements. Rate. The frequency of occurrence of a behavior during a unit of time. Rate is calculated by dividing the frequency of occurrence by the amount of _ frequency time spent working on the material (rate ) and is usually expressed in responses per minute. Reinforcement. Any consequence of a behavior that increases the probability of occurrence of that behavior. Reliability. See Interobserver Agreement Measures. Reversal (ABAB) Design. A design for behavior analysis with four conditions: baseline,, intervention,, baseline, intervention. This design can be used to show a functional relationship between teaching and student improvement.

Target Behavior. See Pinpointed Behavior. Terminal Behavior. The final criterion for student behavior. Trials to Criterion. Measurement concerned with the number of times response opportunities are presented before the student achieves a preestablished level of competence or proficiency. X-axis. With standard graphic arrangements, the x-axis is the horizontal axis and usually represents units of time (e.g., minutes, sessions, days, weeks). Y-axis. With standard graphic arrangements, the y-axis is the vertical axis and usually represents amount of behavior (e.g., frequency, rate, percent,

proportion, duration).

&

a

Hf

y

Index

AB design, 99-103 ABAB design, see Reversal design Alternating conditions (multielement

baseline)

definition of, 99

Behavioral-analysis techniques AB design, 99-103 changing criterion design, 15763

design,

165-72

classroom use, 174

multielement baseline design, 165-72 See also Multiple-baseline designs; Reversal design Bijou, S. W., study by, 37-40

Ault, M. H., study by, 37-40 Averages mean, 85

median, 85-87 Baer, D. M., 142, 153 study by, 108-10 Bailey, J. S., 24

Boyd, S. B., 25 Bushell, D., Jr., study by, 61-62

Barlow, D. H., 165 note, 166 Baseline logic applications of, see Behavioralanalysis techniques

Camera copy, preparing, 197201 Central tendency, measures of, 85-87 207

208

Index

Changing criterion design, 15763 Christensen, A., 46

Clark, hi Be. 2) Complexity of items, measuring, 22-23, 27 Condition-change _ interactions, 165 Contingent reinforcement of incompatible behavior, 11014 Continuous recording, 36-43 Cooper, J..O.; 21

studies by, 111-14, 123, 12527, 129-32 Correction fluid, 198-99

Counters hand-tally, 187 Criterion levels changing, 157-63 trials to, 21-22, 27 Cumulative graphs, 74-78 Delayed multiple-baseline design, 145-53 Different-behaviors multiplebaseline design, 12-26, 127 Different-settings multipledesign,

32 Different-students

other

128-

multiple-

baseline design,

Differential

126,

132-40

reinforcement behaviors

Garton, K. L., study by, 12-13,

14, 18, 19 Garvin, J. S., study by, 10, 11 Graphic presentations, 72-93 of central tendency and range, 85-88 cumulative,

74-78

guidelines for, 89-92 multiple concurrent, 79-84 noncumulative,

73-74

preparation for printer, 196-201 Guilford, J. P., 18

Hall, (R.Y 7) 33, 34, 100157. 160 study by, 57, 58 Hand-tally counters, 187 Handicapped students, 64

wrist, 186-87

baseline

Foxx, Ro M., 192

Frequency of response, measuring, 10-12, 26

of

(DRO),

114-16 Dry-press transfer, 199-201 Duration recording, 48-52 instruments for, 48, 189-90 Eaton, M. D., 65-68

Egner, A. N., study by, 10, 11 Event recording, 45-46 instruments for, 45, 186-89

Harris, J., study by, 137, 139-40 Hartmann, D. P., 157, 160 Hersen, M., 165 note, 166 Heward, W. L., 145

Hopkins, B. L., study by, 12-13, 14, 18, 19 Horner, Ry D142 5153 Incompatible behavior, contingent reinforcement of, 11014 Induction effect, 140

Instrumentation for observational recording, 45, 48, 56, 61, 186-93 Interaction effects, 165 Interobserver agreement, 33-36 with continuous recording, 40 Interval recording, 53-58 instruments for, 56, 190-91, 192

Index

Jackson, Jacobsen, 27, Johnson, Johnson,

D., study by, 57, 58 B., study by, 123, 125129-32 J., 21 M., 24

Jorgenson, H. A., study by, 56eo ets Knight, M. R., study by, 10, 11 Kunzelmann, H. P., 2, 15

La Barbera, J., 46 Lindsley, O. R., 186 Line graphs, 73-74 Lund, D., study by, 57, 58 McKenzie, H. S., study by, 10, 11 Macrae, J. A., 25

Martin, P. L., 192

different settings, 126, 128-32 different students, 132-40 multiple-probe, 142-45, 150, i353 Multiple concurrent graphing, 79-84 Multiple-probe baseline design, 142-45, 150, 153 Multiple schedule (multielement baseline) design, 165-72 Noncontingent

reinforcement,

107-10 Noncumulative graphs, 73-74

Observational recording, 32-63 continuous recording, 36-43 duration recording, 48-52 instruments for, 48, 189-90

event recording, 45-46 instruments for, 45, 186-89

Mayer, G..R., 36, 171 Mean, arithmetic, 85 Measurement tactics

instruments

aids to proficiency with, 65 limitations, 98 occasions for, 63-64

problems, 64 use of data, 65-68 See also Observational record-

ings; Permanent-product measurement Median, 85-87 Michaelis, M. L., study by, 61-

62 Momentary time sampling, 6062 instruments for, 61, 191-93 Multielement baseline design, 165-72 Multiple-baseline designs, 122 advantages and disadvantages,

140-42 delayed, 145-53 different behaviors, 127

209

122-26,

for, 45, 48,

61, 186-93 interobserver agreement, 36, 40 interval recording, 53-58

56,

33-

instruments for, 56, 191-92,

192 momentary

time

sampling,

60-2 instruments for,61,191-93 Payne, J. S., study by, 123, 125-

Deal 29—32 Pen and ink drawing, 197-99

Percentage of response, measuringali=20,-20 Perelman, P. F., study by, 10, 11 Permanent-product measurement, 8-29 complexity of items, 22-23, 27 frequency of response, 10—i2, 26 percentage of response, 17-21, 26

210

Index

Permanent-product (continued) pre- and posttesting, 23-24, ZT probes, 24-25, 27 rate of response, 12-17, 26 trials to criterion, 21-22, 27 Peterson, R. F., study by, 37-40

noncontingent reinforcement, 107-10 Reynolds, N. J., study by, 114LoL ty Risley, T. R., 1, 98, 99 study by, 114-16, 117

Postcheck data, graphing, 92-93 Pre- and posttesting, 23-24, 27, 82-84

Schneider,

Probes, use of, 24-25, 27

14, 18, 19 Signal-generating devices, 56, 61, 190-93 Stanberry, M., study by, 137, 139-40 Stephens, T. M., 99 Stopwatches, 189-90 Sulzer-Azaroff, B., 36, 166, 171

multiple-probe baseline design, 142-45, 150, 153 Range scores, 87-88 Rate of response, measuring, 12-

175-26 Reinforcement contingent, of incompatible behavior, 110-14 differential, of other behaviors (DRO), 114-16 noncontingent, 107-10 Reversal (ABAB) design, 104— a, advantages and disadvantages, 116-19

in changing criterion design, 163-64 contingent reinforcement of incompatible behavior, 11014 differential reinforcement of other behaviors (DRO), 114-16

B. M., study by, 10,

ti Schutte, R. C., study by, 12-13,

Tallying devices, 45, 186-89 Time sampling, momentary, 60—62 instruments for, 61, 191-93 Timing methods, 48, 56, 57, 61, 189-93 Trials to criterion, 21-22, 27

Wolf, M. M., 99 study by, 108-10 Worthy, R. C., device described by, 56, 190-91, 192 Wright, H. F., quoted, 36-37 Wrist counters, 186-87 Wrist tally board, 187-88 Wrobel, A., study by, 61-62

Ny

—

i

—

¥

>

-

= )

a“

a

a

=

v

=

™

wy

Ss

H

”

hae 2

i

ree

| —

Wat re a 1

P

oy

| { SW oF Uae7

yah 7)?

The Charles E. Merrill Series on Behavioral Techniques in the Classroom Thomas M. Stephens, Consulting Editor

John O. Cooper

Measuring Behavior (Second Edition)

John O. Cooper &

Parenting: Strategies and

. Denzil Edge

Educational Methods

John B. Glavin

Behavioral Strategies for Classroom Management

Stainback, Payne, Stainback, & Payne

Establishing a Token Economy in the Classroom

Thomas M. Stephens

Directive Teaching (Second Edition)

Thomas M. Stephens

Implementing Behavioral Approaches in Elementary and Secondary Schools

Thomas M. Stephens

Teaching Children Basic Skills

Thomas M. Stephens

Teaching Skills to Children With Learning and Behavior Disorders

Charles E. Merrill Publishi | A Bell & Howell Company Columbus, Ohio 43216

fA) Bette Howe