Group Title: free operant analysis of programed insturction performance with reading disabled children
Title: A Free operant analysis of programed insturction performance with reading disabled children
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00098407/00001
 Material Information
Title: A Free operant analysis of programed insturction performance with reading disabled children
Physical Description: xi, 152 leaves. : illus. ; 28 cm.
Language: English
Creator: Edinger, Dennis Lloyd, 1943- ( Dissertant )
Walberg, William D. ( Thesis advisor )
Pennypacker, H. ( Reviewer )
Cunningham, Myron A. ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 1969
Copyright Date: 1969
 Subjects
Subject: Vision   ( lcsh )
Reading disability   ( lcsh )
Special Education thesis Ph. D
Dissertations, Academic -- Special Education -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis--University of Florida, 1969.
Bibliography: Bibliography: leaves 148-152.
General Note: Manuscript copy.
General Note: Vita.
 Record Information
Bibliographic ID: UF00098407
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000871643
notis - AEG8866
oclc - 014279529

Downloads

This item has the following downloads:

PDF ( 7 MBs ) ( PDF )


Full Text










A FREE OPERANT ANALYSIS OF

PROGRAMED INSTRUCTION PERFORMANCE

WITH READING DISABLED CHILDREN












By
DENNIS LLOYD EDINGER


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY













UNIVERSITY OF FLORIDA
1969






























Copyright, 1969
by
Dennis L. Edinger































This dissertation is dedicated to

Ogden R. Lindsley













ACKNOWLEDGEMENTS


This dissertation is the first installment on

my debt toProfessor Ogden R. Lindsley.

Carl Koenig and John Nichol of the Behavior

Bank (P.O. Box 3351, Kansas City, Kansas) deserve enthusiastic

applause from the Florida group for their brilliant efforts

in educational science. The data in this dissertation could

not have been presented without their invaluable assistance.

My gratitude, thanks, and profound regard to

my chairman,Dc. W. D. Wolking, and to my minor director,

Dr. H. S. Pennypacker, for the superb direction and

leadershipof my infantile gropings for a precise science of

human behavior. If I am indeed a scientist, I am of their

seed,

Dr. Myron A. Cunningham is directly responsible

for the maintenance of my doctoral program. Without his

navigation my ship of science would have foundered on

the shoals of bureaucratic gerrymandering.

The teachers who taught deserve a special note

of gratitude. They are: Mrs. June Sutton Annis, Mrs.

Marilyn Miilanich, Mrs. Connie Shea, Mrs, Anne Storch,

Mrs. Cenevievebbor, Mrs. Emily Wellborne, I-iss Lynnanne

Darnall, Miss Holly Gladstone, Miss Patty LaBrot, Mr.

Bill Geiger, and Mr,. Jeff Kanov.

iv
















TABLE OF CONTENTS


ACKNOWLEDGEMENTS .......................

PROLOGUE.............................. .

LIST OF TABLES..... ................. ....

LIST OF FIGURES .......................

APPENDIX E TABULAR CODE................

CHAPTER


I.


II.

::rI.T

IV.

AP~P~lDI CEC


INTRODUCTION

Related
Stateme:
Method.
Stateme:

RESULTS....

DISCUSSION.

SUMMARY.,..

... 0........

Appendix A.
Appendix B.


Appendix C.
Appendix D.
Appendix E.

BIBLIOGRAPHY .........


*Q. *

0**@


* .. 0*

. *00

'.4..,..


N..............................

Research ........................
nt of Purpose.......... .......
S..... ...... P................
nt of the Problem ...............

................................




....4.....4... ...$. 000*0*.. .. ..






4... 44.... ..... .. .* .. 4... .. ..
44. o. 4.o.. .. 4 4. 4o0
..o............o....o,.....,,.o.
oooooooooooCoeeeoooee~eooo~o


Pago
iv

vi

viii

x

xi



1

2

11
I

20

22

57

65

67

68
72
74
79
88

148


vii










LIST OF TABLES


TABLE Page
I A Simple Analysis of Variance for Differences
Between Pre-Placement Test Scores, last Completed
Programmed Reader Booklet Number, and Post-
Placement Test Scores............. .... ... .......... 23

II A Lindquist Type I Analysis cf Variance for
Differences Between Correct and Incorrect
Programmed Reader Response Rates.................. 25

III The Direction and Magnitude of Differences
Between Correct and incorrect Programmed
Reader Response Rates............................ 26

IV A Lindquist Type VI Analysis of Variance for
Differences Between Programmed Reader Response
Rates and Diagnostic Test Response Rates.......... 29

V A Wilcoxon Signed Rank Test for Differences
Between Before Phase Correct Programmed Reader
Response Rates and During Phase Correct
Programmed Reader Response Rates.................. 32

VI A Wilcoxon Signed Rank Test for Differences
Between Before Phase Incorrect Programmed Reader
Response Rates and During Phase Incorrect Programmed
Reader Response Rates. ...... ....................... 34

VII A Wilcoxon Signed Rank Test for Differences
Between Before Phase Correct Programmed Reader
Response Rates and During Phase Correct
Programmed Reader Response Rates.................. 36

VIII A tWilcoxon Signed Rank Test for Differences
Between Before Phase Incorrect Programmed
Reader Response Rates and During Phase Incorrect
Programmed Reader Response Rates .................. 37

IX A Lindauist Type VI /Analysis of Variance for
Differences Between Before Phase Diagnostic
Test Response Rates and During Phase Diagnostic
Test Response Rates ..... .... .. .... ........ 40

X The Direction and Magnitude of Differences
Between Before Phase Correct and Incorrect
Programmed Reader Responses and During Phase
Correct and Incorrect Prografmned Reader
Responses ..... ............ ....................... 41


viii







LIST OF TABLES (continued)


TABLE Page
XI A Wilcoxon Signed Rank Test for Differences
Between Before Phase Incorrect Programmed Reader
Response Rates and During Phase Incorrect
Programmed Reader Response Rates.................. 44

XII A Wilcoxon Signed Rank Test for Differences Between
Before Phase Correct Programmed Reader Response
Ratesand During Phase Correct Programmed Reader
Response Rates........................ .......... 46

XIII A Wilcoxon Signed Rank Test for Differences
Between During Phase Correct Programmed Reader
Response Rates and After Phase Correct Programmed
Reader Response Rates............................. 48

XIV A Wilcoxon Signed Rank Test for Differences
Between During Phase Incorrect Programmed
Reader Response Rates and After Phase Incorrect
Programmed Reader Response Rates.................. 50

XV A Wilcoxon Signed Rank Test for Differences
Between During Phase Correct Programmed Reader
Response Rates and After Phase Correct Programmed
Reader Response Rates.............. ............... 53

XVI A Wilcoxon Signed Rank Test for Differences
Between During Phase Incorrect Programmed
Reader Response Rates and After Phase Incorrect
Programmed Reader Response Rates.... ........... 54











LIST OF FIGURES


FIGURES Page
I A schematic illustration of the within-subject
design, with replications, used in this study... 15

II Correct and incorrect Programmed Reader
response rates. (Before Phase)................. 28

III The effect of presenting a subsequent
event following correct Programmed Reader
responses on correct Programmed Reader
response rate.............. ...................... .. 33

IV The effect of withdrawing a subsequent
event following incorrect Programmed
Reader responses................................ 35

V The effects of simultaneously presenting
and withdrawing subsequent events following
correct and incorrect Programmed Reader
responses....................................... 38

VI The effect of presenting a subsequent event following
incorrect Programmed Reader responses on incorrect
Programmed Reader response rate................. 45

VII The effect of terminating an arrangement made
to correct Programmed Reader responses on
subsequent correct Programmed Reader response
rate.................................. ......... 49

VIII The effect of terminating an arrangement made to
incorrect Programmed Reader responses on
subsequent incorrect Programmed Reader response
rate ............. ... ................. ........ 51

IX TIL effect of simultaneously terminating
arrangements made to correct and incorrect
Programmed Reader responses on correct and
incorrect Programmed Reader response rates...... 55












Appendix E
TABULAR CODE


SRP-BC Programed Reader, Before Phase Correct
SRP-DC Programed Reader, During Phase Correct
SRP-AC Programed Reader, After Phase Correct
SRP-BI Programed Reader, Before Phase Incorrect
SRP-DI Programed Reader, During Phase Incorrect
SRP-AI Programed Reader, After Phase Incorrect


DT-BC Diagnostic Test, Before Phase Correct
DT-DC Diagnostic Test, During Phase Correct
DT-AC Diagnostic Test, After Phase Correct
DT-BI Diagnostic Test, Before Phase Incorrect
DT-DI Diagnostic Test, During Phase Incorrect
DT-AI Diagnostic Test, After Phase Incorrect













Chapter I


INTRODUCTION


Education is exclusively concerned with behavior

change for the purpose of developing and maintaining complex

repertoires of culturally-valued human behavior. In order

to evaluate the effectiveness of its procedures, education

must have a reliable and sensitive method for describing

and measuring the behavior changes it produces. Currently,

such evaluation is performed almost totally by psychoinetry --

psychc-educational tests and rating scales. Despite its

undoubted importance historically, psychometcy has now been

shown to have serious shortcomings of both an empirical

and ethicalnature. These deficits have been examined in

great detail by Hoffman (1962, 1965, 1967). Hoffman's

documented discussions include statistical misuse in the

development of tests, misuse of test results, fallacies

in the use of pre-tests, as well as the corrupting effects

tests have on educational practice.

Despite these deficits, educators continue to

use psycho-educational tests and rating scales in most

evaluation research, apparently because they are unaware of

more acceptable alternatives. Tests and rating scales

are used to evaluate behavior changes in children, teachers,

and school administrators, as well as to assess the effects

1







of curricula and teaching methods on behavior. With the

partial exception of achievement tests, psychometric

procedures never directly measure the behavior in question.

Unfortunately, the trend in recent years would seem to be

away from more simple and direct measures of behavior and

toward procedures which rely highly on indirect measurement

and a complex chain of inferential statements mediated by

hypothetical constructs. The Illinois Test of Psycholinguistic

Abilities is an example of this trend.

The present dissertation seeks to utilize free

operant technology, in particular the direct and continuous

recording of pupil performance, to evaluate and analyze a

programed instruction curriculum material. At the same time,

the records gathered can be used to further understand free

operant technology itself. Therefore, the dissertation will

describe not only human behavior change as a function of

the programed material, but it .iill '-escribe the effects

of the technology on pupil performance rates. The findings

of this type of research are directly applicable to children

and may he directly related to further applications in

which the evaluation of educational procedures is of concern.


Related Research

The Evaluation of Programed Instruction

A review of the literature relating to the evaluation

of programed instruction reveals; hat there is no generally

accepted method available. There are, however, several

methods which are used with some frequency.







The first of these is the checklist. Although

Newman (1965) has recommended against the development and

use of checklists without first having conducted exhaustive

research on their reliability and validity, the checklist has

nevertheless become the most popular and convenient method

of evaluating programed instruction.

Before 1963, checklists were commonly created by

individual researchers for their own purposes (Fry, 1963;

Hughes, 1962). Most of these checklists were quite

heterogeneous, and rested more on considered opinion and

theoretical orientation than on empirical research.

In 1962, The Joint Committee for Criteria for

Assessing Instructional Programs began publishing checklists

for use in evaluating programmed instruction. These lists

were revised yearly by the committee, on the basis of

their demonstrated utility in the applied situation. The

absence of systematic research in evaluating checklists

produced an instrument that was, at best, crude.

Other checklists were later developed (Jacobs, Maier,

and Stolurow, 1966), but in the final analysis, the Joint

Committee's checklist has become the standard in the field. It

has been endorsed by both the National Society for the Study of

Education (NSSE) and the Division of Audio-visual Instruction (DAVI)

of the National Educational Association, the two organizations

most concerned with quality standards in education.

Ironically enough, the chief rival of the Joint

O mmittee's checklist is another well-rospacted educational

standard that was not developed specifically for evaluating







instruction. Bloom's Taxonomy of Educational Objectives

(1956) has been repeatedly invoked as the evaluative

standard form a "cognitive" point of view. The chief

spokesman for this type of evaluation is Louise Tyler

(1966). Not satisfied with theory only, she has reported

some data on its use by teachers already familiar with the

taxonomy. Newman (1965) has used it to evaluate programed

instruction in the Social Studies.

The second method of evaluation that can be

distinguished might more properly be called the comparison

procedure. It generally takes the form of comparing

programed instruction with traditional teaching methods,

or programed texts with standard texts.

Schramm (1964) indicates that much of the

evaluative research done in programed instruction is of

this nature. An examination of Educational Abstracts

confirms this statement for the subsequent years. The

difficulty, however, is that comparison, as a research

method, reveals little or nothing about the programed

instructional material per se. Nevertheless, the Joint

Committee recommends comparison as a method for the external

validation of programmed material, and it may be an acceptable

procedure for this task.

Another difficulty with comparison studies lies

in the nature of the research design employed. It is almost

impossible, in this type of research, to control for

individual differences in children and in teachers, and

its even more difficult to control for differences in content.





5

The majority of research of the comparative type, as Schramm

notes, is so poorly done that little faith can be placed

in the validity of the results. There are fortunate exceptions,

particularly with regard to exceptional child populations.

Blackman and Copobianco (1965), for example, report on

the use of a specific programed material with retarded

children. Likewise, Rainey and Kelly (1967) report

the use of a time-telling program with educable retardates,

and Streng (1964) reports evaluating a program with deaf

populations. This research is child-oriented and involves

the determination of the utility of a specific program for

developing a defined behavior in a given exceptional child

population. The utility of this type of research for the

practicing classroom teacher should not be under-estimated.

Closely related to the programed instruction with

"other" comparison, is the programed instruction with

achievement test comparison. Normally, the test used is

one of the standard achievement tests such as the Wide Range

Achievement Test or the Metropolitan Achievement Test. In

this case, the research questions are directed to difference

scores on the specific test before and after the administration

of the programed instruction material. The discontinuous

(before and after) nature of this procedure is a serious

shortcoming because it does not allow a point to point

analysis of the relationship between the program and the

child's behavior. This method, like the programmed instruction

with "other" comparison, is recommended by the Joint

Committee for the external validation of the material.







The careful reader of research is quick to note

that the dependent variable in these studies is not programed

instruction performance, but achievement test performance.

Programed instruction performance is then inferred from

the test performance.

In his text on evaluating programed instruction

Jacobs, et al., (1966), mentions the Denver Study as the classic

model for evaluation. This study, reported in toto by Jacobs,

utilized boththe programed instruction with "other" comparison

and the programed instruction with achievement test types

of evaluation. The questions asked in this study were:

1) Do classes taught by the program only,
by a teacher only, and by a combination
of a teacher plus the program differ in
the outcomes of levels of achievement,
attitudes toward programed instruction
and interest in Spanish?

2) How are the input characteristics of initial
achievement, academic aptitude, and attitude
towards Spanish related to the outcomes
of achievement in interest in Spanish in
each instructional group? For example, do
the brighter classes learn more than the
slower ones in each group?

3) Are teacher's attitudes towards Spanish
and various teaching methods related to
the instructional methods used in the study?

A careful examination of these questions reveals

that in no case is programed instruction performance the

dependent variable. This study, the apparent classic in

the evaluation of progra-:ed instruction, is in fact not

an evaluation of the internal aspects of programs. Because

it is an evaluation about programmed instruction as it

relates tovarious dependent variables such as achievement







test and attitude scale scores, no direct statement may

be made regarding the behavior change on the program itself.

Also mentioned in the evaluation literature,

but clearly not research, is a caution to the program

user to check the credentials of the program author and

the publisher. The Joint Committee advises all publishers

to include with each program sold, complete developmental

and utilization testing data. The advisement, in practice,

is little heeded.

In summary, there are four main methods of evaluating

programed instruction material. These are:

1) the checklist,

2) Bloom's Taxonomy of Educational Objectives,

3) programed instruction with "other" comparison,

4) programed instruction with achievem.ent test
evaluation.

It is clear from this review that programed

instruction performance is typically not the dependent

variable in the evaluation of programed instruction

materials.


Fre Operant Research in Education

In the history of psychology, the u.se of free

operant techniques in the analysis of human behavior is

relatively recent. It was only in 1949 that Fuller demonstrated

experimental control over a vegetative mental retardate.

Skinner's text, using knowledge acquired in the study of

free operant behavior of animals to describe human behavior,

appeared in 1953. In it, an empirical framework for the





8

experimental analysis of human behavior was presented. Sound

experimental data were not forthcoming until Lindsley's classic

study with chronic psychotics (1956). Skinner (1958), reporting

on his research with programed instruction technology, an

extension of free operant techniques with animals, excited

much interest in the educational community. Bijou (1957, 1958)

developed observation techniques for young children patterned

closely after those used by Lindsley.

The marriage between the educator and free operant methods

was not long in coming. Birnbrauer, Bijou, Wolf, and Kidder

(1965) demonstrated the application of free operant techniques

in a classroom situation using programed instruction as a

curricular core. Zimmerman and Zimmerman (1962) also applied

free operant techniques in a classroom with much less structure

than Birnbrauer's classroom.

Ayllon and Azrin (1964) demonstrated the functional

utility of token economies in shaping the behavior of patients in

a mental hospital. Girardeau and Spradlin (1964) used the

same type of token control with retardates at the Parsons

State Home and Training School.

At that time, however, there was no systematic or

standard method in human free operant research. Although a

precise language was available for the description of animal

behavior (Ferster and Skinner, 1957), its application to the

human situation was confusing and left much to be desired.

Research reports were presented in the literature with method

and discussion sections so radically non-standard that scientific

replication (Sidman, 1964) was virtually impossible.







O. R. Lindsley (1964) put forth a numerical-

temporal descriptive language. Its purpose was to precisely

describe behavior and those events related to behavior,

either in number or in time. This was followed by a

major strategy statement indicating that behavior change

must be produced by teachers and parents in order to meet

the existent need (Lindsley, 1968). Lindsley (1966)

has coined the term precision teaching to describe the use

of free operant methods by teachers. The details of

precision teaching and the descriptive language are available

elsewhere (Koenig, 1967; Caldwell, 1967; Haughton, 1967)

and will not be treated at length here.

Although rate was the accepted datum unit for

animal free operant research, researchers using human

subjects felt little obligation to follow suit. Instead,

many relied on the more standard educational datum units

of absolute number and percent. In a study of the

sensitivity of the various datum units to behavior change,

Holzschuh and Dobbs (1966) demonstrated that rate was

consistently more sensitive than other units examined.

Subsequent research (Caldwell, 1966; Johnson, 1967b) has

supported this finding.

Free operant techniques in general, and precision

teaching in particular, have found wide acceptance in special

education (Haring and Schiefelbusch, 1967). Operant techniques

have been used almost exclusively with single organisms.

Because special educators have long been tailoring curricular

programs for individual children, the techniques seemed





10

ideally suited for that area of education concerned primarily

with individual differences.

Koenig (1967) used precision teaching in a

classroom setting with emotionally disturbed children to

examine a wide range of academic and disruptive behaviors.

Haughton (1966) demonstrated the functional utility of

direct and continuous recording of behavior compared with

an examination of achievement tests in predicting pupil

performance. Johnson (1967c) demonstrated dramatically

that achievement tests and pupil performance on similar

material generated different performance rates when both

pupil performance and achievement test performance were

directly recorded. Johnson (1967a) found that teacher-

planned rates (number of problems assigned by the teacher

divided by the number of minutes allotted by the teacher

for their solution) were, in part, determiners of subsequent

pupil performance rate.

Johnson (1969) is currently engaged in the use

of precision teaching to evaluate the Science Research

Associates arithmetic series. This study, as yet unpublished,

is the only reference known to this writer on the evaluation

of any curriculum through the use of free operant techniques.

In summary, free operant techniques originally

used exclusively in the study of animal behavior, have, in

recent years, seen wide application to human performance.

When used in classroom settings, the techniques may be

collectively referred to as precision teaching (Lindsley,

1966). Although precision teaching has been applied to







a broad range of educational problems, only in one case

has it been used in the analysis of curricular material.


Statement of Purpose

The purpose of this dissertation is to assess

the usefulness of free operant technology (precision

teaching) in the analysis of programmed curricular materials.


Method

Subjects

Fifty-nine children, thirty-seven boys and

twenty-two girls from two Alachua County, Florida, elementary

schools, Stephen l'oster Elementary and Duval Elementary,

participated. The children were placed in grades three

through six and were referred to the writer by their regular

teachers. The sole referral criterion was that the child

be two or more years behind in his reading performance.

Children were not excluded because of low 10 test scores.


Teachers

Eleven teachers participated in this study. All

were graduate students in the Department of Special Education

at the University of Florida. Nine were Mastefs candidates,

one an Ed.D. candidate, and one a Ph.D. candidate. All

teachers received some form of graduate credit for their

participation in the study.


Teaching Situation

The children in the study left their regular

classes at a predetermined time to meet in a group with the





12

special teacher. The teaching situations varied from teacher

to teacher, but in general can be described as poor. Lack

of space necessitated some classes meeting in cafeterias,

gyms, halls, and the like. In all cases, however, the school

administrators in each school made every effort to insure

the best teaching situation available in the specific

circumstances


CurricularMaterial

The Sullivan Reading Program (1964) was selected

for use in this study. Consultation, initially, with

educational specialists and, later, with teachers using the

program, indicated that it enjoyed wide acceptance and was

considered educationally sound.

In its entirety, the Sullivan Reading Program

is not all programed instruction. Also included are

storybooks, filmstrips, and end-of-book tests. These

materials are not programed. They were not used in this

study.

The programed instruction portion of the Sullivan

Reading Program is presented in three series. Series I

includes Programnmed Reading Booklets 1-7 (Grade 1), Series

II includes Programmed Reading Booklets 8-14 (Grade 2),

and Series III includes Progranmmed Reading Booklets 15-21

(Grade 3). A sample of the program from Series I is

presented in Appendix A.

Following approximately each fifty frames, a

Diagnostic Test is scheduled. The Diagnostic Test, unlike







the programed text proper, has no answers in the answer

column. The purpose of the Diagnostic Test is to give the

teacher a check on the progress of the student by presenting

a sample of the content presented in the preceding fifty

frames. A sample Diagnostic Test is presented in Appendix B.

To determine the individual child's proper

starting point in the Programmed Reader, a Placement Test

is provided with the Sullivan Reading Program. The

Placement Test is similar to the Diagnostic Test in that

it has no answers. Each two pages in the Placement Test

summarize the content in one Programmed Reader. The child

begins in the booklet indicated by his first error in the

Placement Test. A sample Placement Test is presented in

Appendix C.


Experimental Design

The purpose of this study was to assess the

usefulness of free operant technology, precision teaching,

in the analysis of programed instruction material, specifically

the material in the Programmed Reader of the Sullivan

Reading Program. The experimental design had to meet two

requirements:

1) it had to use the Programmed Reader exactly

as indicated in the Teacher's Guide, and

2) it had to apply free operant methods to the

recording and modification of Programmed

Reader Performance.

The within-subject design, using each child as his own

control, seemed an ideal choice. Following Sidman's (1964)






suggestion, the experiment was conducted :in phases (the

Before Phase, the During Phase and the After Phase) with

replications across children and teachers. This particular

design also allowed subsequent between-subject parametric

analysis of many of the questions. This is schematically

illustrated in FIGURE I.

In the first phase, the Before Phase, the

Placement Test was administered and each child began in

the Programmed Reader Booklet indicated by his Placement

Test score. This phase lasted approximately eleven days

and established the baselines needed to evaluate the effects

of the curriculum and the teaching procedure. In the second

phase, the During Phase, all experimental manipulations

of independent variables took place. This phase lasted

approximately eleven days. The last phase, the After

Phase, was a replication of the Before Phase with no

experimental manipulations in effect, and with the Placement

Test being readministered at the conclusion of approximately

eleven days.


Experimental design applied in the classroom

The teacher in the classroom intending to use

the Programmed Reader from the Sullivan Reading Program

must first determine in which Programmed Reader, (Booklet

1-15) each child is to begin. The Placement Test is provided

for this purpose, If performance on the Placement Test is

not equivalent to performance in the Programmed Reader,

then the Placement Test is not performing its stated function

with respect to placing the pupil in his correct be-gi.ining booklet.


































4-J





4J
0*
*nt







J4J-
rna









.4
4-1
I








0
fO







OC
Ul Cf





4Jl





HQ
KcI





4J




OC
S(
^ ^
rtl O
rln
14-1 k


C en
0 ^
*H
4J-
n34-
!- n


H
.-4 (N y
<<< 0 .




,-I












ro
,-I



U











H








r-4








'8


,- (N














rH-
*-i






























u
0
rNm







*C1









C
U



*H


U







The careful teacher will want to know if, indeed,

the Placement Test score does approximate the level of

performanceof the booklet indicated. To investigate this

question, the Placement Test was readministered on the

last day of the After Phase. If the Placement Test was

correctly indicating Programmed Reader performance level, the

final administration of the test should have indicated the

book in which the child was last performing.

Skinner points out (1954, 1958) that one of the

central goals of programed instruction is to maximize

correct responding and minimize incorrect responding.

This is accomplished by carefully constructing each programed

frame (antecedent event) in the program. The teacher using

the Sullivan Reading Program naturally wishes to know whether

the Programmed Reader does, in fact, differentiate correct

responding from incorrect responding.

To examine this question, correct and incorrect

response rates on the Programmed Reader in the Before

Phase were analyzed. If the Programmed Reader differentiated

correct from incorrect responses, then a difference should

exist between the two response rates.

The most recent literature in programed instruction

(NSSE, 1967) concerning the development of programs suggests

"in-program" checks with frames similar to those in the

regular program, but without the correct solutions available.

The Programmed Reader of the Sullivan Reading Program has

incorporated this principle in the Diagnostic Test, The

Diagnostic Test, according to the Teacher's Guide to







Programmed Reading (1964), is not to be graded, but instead

to be used as a guide to check on the quality of the students'

work.

The careful teacher will want to know if performance

on the Diagnostic Test is equivalent to performance on the

Programmed Reader before she alters her supporting curriculum

on the basis of this quality check.

To evaluate this problem, correct and incorrect

performance rates, on both the Diagnostic Test and the

Programmed Reader, collected in the Before Phase were

analyzed. If the Diagnostic Test and the Programmed Reader

are equivalent, there should be no difference between the

performance rates on the two programs.

The teacher who has taught with the Programmed

Readers for a period of time is able to assess the performance

of her students. By examining each child's rate correct

and rate incorrect as plotted on his six-cycle semi-log graph

and recorded on his data sheets (Appendix D), she can

decide on the best tactic to maximize his performance

accuracy. That is, the teacher can plan how to increase

the difference between correct response rate and incorrect

response rate. One of the most common tactics to this end

is the arrangement, the presentation orwithdrawal, of a

subsequent event following each response or series of

responses made in the curricular material. This arrangement

can be nade with :-he intent of accelerating correct response

rate by presenting the subsequent event; or decelerating

incorrect response rate by withdrawing the subsequent event,

or both.





18

The teacher who makes such an arrangement for each

child will want to know precisely what effect it has on

the child's behavior. In addition, she will want to know

what effectthe arrangement procedure, itself, has on the

performance of her class as a whole.

To investigate this problem, the data gathered

in the Before Phase, where no experimental conditions were

in effect, and in the During Phase, where arrangements

were in effect, were compared. If the subsequent event

arranged to follow the Programmed Reader correct and/or

incorrect response produced an effect on response rate,

this wouldbe seen as a difference between response rates

in the Before Phase and response rates in the During Phase.

The classroom teacher is only too aware that often

pupils are not motivated to perform on test items. It

might interest her to know whether, on the Programmed

Reader, a subsequent event presented to the child following

each errorless Diagnostic Test will alter performance rate

on the Diagnostic Test. To investigate this problem,

correct and incorrect performance rates on the Diagnostic

Test in the Before Phase and in the During Phase were

analyzed. If the subsequent event following each errorless

Diagnostic Test produced an effect on correct and/or

incorrect response rate, this would be seen as a difference

between response rates in the Before Phase and response

rates in the During Phase.

The inquisitive teacher, when she presents a

subsequent event following only the correct Programmed







Reader response will want to know whether the correct

response rate accelerates or not, and exactly what happens

to incorrect response rate in this condition. In order to

answer this question, incorrect response rates on the

Programmed Reader in the Before Phase and in the During

Phase were compared. If the subsequent event presented

following the correct Programmed Reader response had any

effect on incorrect Prcgrammed Reader response rate, this

would be seen as a difference between incorrect response

rates in the Before Phase and incorrect response rates in

the During Phase.

Similarly, the teacher will want to know what

the effect on correct response rate is when she withdraws

a subsequent event following the incorrect Programmed

Reader response. The same procedure was repeated for the

correct response rates in the Before Phase and in the

During Phase for this comparison. If the subsequent

event withdrawn following an incorrect Programmed Reader

response had any effect on correct Programmed Reader

response rate, this would be seen as a difference between

correct response rates in the Before Phase and correct

response rates in the During Phase.

Since the purpose of arranging a subsequent

event is to produce a lasting change in the magnitude of

the difference between correct and incorrect response rates

maintained by the natural consequences of superior achievement,

the careful teacher will want to know precisely what the

effect of the removal of an accelerating or decelerating





20

consequence has on subsequent performance. The problem

was investigated by comparing correct and incorrect Programmed

Reader response rates in the During and After Phases. If

the removal of a consequating condition following Programmed

Reader performance produced a subsequent change in Programmed

Reader response rate, that change would be observed in a

difference between performance rate in the During Phase and

performance rate in the After Phase.


Statement of the Problem

The analyses indicated above may be condensed and

summarized in the following eight questions:

1) Is the Placement Test score equivalent to

indicated-Programmed Reader book number?

2) Is correct Programmed Reader response rate

different from incorrect Programmed Reader

response rate?

3) Is performance rate on the Diagnostic Test

equivalent to performance rate on the

Programmed Reader?

4) What is the effect of the arrangement of

a subsequent event following the Programmed

Reader response on Programmed Reader response

rate?

5) What is the effect of a subsequent event

following each errorless Diagnostic Test on

Diagnostic Test performance rate?

6) What is the effect of the arrangement of a

subsequent event following each Programmed Reader





21

correct response on Programmed Reader incorrect

response rate?

7) What is the effect of the arrangement of a

subsequent event following a Programmed

Reader incorrect response on Programmed

Reader correct response rate?

8) What is the effect of the removal of the

accelerating and decelerating consequences

following the Programmed Reader response on

Programmed Reader performance rate?













Chapter II


RESULTS


Placement Test Accuracy

The Sullivan Reading Program is organized so that

a child may begin at any performance level from grades 1

through 4. The Programmed Reader booklet in which he

begins is determined by his performance on the Placement

Test. The child begins in the booklet indicated by the

location of his first error in the Placement Test.

To determine the accuracy of the Placement Test,

the scores on pre-Placement Tests administered in the

Before Phase, the scores on post-Placement Tests administered

at the conclusion of the After Phase, and the booklet

numbers of the last Programmed Readers completed in the

After Phase, were compared by means of a simple analysis

of variance. The results, presented in TABLE I, suggest

that a reliable difference existed among the three measures.

A t test between the last Programmed Reader booklet

completed and the post-Placement Test yielded a t of 12.66,

p .01. Strong support is lent to the conclusion that

the Placement Test was not a reliable estimate of Programmed

Reader performance.






















TABLE I

A Simple Analysis of Variance for Differences Betweenn
Pre-Placement Test Scores, last Completed Progranmed
Reader Booklet Number, and Post-Placement Test Scores




Source of Variation df Variance F
Estimate


Between Groups 2 278.48 45.2*

Within Groups 56 6.27

Total 58 15.55

*P-<-. -----------------------------------







The Difference Between Correct and Incorrect Programmed Reader
Performance Rates

In setting up guidelines for writing programs,

Skinner (1954) pointed out that the well-constructed

program should maximize correct responding while permitting

a minimum number of error responses. The question of this

difference was examined in the Programmed Reader by treating

the correct and incorrect Programmed Reader response rates

collected in the Before Phase with the Lindquist Type I

Analysis of Variance (Lindquist, 1953). TABLE II indicates

that correct and incorrect response rates on the Programmed

Reader were well differentiated. TABLE III and FIGURE II

display the direction and magnitude of this difference for

the entire population and for one typical child, respectively.


The Difference Between Programmed Reader Performance Rates
and Diagnostic Test Performance Rates

In the Programmed Reader booklets, a Diagnostic Test is

scheduled approximately each 50 frames. The Diagnostic

Test is intended to be a quick check on the accuracy of

pupil performance. If this check is to be meaningful,

the correct and incorrect Diagnostic Test performance rates

must be approximately equal to the correct and incorrect

performance onthe Programmed Reader itself.

The extent of the difference was analyzed with

the Lindquist Type VI Analysis of Variance. This analysis,

presented in TABLE IV, considered class units, in addition

to both correct and incorrect response rates on the Diagnostic

Test and the Programmed Reader. It is clear that a reliable























4J

N 0 CL
M M o Im Ln c.
u iLn Il CN
C 0 coL


i ro A r r





5CO
w w 0
"A 0

0 W
ciC)









"W ( 0 3 0J1 q 0 -
a O .o




O C- O
H a I II

H 0









00 0
E-i 44 O r- ^ i r-1< i l
4J 9 i II

0 40
CH 0
0 04

0 C 0 0 z ( 0



* 04 -4< 0) 4 C p
Y 'IOo c, 4CI4 L O

di) (0 0 U HEl p 91 0
U r-H ) 0 H- 4 0 0 r1 0
:1 -I 0 H3 4 *^-ci pO X p4 4-' 0
0 (d .C p X. 0 4 0
En> U WI UUl--< W E-4







TABLE III

The Direction and Magnitude of Differences Between Correct
and Incorrect Programmed Reader Response Rates




Direction of Magnitude of
Child Change Change


Allen + 3.20
Ballard + 12.60
Carver + 6.35
Hathway + 5.15
Williams + 9.70
Langston + 6,35
Hines + 3.50
Willford + 2.90
Young + 5.45
Godbolt + 2.40
Nattiel + 3.70
Ford + 2,70
Brown, M. + 13.75
Wright + 6.95
Jones + 6.95
Jeffcoat + 4.50
Fogarty + 6.85
Howell + 14.83
Jackson + 7,20
Wimms + 5.75
Thomas + 4.85
Klickly + 4.55
Lesene + 4.50
Webb + 4.40
Morris, P. + 4.06
Morris, B. + 9.67
Johnson + 4.35
Hardwic + 4.15
Cray + 5.71
Iayes + 7.29
Camps + 6.02
Kelly + 5.90
Railey + 1.90
Sperring + 7.15
Taylor + 3.50
Bruce + 4.65
Hague + 9.10
Howell, D. + 4.90
Beals + 4.80
Bishop + 3.70
Somese + 3.80
Lee + 3.40







TABLE III Continued


Direction of Magnitude of
Child Change Change


Pattison + 4.70
Ross + 5.90
Brown -4. 13
James + 5.68
Stewart + 4.68
Walker + 5.93
Berry + 4.62
Haile + 2.20
Alexander + 2.55
Burke + 2.30
Brown, K. + 4.05
Davis + 2.20
Smith, J. + 2.05
Bass + 2.95
Fogarty + 2.90
Howard + 5.90
Shaupe + 3.48















0
0
I.. cU o"
jIii IT 1TTtT FI1 tillirii J 'I' I w i 1
I -TT I ii lIT1


lli- : II '[ ir I I
*1 / I i fi t t' f .. J
':~ i-I I rj'!1 I jI I 8
L o


4---r~-. I I- i I:
S- l -IIL M
1 77- 1 T-7
A::


Hji i i ~~r'; I"~ I -f~ I
(Q41!i 1 4 i
4 0.






P-4 (~1) La~rl~
04- I1J'ci '.1tlTt----- ---- -v' r_Kt1 H- --






i C/i
0 ~~ -l) __ ___-U
4j OKI~_--







41 u
rd C., u ~ -~-
-.3 Ic,

~ Ii







1 'I I L il L ---
,,j I 11 ii I ~--;]_ : -_-- It










T d I L k i ll J 0

01 0 8 0
J3fl'IV Wjd SINIMhAON
























L* a i N co
0H H


.i'
It





























00 00
;n r-4 C"


S0
O0

0 E-4

& 0
G


O 0
o1 *(-1






* 0
4O


kC


C)
r4 d






4-4 0



H 0
(0
, ra







r0
0i
n3


e-4
















4-
1


NO
0 CJ

CN
I U


re Co Ln zr itn kD
N un -A Hn C tn















H in oo H
r-( r-i rH


SQ -
r<

C
0 o 0
O (
HE n
rd<
*r^ (J
Lc ^
1U U


CO0000
d' co
* *
















-I


4-)
U
'-4




14 0


S0 Ut
u 0 *X

0 tn
U 0 : <
M 3QO
CO r
Otnto


0 0
*-1

U rd

0 0i
r f"


4v -1

S0

w W
OO3-

3-3-


c m

00
33
^ i:1:







difference did exist between performance rates on the

Diagnostic Test and on the Programmrned Reader, with the latter

having the higher rates. In addition, differences were

observed between correct and incorrect response rates

(a result consistent with TABLE II) and in performance

rates among class units.

A graphic display of the means in the interaction

terms revealed that: 1) the B x C interaction could be

attributed to the differences observed between the mean

incorrect response rate on the Diagnostic Test and the mean

incorrect response rate on the Programmed Reader, and 2)

the A x B interaction could be attributed to a pooling of

four distinct class units on correct response rates for

both the Diagnostic Test and Programmed Reader.


The Effects of Arranging a Subsequent Event Following
Programmed Reader ResIponses

The precision teacher who wishes to change the

rate of correct and/or incorrect responding often achieves

this end by arranging a subsequent event (e.g., penny,

star, M&M, etc.) to follow the desired response. This

procedure of presenting or withdrawing subsequent events

may be used to either accelerate or decelerate the response

rate. The first possible procedure, presenting a subsequent

event following correct Programmed Reader responses, while

ignoring incorrect Programmed Reader responses, was

examined by treating the correct Programmed Reader response

rates collected in the Before and During Phases of the experiment

with the Wilcoxon Signed Rank Test (Wilcoxon and Wilcox, 1964).






The result, presented in TABLE V, suggests that a reliable

difference was present between the performance rates in the

two phases. An examination of column "d" shows that most

of the response rates were accelerated in the During Phase.

It should be noted that the magnitude of each of the seven

accelerations was far greater than the magnitude of each

of the two decelerations. A graph of this acceleration is

presented in FIGURE III.

The second possible procedure, withdrawing a

subsequent event after incorrect Programmed Reader responses,

while ignoring correct Programmed Reader responses, was also

treated with the Wilcoxon Signed Rank Test. The result as

presented in TABLE VI, indicates a reliable difference

between the performance rates in the two phases. An

examination of column "d" shows that the response rates

here were decelerated in the During Phase. Once again,

the magnitude of the changes was largest in the expected

direction of the change, deceleration in this case. A

graph of this deceleration is presented in FIGURE IV.

The third possible procedure, the combination of

the first two arrangements, was also treated with the

Wilcoxon statistic. TABLES VII and VIII present the

results of the analyses. It should be noted the results of

this combination of arrangements are strikingly similar to

each of the arrangements alone for their respective

responses. FIGUPRE V displays this simultaneous acceleration

and deceleration.


























0 l
O a-i
S M-l 4-4 -I
a) a 04 -
4- 0 $4




Q-4 E E-
m 0

zO M,
E-) I
U 0 p 4

















5.O
SU) 0 0

4-H 4 H (0




p co ***)*** *) V



o0 -o
4-4O M M Nr-nI MN m

1.r, 44



O 0 fU 9 4 a a 1 *
r 0 (p a)0 4 oI'D r-m om -0 r- yrm
M(3 QP r-I



P.0 U) 0 UC VmmN
C-i O



C (C g

U Pf pk 0 4J 0
El) a) >1 ul M -H 0
U 0 < i r-I P > >- n
CQ J M 0) r>U ( -
r, mw I mu -mm

















d
aiRO
U









*ill I. liii
)7~f ..I~i.. I I: 7 :I ......-...--. I!-
9H_! K.ii : H' j: I ..............
U :. I I iVI



J: :ii I i i 'i


I--~~ I: i

cnI .. I *!ii.,: 1 1I

I~ I.





11L

J1OJ~J.I33 .1N''JA0


Iccr wta.




0 =





[V I N


II
III


r4

r-40a)

0
> 0

>0




43





0 r_
ui

UI4
EC-


4rdl




0)
4-Ad
4J4
00
Wc)c




C)0
'C


.0
CL

o




00


00

0)
O








ol
N, .
-J
(A
(A








-o
7
C'













TABLE VI

A Wilcoxon Signed Rank Test for Differences Between
Before Phase Incorrect Programmed Reader Response Rates
and During Phase Incorrect Programmed Reader Response Rates



Before During Ranirk f Rank of D
Child Phase Phase d D with Less
_____ __Frequent Sign

Willford .75 .40 -.35 16
Ballard .39 .05 -.34 15
Williams .24 .08 -.16 10
Brown .29 .24 -.05 4
Howell .23 .46 +.23 13 13
Kelly .54 .03 -.51 17
Davis .10 .12 +.02 1.5 1.5
Brown .08 .02 -.06 5.5
Burke .22 .07 -.15 9
Alexander .09 .15 +.06 5.5 5.5
Haile .36 .14 -.22 11.5
Morris, B. .42 .10 -.32 14
Morris, P. 1.06 .20 -.86 18
Jackson 21 .13 -.08 7
Klickly .25 .24 -.01 1
Lesene .42 .20 -.22 11.5
Thomas .69 .59 -.10 8
Wimms .09 .11 +.02 1.5 1.5

T = 21.5*


*P < .005


















rnl)jcr Uo.


0
r-i
IH
0



tj .



tr
>0

4i-4


IDrd
.Q Q)




rcQ


17,

*Z 0
.Ir-

00

44~
4I4
'4-I


a
















I-
U


1 I

~~~1~ ~ t jjI i t
I LI II: I- j



I.-t.'oiII::: i II IiIII -I I'.-'

Ii I I '4J




j :i~1 I I lti1 lr .




~-I. 0


:1.ny
1iiiL "2 j :t 111 i;i I-iH
Ii ~~1Li 10:


3IlNV : 3 SL)JA1


.,-














w


Ci)
w
0









I:


C..










TABLE VII

A Wilcoxon Signed Rank Test for Differences Between Before
Phase Correct Programmed Reader Response Rates and During
Phase Correct Programmed Reader Response Rates



Before During Rank of Rank of D
Child Phase Phase d D with Less
Frequent Sign

Shaupe 3.74 3.08 -.66 9 9
Smith, J. 3.01 2.31 -.70 10 10
Bass 3.45 5.94 +2.49 25
Fogarty 3.20 5.23 +2.03 22
Howard 6.24 4.47 -1.77 21 21
Ford 3.35 3.58 +.23 4
Godbolt 3.00 3.36 +.36 7
Hines 3.91 5.23 +1.32 19
Nattiel 3.65 4.71 +1.06 16.5
Young 5.25 5.87 +.62 8
Allen 2.41 1.67 -.74 12 12
Carver 6.51 6.17 -.34 6 6
Hathway 5.70 6.73 +1.03 13.5
Langston 6.52 7.55 +1.03 13.5
Jones 6.78 7.49 +.71 11
Wright 7.35 11.10 +3.75 30
Jeffcoat 4.86 7.63 +2.27 29
Fogarty 7.21 12.57 +5.36 31
Beals 5.26 4.19 -1.07 18 18
Lee 3.78 3.97 +.19 3
Ross 6.23 6.24 +.01 1
Somese 4.21 6.86 +2.65 26
Cray 6.77 4.11 -2.66 27 27
Hardwic 5.06 3.12 -1.06 16.5
Johnson 6.12 3.39 -2.73 28 28
Webb 5.27 4.22 -1.05 15 15
Railey 2.48 2.43 -.05 2 2
Sperring 7.50 9.20 +1.70 20
Taylor 4.05 6.31 +2.26 23
Bruce 4.78 7.25 +2.74 24
Hague 9.10 15.20 +6.10 32
Howell 5.64 5.38 -.22 5 5

T = 153*
z = 2.07

*P < .02










TABLE VIII

A Wilcoxon Signed Rank Test for Differences Between Before
Phase Incorrect Programmed Reader Response Rates and During
Phase Incorrect Programmed Reader Response Rates



Before During Rank of Rank of D
Child Phase Phase d D with Less
.Frequent Sign

Shaupe .27 .11 -.16 14.5
Smith .08 .04 -.04 1.5
Bass .20 .11 -.09 5
Fogarty .18 .07 -.11 9.5
Howard .16 .28 +.12 11 11
Ford .06 .42 +.36 27.5 27.5
Godbolt .54 .33 -.21 19
Hines .34 .13 -.21 19
Nattiel .21 .01 -.30 24
Young .20 .10 -,10 7
Allen .45 .11 -.34 25
Carver .45 .08 -.37 29
Hathway .36 ,21 -.15 13
Langston .28 .12 -.16 14.5
James .56 .16 +.04 1.5 1.5
Wright .37 .16 -.21 19
Jeffcoat .25 .15 -.10 7
Fogarty .15 .15 -0
Beal .43 .21 -.22 21
Lee .30 .15 -.17 16
Ross .41 .16 -.25 22
Somese .16 .08 -.08 4
Cray .32 .18 -.14 12
Hardwic .50 .14 -.36 27.5
Johnson .56 .21 -.35 26
Webb .92 .20 -.72 31
Railey .44 .04 -.40 30
Sperring .40 .14 -.26 23
Taylor .19 .09 -.10 7
Bruce ,12 .01 -.11 9.5
Hague .18 .00 -.18 17
Howell .09 .14 +.05 3 3

T = 48*
z = 4.01

*P < .000 3


















PRO


-~- L


; I I H' I: U 'lI: :i,
],:1l





L:1. ';-- I i






tt


*'h1 1 ,. t.. 1 I


:itI i 1:

..1~1


rj 41



.H 0

ru



Q)

nla,


r-1 H ru
C)Q)





r- 0' rd
ukrd
:jkv









(4-4 W)
-Ii


Ud,









C)4~
tn~
Eo-
( c) C)
(9r) E
Uvr
c)it


-I--








-i.

_i_














I
'r i:-"




r-r--



C1;


tCT IS._


j


3Jn-4-- H IS
JJ.flNIVJ~ 1td S1N3VJYJAO~J


U,
t 0


7
-J
cd


8 0
to
U)


0 .
OC)
0 0



f.

1-0

















4 <
I.C






::
I) w
-o



I:

cc







-o
V
S0.i1







The Effects of the Arrangement of a Subsequent Event
Following Each Errorclss Diagnostic Test

It has just been demonstrated that the presentation

and withdrawal of a subsequent event following Programmed

Reader performance reliably altered the rate of that performance.

What, then, is the effect of presenting a subsequent event

following each errorless Diagnostic Test? To answer this

question, a Lindquist Type VI Analysis of Variance was

performed on all Diagnostic Test response rates, correct

and incorrect. The data were analyzed by phases, Before

Phase and During Phase; with a subordinate analysis by

class units. It can be seen in TABLE IX that, although

the main effect for class units and the main effect for

correct and incorrect response rates were significant, the

Before and During Phase main effect was not significant.

This result is clearly evident in the lack of uniformity

in the direction and magnitude of changes presented in

TABLE X.


The Effect of Presenting a Subsequent Event Following
Correct Programmed Reader Responses on Incorrect Programmed
Reader Response Rate

It was previously demonstrated in TABLE V that the

presentation of a subsequent event following the correct

Programmed Reader response accelerated that response. In

that condition, the incorrect Programmed Reader response

was ignored. To investigate the effect of the presented

subsequent on the previously ignored incorrect response rate,

a Wilcoxon Signed Rank Test was performed on the incorrect

Programmed Reader response rates observed in the Before and































o n'.c LA'"
* *



CN




co n m 0n m L n
Ua mamwa
Sc r^D r~ m o

co cm M N
0- N .
CO
1-A








CO





H H 44 4


Ofn


0 0C

0C~0

O
a 0







0
*Ho E
(:> .E-I

4-4 U -L
o Q4-l


400
H 0)
O04 W


O -I
S0 Z
m* t0
>1 r.
0 *d




> .
00 (

0, 0 t
>i4-4 Z
E- 'a)
4M c




S4J (!
<)P


CQU

xS
<<


O
-)w

I
0 -H
0 r
44


rn


mmo

0cr


CON


L
n v
0


U0
LaO

Orfl


n 0 C0 NM

N Cn r













* *0 *0
*1O n
N0~n


crmm


wwww 0
b< ~~


0NM

00
M
Ok


O


0
U

P4







0
O
*Pl

U
O)


0

.3





41

TABLE X

The Direction and Magnitude of Differences Between Before Phase
Correct and Incorrect Programmed Reader Pesponses and During
Phase Correct and Incorrect Programmed Reader Responses




Correct Incorrect
Directiono3MTiagnitude of Direction of Magnitude of
Child Change Change Change Change


Allen .40 .15
Ballard -4.00 1.00
Carver 00 .92
Hathway 3.00 00
Williams + 1.00 .71
Langston + 2.80 + .23
Hines + 5.40 .60
Willford + 2.70 .72
Young + 5.40 .30
Godbolt + .60 .84
Nattiel + 3.00 .60
Ford + 2.40 .45
Brown, M. 1.70 + .11
Wright + 1.00 -.58
Jones -.70 -.34
Jeffcoat -2.70 + .32
Fogarty + 2.00 -.33
Howell 00 + .35
Jackson + 1.50 + .50
Wimms + 2.60 + .80
Thomas + .90 .33
Kickly + 2.20 -.33
Lesene + .90 .52
Webb .60 + .36
Morris, P. 2.40 1.20
Morris, B. + 1.20 -.90
Johnson + 4.20 + .60
Hardwic -1.50 -1.20
Cray + .90 + .30
Hayes + 3.00 + .39
Camps + 7.30 -.09
Kelly + 1.00 -1.50
Railey + 4.30 -.80
Sperring + 3.40 + .15
Taylor + 3.10 -.01
Bruce + 1.20 + .44
Hague + 2.00 -.22
Howell, D. + 1.40 -.57
Beals + .30 + .06
Bishop + 1.80 -.42
Somese 2.30 + .27
Lee + .90 + .54







TABLE X Continued


Correct Incorrect
Direction of Magnitude of Direction of Magnitude of
Child Change Change Change Change


Pattison + 1.30 + .54
Ross 00 -.49
Brown + 4.50 + .63
James 00 00
Stewart + 1.00 + .25
Walker .20 1.79
Berry .80 1.05
Haile 3.50 00
Alexander .40 .33
Burke 00 .50
Brown, K. + .40 .94
Davis 00 3.50
Smith, J. + .20 .15
Bass + .21 00
Pogarty -.30 + .34
Howard 2.60 -3.45
Shaupe 2.00 -.88







During Phases. TABLE XI shows that incorrect Programmed

Reader response rates were reliably decelerated when an

arrangement was made to follow correct Programmed Reader

responses. An examination of column "d" reveals that in

only two cases was an acceleration observed, and these

cases showed the least changes in magnitude. FIGURE VI

displays this deceleration as observed in one typical child's

performance.


The Effect of Withdrawing a Subsequent Event Following
Incorrect Programmed Reader Responses on Correct Programmed
Reader Response Rate

TABLE VI demonstrated that withdrawal of a subsequent

event fo lowing the incorrect Programmed Reader response

decelerated incorrect response rate. In that condition, the

correct Programmed Reader response was ignored. To

investigate the effect of the presented subsequent event on

the previously ignored correct response rate, a tWilcoxon

Signed Rank Test was performed on the correct Programmed

Reader response rates observed in the Before and During

Phases. TABLE XII shows that correct Programmed Reader

responses were not reliably affected by the withdrawal

of a subsequent event following the incorrect Programmed

Reader response. An examination of column "d" in TABLE XII

reveals that, although the direction of changes was

predominately accelerating, the magnitude of changes

observed in the four cases of deceleration was sufficiently
/
large to counteract the directional predomenance of the

data and yield a non-significant result.





















TABLE XI

A Wilcoxon Signed Rank Test for Differences Between Before
Phase Incorrect Programmed Reader Response Rates and
During Phase Incorrect Programmed Reader Response Rates




Before During Rank of Rank with
Child Phase Phase d D Less
Frequent
Sign


Brown .38 .28 -.10 3.5
Stewart .20 .07 -,13 5
Jamnes .12 .16 +.04 2 2
Walker .39 .20 -.19 6
Berry .51 .25 -.26 8
Camps .20 .10 -.10 3.5
Hayes .29 .08 -.21 7
Bishop .18 .19 +.01 1 1
Pattison .06 .06 0

T = 3*

*P ( .05
















14 t


it, ,1.r I Iiii r "


r. ro~i :--
0(a) I I I 1 ri 1 '
I i i I t ;; I'i r
urc31'II Ij ftI 1 'i !_
\ I I j *] I **'*I-I 1. i
Ord7 'I'I I ;t _.i I 11I
Ii:; I 2-
-1 tf. *1I.
Ir a ,_ :; i 1
I1 :' 1
r~- 92 1::'-:'.t .j-- Iiii
0t QI -:.1L: _I-l..~- :-
3_.~;_r I! n ILIli 1 ~ i. ::
1t0
tn I:II~-!.1.17:!:-



kIA -I~. ... ji~ I
CC
H *4
4411 0
4 -)

i..1 -. .
0 0 0)



0 M LI.., L: 7
I z I --- t -I
00



IV lc;:II- ..iif1-;T o "
rd Er U) I.Jr (nII



-J 'I I f
b- 0 I
M -7
ukoi J:..
(U n I4 5! I iII.

LR4 -LIt, 7LI
44 Lit -ii It I ~ -1 f'I-Io


0
01 tin 0
cU 3U
0 Se
o t r-i- r I..~.. iIr : 1'.


P43W1IIlH31 JU 1
EAi ij1:'3~ j















TABLE XII

A Wilcoxon Signed Rank Test for Differences Between
Before Phase Correct Programmed Reader Response Rates
and During Phase Correct Programmed Reader Response Rates




Before During Rank of Rank with
Child Phase Phase d D Less
Frequent
Sign


Willford 3.77 4.32 +.55 10
Ballard 13.41 12.79 -.62 11 11
Williams 9.82 9.87 +.05 1
Brown 14.05 17.76 +3.71 7.5
Howell 15.15 15.50 +.35 6
Kelly 6.22 4.43 -1.79 14 14
Davis 2.26 3.20 +.84 12
Brown 3.98 5.63 +1.65 13
Burke 2.49 2.78 +.29 5
Alexander 2.65 3.15 +.50 9
Haile 2.82 2.91 +.09 3
Morris, B. 9.89 7.50 -2.39 16 16
Morris, P. 5.51 5.66 +.15 4
Jackson 6.86 10.48 +3.62 17
Klickly 4.81 4.44 -.37 7.5 7.5
Lesene 4.97 6.95 +1.98 15
Thomas 5.90 5.98 +.08 2
Wimms 6.83 11.07 +4.24 18

T = 48.5*

*not sIgnif ic'ant







The Effects of Terminating Arrangements on Subsequent
Programmed Reader Performance

Normally it is not practical to maintain a classroom

under continual synthetic consequation. The effects of

presenting subsequent events following Programmed Reader

responses have been previously examined. What is the

effect of removing the arrangement entirely?

In the first condition, where the subsequent event

was presented following the correct Programmed Reader response,

the data collected in the During and After Phases were

treated with the Wilcoxon Signed Rank Test. TABLE XIII

indicates that the already accelerating correct Programmed

Reader r sponse rates continued accelerating. An examination

of column "d" reveals that in no case was a deceleration

observed. This effect is shown for a typical child in

FIGURE VII.

In the second condition, where the subsequent

event was withdrawn following the incorrect Programmed

Reader response, the incorrect response rates for these

children that were collected during the During and After

Phases were treated with the Wilcoxon statistic. No change

in incorrect response rate is observed in TABLE XIV. An

examination of column "d" reveals no consistency in

either the direction or magnitude of the changes. FIGURE

VIII demonstrates this maintenance of response rate.

The third condition, where both of the above

arrangements were in effect, was also treated with by

applying Wilcoxon Signed Rank Test to the performance rates




















TABLE XIII

A Wilcoxon Signed Rank Test for Differences Between During
Phase Correct Programmed Reader Response Rates and After
Phase Correct Programmed Reader Response Rates




During After Rank of Rank with
Child Phase Phase d D Less
Frequent
Sign


Brown 6.83 11.17 +4.34
Stewart 7.97 9.97 +2.02
James 7.16 10.22 +3.06
Walker 8.92 12.10 +3.18
Berry 7.82 8.68 +.86
Camps 10.56 11.87 +1.31
Hayes 11.93 13.65 +1.72
Bishop 3.54 7.22 +3.68
Pattison 4.45 8.23 +3.78

T = 0*


















PPOJECT .a.


0 0
L4



0
4-J 4J
1)



4J)

C) 4-


ro




aC)
44 qC
0 tO
u










1-4 0
p
04.)
(7 C




C)n
rdC)
4Jp
U )
OrC d


E- rC


Ii 11: : I :i:: i- r T.:j


.11
A-i jj I(_ :_.( Li -






I- jj7-
::1': -1~ i*il' 1 '1l
:1_i.



.~. I..:I.i: 2. I7P4B~tI1_






:-7-- 1: ~ : .7- __ '77





7-7

L:.i~i ti r1
iii r:. 1-_.A! i J. ..Jr ; 3 1 i i_l ; 1 _:.~r A
~ ::i- I j .:: I I i









: H-ii I -i I




4- --- L-) *
rr_ I.La .


3:InifNiv ?Jd SPIR NJVOAO


O


O






r3
-8

0
I





-o
.0










o











O O


0




(4


SI,


U.'

'i




U,
'U
(2
(a
Cd


4'1

ii


ci















TABLE XIV

A Wilcoxon Signed Rank Test for Differences Between During
Phase Incorrect Programmed Reader Response Rates and
After Phase Incorrect Programmed Reader Response Rates


During After Rank of Rank with
Child Phase Phase d D Less
Frequent
Sign


Willford .40 .17 -.23 14 14
Ballard .05 .03 -.02 3 3
Williams .08 .18 +.10 10
Brown .24 1.46 +1.22 18
Howall .46 .47 +.01 1.5
Kelly .03 .07 +.04 4
Davis .12 .07 -.05 5 5
Brown .02 .01 -.01 1.5 1.5
Burke .07 .15 +.08 9
Alexander .15 .02 -.13 13 13
Haile .14 .21 +.07 7
Morris, B. .10 .21 +.11 11
Morris, P. .20 .61 +.41 17
Jackson .13 .45 +.32 15.5
Klickly .24 .17 -.07 7 7
Lesene .20 .52 +.32 15.5
Thomas .59 .66 +.07 7
Wimms .11 .23 +.12 12

T = 43.5*

not signicant
























C)




C) 4J

C- 4
4C)


kU)C



4-) -rO
4 Q)
U)C)





51 U)
0 CD
u
4 1,4-









CO- C)
OF:O
o o


sC)tC



'40)
(0
_: (1
.1i ko
r-J 1,
C) Z7)
Q) P



C) U
44


44


I I 'I


"II j I I .I I:. I I --I I
'il 7


1l i i1


.,-i,..,.
'i i 1 VV7

I,~ ri




I Iq.II 2 i il .~ii,

j ~ ~ i ~ ~ F 1 I I; _
0i~- Ifl0 -l--1:
00i Q:3 t

:':: ~ ~ Jfl'? ~;:Iiz f-:1 ld L? .AO '


ti






S-1
a

0-

(y
0- o
0)

1:j


*i fl iJJ

. Ii' I



















:I Pi.

11h~: I

(, i


0



CA~~
w
La
or c
LCJ c
-J


CA




La
-J

L'i

V)
CA
U
U)


0


.1-














't


-0


.0
O Q w



0 to
>
O3 01



(4
C:
a)






for both correct and incorrect Programmed Reader responses

in the During and After Phases. The results of these

analyses are presented in TABLES XV and XVI. Note the

striking similarity of these results to those observed in

TABLES XIII and XIV. In both cases an acceleration was observed

for correct response rates and no effect was observed for

incorrect responses. FIGURE IX demonstrates this simultaneous

acceleration and maintenance.


Summary of Results

The results of this study may be summarized as

follows:

1) the Placement Test is not an accurate index

of Programmed Reader performance

2) correct and incorrect Programmed Reader

response rates are reliably different

3) the presentation of a subsequent event

following the correct Programmed Reader

response reliably accelerates that response

4) the withdrawal of a subsequent event following

the incorrect Programmed Reader response

reliably decelerates that response

5) the presentation of a subsequent event following

each errorless Diagnostic Test does not

reliably effect Diagnostic Test performance

rate

6) the presentation of a subsequent event

following the correct Programmrd Reader







TABLE XV

A Wilcoxon Signed Rank Test for Differences Between During
Phase Correct Programmed Reader Response Rates and After
Phase Correct Programmed Reader Response Rates




During After Rank of Rank with
Child Phase Phase d D Less
Frequent
Sign

Shaupe 3.08 1.92 T T6 -T0 T0-


Smith, J
Bass
Fogarty
Howard
Ford
Godbol-
Hines
Nattiel
Young
Allen
Carver
Hathway
Langston
Jones
Wright
Jeffcoat
Fogarty
Beals
Lee
Ross
Somese
Cray
Hardwic
Johnson
Webb
Railey
Sperring
Taylor
Bruce
Hague
Howell


2.31
5.94
5.23
4.47
3.58
3.36
5.23
4.71
5.87
1.67
6.17
6.37
7.55
7.49
11.10
7.63
12,57
4..19
3.97
6.24
6.86
4.11
3.12
3.39
4.22
2.43
9.20
6.31
7.25
15.20
5.38


3.46
5.74
4.09
6.65
.74
4 61.
6.1.5
7.37
10.10
7.32
7.47
8.57
10.82
6.73
8.95
2.52
15.76
5.97
5.30
7.45
8.64
4.35
5.92
2.84
5.84
2.76
9.21
4.35
9.99
11. 34
7.83


+1.15
-.20
-1.14
+2.23
-2.84
+1.25
+.92
+2.66
+4.32
+5.65
+1.30
+1. 84
+3.27
-.76
-2.15
-5.11
+3.19
41.78
+1.33
+1.21
+1.78
+.24
+2,80
-.55
+1.26
+.33
+.01
-1.46
+2.74
+3.86
+2.45


7
2
8
21
26
12
7
23
30
32
14
19
28
6
20
31
27
17.50
15
11
17.5
3
25
5
13
4
1
16
24
29
22


T = 124*


-iTO 2








TABLE XVI

A Wilcoxon Signed Rank Test for Differences Between During
Phase Incorrect Programmed Reader Response Rates and
After Phase Incorrect Programmed Reader Response Rates




During After Rank of Rank with
Child Phase Phase d D Less
Frequent
Sign

Shaupe .11 .07 -. 2 6.
Smith, J. .04 .10 +.06 16.0
Bass .11 .21 +.10 20.5
Fogarty .07 .04 -.03 10.5 10.5
Howard .28 .06 -.22 28 28.0
Ford .42 .40 -.02 6.5 6.5
Godbolt .33 .15 -.18 26 26.0
Hines .13 .19 +.06 16
Nattiel .21 .01 -.20 27 27.0
Young .10 .02 -.08 19 19.0
Allen .11 .06 -.05 13.5 13.5
Carver .08 .24 +.16 22.50
Hathway .21 .05 -.16 22.5 22.5
Langston .12 .05 -.07 18.0 18.0
Jones .60 1.45 +.85 32
Wright .16 .33 +.17 24.5
Jeffcoat .15 .17 +.02 6.50
Fogarty .15 .14 -.01 2.0 2.0
Beals .21 .24 +.03 10.5
Lee .13 .14 +.01 2.0
Ross .16 .17 +.01 2.0
Somese .08 .13 +.05 13.5
Cray .18 .50 +.32 31
Hardwic .14 .24 +.10 20.5
Johnson .21 .27 +.06 5.0
Webb .20 .50 +.30 29.0
Railey .04 .02 -.02 6.5 6.5
Sperring .14 .63 +.49 31
Taylor .09 .11 +.02 6.5
Brvce .01 .03 +.02 6.5
Hague .18 .01 -.17 24.5 24.5
Howell .04 0 -.04 12 12.0

T = 216.5*

*not significant





















to
4J
Q)


4.
0

C1) 0)*-
Q)V)-4J


0 C


SC)0


4- ) S-)

0-) C)
'o
4 1

Orn
Ic~









o 0 ro
0) 4
r, P- L






dj 0
4J 4J It
r1 0

'H 4 0
[ A 0 Q



'4.4 S:
o H 0
4 J
rsUF




"41-
t4.. -0
EC)





Q) $4
P0
u


* ho. I;







0

-cl


I fiOJicI




: iii!., 'mu






1 I .. I i
I''':-; :::';~11111i- -L~t_: II-i/ f


jIL:I j:: II.iil
F:_i_!ii ~' i .1-
: I 1r.;: i
:1 i. 1- I.:: .i ii t

It .... i t
ii ~ ,.-.,-






J~ii I : I i-1 I~: tI i ~ii'ii --1r
1U I 1---; 11 ~I
: ~i' I! '~ :I]

LI1~1 I..


I ---I j'.;i ::j .
j t1~' L%2 4I--{1 I

iii .LflNIV1 ~JJ S lJJ1AOi


- c, ,


-o



0i) (-UJ
J.



C.)






-0 Ii

sij i


A-







response reliably decelerates the ignored

incorrect response rate

7) the withdrawal of a subsequent event following

the incorrect Programmed Reader response

does not reliably affect the ignored correct

response rate

8) the termination of an arrangement made

following correct Programmed Reader responses

is associated with a subsequent acceleration

of correct Programmed Reader response rate

9) the removal of an arrangement made following

incorrect Programmed Reader responses is

associated with a maintence of incorrect

Programmed Reader response rate.













Chapter III


DISCUSSION


The issues with which this dissertation is

concerned may be pooled into three major areas:

1) the demonstrated utility of free operant

techniques in the measurement and analysis

of programed curricular material

2) the effects of reinforcement on programed

instruction response rate

3) the concurrent nature of correct and incorrect

programed instruction performance.


The Utili ty of Free Operant TechnicTues

As was pointed out in the Review of Related

Research concerning the evaluation of programmed instruction

performance, there is a remarkable lack of precision and

standardization in the analysis and evaluation of programmed

instruction material. Indeed, in most cases, performance

on the programed material itself is rarely used as the

dependent variab3> in the reported studies.

The important difference in using free operant

techniques in the analysis of programmed instruction

performance is primarily in the use of direct, continuous

recording of all pupil performance.






Markle (1967), Mager (1961) and Trow (1963)

emphasize the importance of using measurable behavioral

objectives in the evaluation of any performance. In the

usual situation, however, the response required in the

curricular material and the response required in the evaluation

are not the same response. It is obvious that the researcher

cannot evaluate directly one response by measuring another.

He must be able to record directly the responses being

observed if he is to make meaningful statements about those

responses. In this study, all responses in the Programmed

Reader and in the Diagnostic Test were continuously and

directly recorded. This recording procedure allows statements

to be made regarding the compatibility of those two portions

of the Sullivan Reading Program.

The direct continuous recording also allows a

precise statement concerning the relationship between

correct and incorrect performance rates. Research reports

(Holland, 1965) traditionally report only percent correct

and percent incorrect, forcing the two measures to be

complementary and pooling them over time. This results in

a tremendous loss of sensitivity to behavior change (Lindsley,

1967; Holzschah and Dobbs, 1966; Koenig, 1967). The

continuous, direct, daily performance rates give both the

teacher and the student immediate knowledge of performance

in all phases of the programed instruction performance.

In this study, the direct continuous recording

of programed instruction performance rate demonstrated

that:







1) performance rates on the Diagnostic Test

were reliably different from performance

rates on the Programmed Reader

2) correct Programmed Reader response rates

were reliably different from incorrect

Programmed Reader response rates.

In addition, a traditional pretest-posttest

evaluation of the Placement Test indicated that Placement

Test performance was not a reliable index of where the

child performs in the Programmed Reader.


The Effects of Reinforcement on Programmed Reader Performance

The question of reinforcement in programmed

instruction has traditionally been discussed from a

theoretical point of view. Bypassing the theory, which is

not relevant for this discussion, and examining the experimental

methodology found in relevant reports in the literature,

it becomes again apparent that only rarely is the dependent

variable in such studies pupil performance (Johnson, 1969).

Generally, a criterion is administered, either one made

especiallyfor the purpose, or one of the standardized

tests commercially available. In no case did the writer

observe a direct continuous measure of pupil performance

rate in the within-subject design type of experiment. The

between-subject design further reduces the likelihood of

those reported studies finding change, should there be one,

thus increasing the probability of Beta error. The within-

subject design has no such sensitivity drawbacks. Studies





60
utilizing thewithin-subject design with free operant recording

techniques have been able to show not only that a change

occurred with respect to some experimental variable (i.e.,

the presentation of a subsequent event following a response),

but they were able to show the precise daily fluctuations

of response rate associated with the experimental conditions

in effect (Johnson, 1967a, 1967b; Koenig, 1967; IIaughton,

1967).

Using the above type of design, this study

observed:

1) the presentation of an experimentally '

selected subsequent event following the

correct Programmed Reader response alone

accelerated correct Programmed Reader

response rate

2) the withdrawal of a subsequent event following

each incorrect Programmed Reader response

alone decelerated incorrect Programmed Reader

response rate

3) simultaneous presentation and withdrawal

conditions both accelerated correct Programmed

Reader response rate and decelerated incorrect

Programmed Reader response rate

4) the presentation of a subsequent event following

the correct Programmed Reader response

decelerated incorrect Programmed Reader

response rate

5) the withdrawal of a subsequent event following







the incorrect Programmed Reader response

did not reliably affect correct Programmed

Reader response rate.

This study also examined the effect on subsequent

pupil performance of terminating the reinforcement condition.

The results are most interesting:

1) the removal of an acceleration consequence

following correct Programmed Reader responses

accelerated subsequent correct Programmed

Reader response rate

2) the removal of a deceleration consequence

following the incorrect Programmed Reader

response did not reliably affect subsequent

incorrect Programmed Reader response rate.

This result is remarkably similar to an effect

experimentallyknown as "behavioral contrast" (Reynolds,

1961). There are, however, certain dissimilarities present.

In behavioral contrast, the effect is noted following the

acquisition of a discrimination. In the present situation,

no experimental discrimination was trained. However,

response differentiation between correct and incorrect

Programmed Reader responses was established. Whether, in

the present situation, this response differentiation provides

the conditions for the necessary stimulus discrimination

to occur is not clear. Proper schedule manipulations to

replicate the phenomena within each subject were not

performed. It can only be noted that this is a behavioral-

contrast-like phenomena and deserves a thorough systematic

investigation.





62

Such an effect has been noted by other researchers.

Lindsley (1969), Koenig (1969), and O'Brien, Azrin, and Henson

(1969) report observing the acceleration. Lindsley calls the

phenomena "after effect," which, on the basis of present

experimental evidence is probably a more accurate descriptive

term than is "behavioral contrast."

For the teacher, this means that a subsequent

event may be arranged for use as a tool to investigate

a rate change, and need not necessarily be used to maintain

that change. In fact, it appears that any particular

arrangement may serve to limit operant strengthening if

left in effect too long. More research will be needed to

determine the guidelines for terminating accelerating

arrangements at the optimal tiate for maximal maintenance

of the movement.


The Concurrent Nature of Correct and Incorrect Responses

Nowhere in the literature did this researcher

encounter an experimental discussion of the behavioral

relationship between correct and incorrect response rate

in programed instruction performance. Glaser (1965) and

Holland (1965) both provide extended disucssions on error

responding and error "rate," in terms which are actually

not rates but absolute numbers or percent. Market (1967),

writing on the empirical aspects of program evaluation,

presents a similar analysis.

This study reports not only the effects of

arranged subsequent events for correct Programmed Reader

responses and for incorrect Programmed Reader responses,





63

but it reports the effects of an arranged subsequent event for

correct Programmed Reader responses on incorrect Programmed

Reader response rate and the effects of an arranged subsequent

event for incorrect Programmed Reader responses on correct

Programmed Reader response rate.

The changes observed indicate that correct

Programmed Reader responses and incorrect Programmed

Reader responses function as concurrent operants (Catania,

1966). Ferster and Skinner (1957) define concurrent operants

as "two or more responses, of different topography at least

with respect to locus, capable of being executed with

little mutual interference at the same time or in rapid

alternation, under the control of separate programming

devices." Educationally, this means that correct and

incorrect responses are not complementary movements with

respect to their contingencies. This further means that

measures forcing complementarity (e.g., percent and

absolute number) are not only insensitive to behavior

change, but inappropriate for precise and accurate statements

concerning correct and incorrect responses.

The great majority of research available on the

nature of concurrent operants has been done with non-human

vertebrates, chiefly the monkey, pigeon, and rat. One

central experimental problem in studying the exact nature

of concurrent schedules of reinforcement is the establishment

of the independence of the concurrent operants in question,

whether they are compatible or incompatible. Procedures







to insure this independence are introduced primarily to

avoid concurrent superstition effects (Catania, 1966,

Ferster and Skinner, 1957). Correct and incorrect concurrent

responses, even though they are technically incompatable,

viz, they can't occur at the same place at the same time,

are not independent, as shown by the data presented in

Tables XI and XII. Specifically, the consequation of the

correct response results in a simultaneous deceleration of

incorrect response rate, but the deceleration of an incorrect

response has no noticeable effect on correct response

rate.

This study employed daily plotting of correct and

incorrect response rates. This allowed only a coarse

grain analysis of the concurrent operants and therefore no

local interactions (variations in performance rate as a

function of switching from one response to the other)

could be observed. The only statement that can be made

concerning correct and incorrect responses in this study

is that they appear to be concurrent operants. Recordings

of individual responses will be necessary to uncover further

information concerning the exact behavior of the operants

with respect to each other for various schedules of

reinforcement.













Chapter IV


SUMMARY


There are apparently no published reports of the

experimental analysis or evaluation of curricular materials

in which pupil performance rate, directly and continuously

recorded, was used as a dependent variable. The present

study applied free operant methods to the analysis of the

performance of 59 reading disabled children cn the Programmed

Reader of the Sullivan Reading Program. A within-subject

design allowed an experimental analysis of each individual

child's performance. In addition, parametric and non-

parametric statistical analyses were performed on the

grouped data.

The extreme sensitivity of the direct and

continuous recording of pupil performance rate as a dependent

variable in the analysis of programed materials was demonstrated.

Further, the efficiency of various contingency arrangements

was examined. In all conditions, the presentation and/or

withdrawal of a subsequent event produced an observable

change in performance rate.

The results of Lhe experimental procedures used

in the study were discussed in the light of their strong

similarity to results observed with non-human subjects.

In particular, the concurrent nature of correct and incorrect

65





66

responses was discussed, and the need for further research

to uncover local interactions for various contingency

arrangements was pointed out. The study also noted a

post-reinforcement condition (after effect) similar to

the behavioral contrast seen experimentally with animals.










































APPENDICES



























APPENDIX A

A Sample of the Programmed Reader
from Series I of the
Sullivan Reading Program























The red car is pass___ the big car.
The red car is f ster than the big car.


Jack and Jill are dressed up

to visit Sam and Ann. Jill has on

a red wig and black glasses.


red
Jack has on a mask with <
Black


shirt
Jill's irt
skirt


whiskers.


has stars on it.


Jack and his sister have a present
to give to Sam and his sister.


-'I


This is the present.

Is it a banana?


93


passing


faster


black


skirt


~~Llr-- U r


*-,-,J.K.^t'AW 4.1J..J".TK,--.l.-* "- --'"^J NEI?'-aMNEMEM



























Jack and Jill are with Sam and Ann. When

Jack and Jill rang the bell, Ann let them in.


Which girl has on glasses?


Jill
Ann


yes
Did Ann get the present? yes
no




Ann thanked Jack and Jill. Then Sam

passed a dish of mints. Jack had seven

mints, Jil had six, and Sam had ten.


S Sam's
Jill is 's
Jack's


sister.


yes
Are Jack and Jill visiting Sam and Ann? yes
no


94


Jill



yes


Jack's



yes


































letter


hand


I.i
I vings

\''t79


I am thinking of a thing that is glass

and has jam in it.


car.
It is a ar.
jar.


I am thinking of a thing that has

a stamp and printing on it.


< adder.
It is a letter.
letter.


I am thinking of the part of my arm

that has fingers on it.


I am thinking of my and.


The things I am thinking of are part

of a bird. A bird can flap them.

It has to flap them to fly.


I am thinking of ings.


95


'r"~~~5717;--~;);*Tlr?~P;
rX*-.


f~i\ ~lmri-i'if-i..i^'i- _.inii~'iL ^^-1" -*7-''? I'^. ^^ ^'T^ ^^ ^^^^.^^^^^^^y~^^ ^^ S.^^L..... .....S .y r *.*1* '? '^ '* ^'


~.'-~;~,~c~,-;~~ ,*IZ~w~t~"T"~~~9~~H"m~'T--rr\~~.T,~


I


.03
[oo



























APPENDIX B

A Sample Diagnostic Test from
the Programmed Reader














TEST 6




1


a i ac r a st r an r

















gr ss. The grass left a m rk on his pants.


A man passed by and fed p_ rt of a cracker
to the birds.


Sam let the birds sit on his rm.

Are Sam's arms par of him? yes
Are Sam's arms part of him?



























APPENDIX C

A Sample of the Placement Test
of the
Sullivan Reading Program














Sample Page


yes
Am I an ant?
no


hat.
I am a <
cat.


Am I fat? yes
no



I am a fat .at.


man.
This is a pan.
can.


;t is a tin cn.


_ __


- ~_ ____,. ,__ __ __











TEST 1


This is Sam. This is Ann.
Sam and Ann can sing.


Is Ann singing? yes
no

yes
Is Sam singing?
no


Is Sam sitting? yes
no.


hot
Sam has a < t
mat


on.


hat.
Sam has a fan in his hand
That is a pn in Ann's hnd.

That is a p .n in Ann's h_nd.



















1



hat

mat
cat


an _nt





a fsh


pan

pin

pig


dish
ship

maps


6


a an





a h nd


TEST 1


I-


--I s~ --


_ -- -


~Lc~---~-~-~ll --- --------------~-CIII~













TEST 2
-----e5- r


bat
bag
pig


kitchen
kitten
mitten


Sa -mng




a drss


trip
rip
drip


I-ps






;Pa-- -icken


- -- --I----


- --


























APPENDIX D

One Complete Precision Teaching
Project for one Response



















PROJrCT Iw si .. a. --
NAMES _EpnnypncCkeL r.dlnfKor Cel ge Jolhnnon s3nP Inc I A-. sr: b PROCEDURE REPORT
t RAIt R AD.,4 l S*NieflB 4aOVI4 --
USE NO 2 tEAD_.E-Cl. 335' %:, .. --T





...--o i . .. ... .. . -
r ne uses .. v *r(, z [l







DAILY RECORDING PLAN, ::::: ^ I :::By-: r5tCrto TB ,..SE g p.. ::A e-"
TO 7I" ^t n' b! ,iif .. t* ; "Wl ::::w &.o A T."W4 eCE B.. r. r ::::r r 4'

TOCCiT. fly- w4" rotf ,0- -*,r :::::tO CtcaHWoCr tt n TO CAE" nFbA1 ra vae" .' E," ri v r::,::r Br 7msrrr n SEI FW P.40p P fl B F S .::- :sf
0 .. 1.** L .. ..> .A. '. I uu* Is K 3- 11. 11>. *I* V- 6
0 1 ..* .c.- *, .. *** s **- Srr r *^ I | *rm as I *S* *ri *M : 2


::&: :3:+
r:S: ::*r


NO o* )4NmTj :lt*aO:f:O (:I:M ):,::
Nog ::a: :&r: ::r.: ::t : ::2:
*ls ::At: ::: ::r.: ::&r t:r:
ons. r:03 :rB rtTP: r8 5535


Ut wur L Bl Vi U vit"t mrfl t. r oi l ::rr: ::r: ::Z: ::.3 ZB knd ::a: S-se 9 t::r:
:t0: ::t :: ::t: :: : ::4: te ::s: :. t : *9:*.
,0.. --. *.)- :: :-.: o : :- <- *r


I T : aM ::
(IT r : : ::1:: :- : :


CF7.2 ::0: :'t:: :'t:


( Irr uO A. .t M1VYEnm A rACHo 1Ar(Muri.


t::: ::t: *as ::t: :ti: ::.: :tr: v::


i to iE NokiT No. E MoYLS's. EACH OAT (M.4d).
: : :- 1 :::F: :;} ; :.2: :" : :?+: :-s: :

: ::I:: ::: : : : :t: o*, ::;= : ^r1 8f ^l


".- -. L --.-r .- ..-.* ;. ": -, .l --> S r ,.". .s'
PROC,9AM ::I== 5= 5, At" muCTED Sy. 41 f" I '
CO-Ollot t Ow toG l 1'f a ::: ToIfo:::: : 0 Gr1 AuC rt V. AR -0 AA-' WMn g::::

PPoUmRaM b# E nO 1 (VEIt PEtw V7- jW 4W 4d" WcHIC PKNs n (i Ma o1* N
:-0: :::I:; : : ::: ::*: 1... *-V =-- IF s* I' O = tV' u
:-0 ::I:: ::I: ::: : :::: r! ::r: ::.: |:: ::On : :r :: ::: ,., : : : : Wf ::

IC START Cta.S.PARAT E VENT BY' Ot:OPar P t ? A. ::TO ::1 .TO : : :- '::O 0' : ': !T

S AN ECCEENT EVEN'Tf ::::: "* ::: I s5 crTt0 .' e r r :::: .. ..
.. .SULLIVA EADING PROGRhAM ::: : .RA..E Cpi:: ;r.ii :' :i:::c":

"'- ... .~ -.... ". "'" "- I'.... i '--' '. nrt < '^"7
TO PRESI-7- V-4 W9 4 O-R- CW VA L0 Orro 001 0MF .




::o: Z::: r:: :: 3:: : 4 5 : S *I : rg:t ::; 0: '*:rs :4:F :455: *' *' :*'t
.. .. ..,- .. ..... .a- - : r:: ,- -, .




AR GEMENT:: :: o 4snE CTtE U :r i i i ..... -'-


C'rUSWU:roNTE 'ce :::: A(t VSk S :::: |I otfC r 1kd Oj iP ,e ;M r, CJ 0*.4 RAT* ::AP

::. t w h 4 a.- ::I; t.-7t ::F. -:9: W 0: tw: 6, S :;V t. : -r :t : :-v
:g: ::I:: ::r: ::I: : ::2:: ::: :or :V: :3:?o. ;: :trw: ? t42 : 51St : r: iF: :SB
: :-: ::s:: ::: ::a: :.4: t 5 : **.* 5* 4 :-wt r t SB r


Z PrOTECct -TO wm p.. ;Soi* 4-10o..t *S S ii' = IM,.s .I) r = tr L: *tl 4: 1.* : s -5* T r:7
-1
Af* ::::* ::: a ftro : crrn C rre w a s m outr'rr n r a: re 7 0 r:> : :5 ; *,* r





S" : :: : tI: :"ING 91 t: : :: 1:: : ::: :::: 3
.K S ..SE.ucNT EVENT- :-.. f,, -.> ., -| n- CT V .-. i _-_* p -" l

TO f-E0 T- NA *v w W0 V ij.9 V~m TO CA FROM, 5T!ff rB



4:. ::1:. :qt: :X. :: t. t;' :;::4 : t5 2:)5 ?rtt t ;-:-B< sum iti toes : I, lie
=^=c = : := = ,0 *

,^ _^_2. -r_.-a -.r^_^^^ ^^I~I_~r~lrl-l-n- ^ ^.. *~i .*^-~'-^- ^-


TOOCa(T:fO A-.

:0:t ::I:: ::r:

s- .;S t- O


..r


-- ,-,-, ..~..


I






81












PROJECT TOM .9, ..*,.,. *. I. l
NAMES: Praypackor Edingor G oArI Johnson SRP in c t h.-O. PROCEDURE REPORT
TIRAINER ADYIS( M N aC PRiOTEG MOYEWEHI TARGET Prt-
39040 7999..ct C*I
USE NO 2 LEIPL..!.CIL R~ ~33 .n...t Ce,

Jr) ,. U) rIr~Frr1


MULYNI C)CLL .. EN z0 550 A" snurtco m.- 6 --I pl p ..::: 0,0 t:


rL4C 1 V0 4 4:0. M, MAC[E By'~..9 .L % wv I"" ItU~ vo ~4
... .. .. ... ..


-1- L-RDIN. F LAN: 15 2,0 3Rru jn SELECTEOVI '.,M 'VDw~f IVC .:: ~L
TO T:9'( BY -04.v W.L. ,oO4* po1111 ,1OJ w,iil nr EA::(04 T:411mI 1(4 FEE 44 ".a'I 94145i 44'4. 4.4-*035 :::~' :40:

TO CC.Y4T By foo,' ~ 5.'40 o ,'.p 4 404B ."E [A04CMuTIOSTfLSY Ww41At ,I4nGA .4MAO gotr :::.~
TO GMRhBY.9$J 3l 8 -\r 'Id P.0 0 A-I'~ O :. V04 alAWYG Sf BY'43 VIA" gn.0" ,n.A0. ,nwaC pol ff :.4.
...r4OY :S 7-,, 4444 -:1. '7 IU '44L0n .45.. ..Jq




4.T S.


:-0 7 ::1-: :t :::, :54 :: : ::: Sot"::Z: 5 .: o~c :35.: 4-*: :-v_ :z5:


'* :: I.;. ::3: -.q: ,'4 I_5: :.G: :.r.: i_ E
0 .. I:. :-f 3 4.N (


P9OGRiAM' El~ ~ MCSBY~4' '*~,J0S r

COWY P NE' 61 NON .1 EV N PEA TOM* o~ 494 po1.S FDA~ AO ,W 0H w.L NO- r


:0: 5: .s :r o : .a: :..: 83.;. 0 T, :5 : :A;: .: j ,- .. .~
2: 4-: 4





c 0 TO TA.44 [EACH SEPARATE EVET.M OY, N&A C N:pRi: ,vag,0,'%4 AMO .TO:3TOP9 ar 4: ^ OR 115.::I4*TS

f ANTECEDENT EVENT: .: 4([(B
*00.m~r~--~~ ,,, ,l~'~~-~iL~~~~l~E~~-~
41 JPORI. !o: ::: _____ I .ncr r r iC LL -n ~ 1~ dY






to P PREIEiNT, 1'f .ON 10.4W ,1" CA C1! TO OFT [5044 ps,' ':v' f,-of it -
:7: J5J: gn:
StFcdo aI-AipOC*jZ ~ i~vt
u -Ir ir t Erd -m j'TIE cEAH NTCLU .7EWN-


3 -4 4.n444S -6


:5:::: 1 :401. :.::-:: .1:: j
. 17: :0. ~ :r .3 : :A- -n. : : :47:


0O. --I1


.B .9-


:i:4:


CLOXAON: 1- :a5 BO TH V 4T! T. ::: :::: tRW4l[5 4I: y.44. L : i 4 '-' 7 ;







PNOTCLTO WAM NO. 1D- ,0-0 0. :4:::.: 3 :4: jn0 :5: :i:: : ::9
: ::BOEFR: G[ry:4G EAIH EA?;NtD StBSEOUENT CVEIMT9 0 fit

SSUBSEQUENT EVENT'" 151. t9N 5R0 4TH SCLEC-0 Ell
IT(- ..C4 ..T ..... .....arr'...z..........
TO PFE 4t tW- OeR*4 o %YTSORA~wT- w"" 04pst '10V T C-6A r~,ww
.- o ,"', t ,,a- M Imr t it

0* 1* 2 3 .4 *J 13 6 .7 1 1 -0 1 5 4
:ID-: :-f:: ::2: ;-S: ::4: f ;:- am 0: t :. : : : S: :4: S. : : :is` :`..: : t :
o 1 a 4 *:s: 1 S S0 5 4 n. 5 f. .j


I-


---





















....cr79 .... --- ... .
""'"J C "'t< | i 7 M I '' -- 8 4 "
Na'YLS: PCIrnvpr er Edincger Gtiger r Johnson SRP inc :4: d:u RAC

u Lt LiAr 11 ('41. Uiiyttrt ( r f a L M yw r* T*** M l,
Pa.-I 14 y t v y Mtlto 1 r fr i Tme COift' AC! 1 *'* T. I 114 to M k 4 KV s T T .t .I ..
i'Y-. *_ UL i n4 V TVyI 3 -) rw4tr TT vryt1 or v' Uy t. l r* *iv*. Pal. i. s ylrm to .,,
lt.- In uIa (o' P T *.\v y ta a riuvn Iv I? *'r. *' Ir* M tt acs-t i OajL *vn (.?j P. 33,,, IPtn a Cyp
"..t r l tt 1* Ly ,ta *F. 4' t*t, i" l Cr e *y t. f rI .T t'*l**rv n.* Vn t r 1 t o yAY 1' .-<. v *** 44105

C j0rto /dO .. I. .ji.. 4. te 0 .. A. i .. ... ...A


| 5 1 o 4 r. *r s | .* .. -
S0 o L- JL. .; -- 0 M.. ..0 _. _


r o a 4 3 a4 a
___ __ __-_ _a -- ;,.z .:5-* : :
S .3 & 6 7 3 I e 5 4:: i : ,. .: 9 9
0 o 7. 3: 4 3 .6 7. a 9 e 3 4 : .- :: :

T ^ 0 o o -.. p .. ..0i :i r o 0 L dL. ,471 .-
S9 6 : 8
0 :A1* : 32 4 S 7


o jti~ f .o. .co-.. 0 t._ .. c-l ,. ; .n.A "_

0 0 .* oO '' r E P e0 a ;: A z;. A
o :* : 2. :3 4 4. :7"- :8 1 : 4: : :: .8 : 9

l 5 .U o 0 .- ... j o l






.- ...... ....... ,,, -- ...a .. .4 O GL : ---7- .5
.4. 4 : 7 a. z [ 4 5: fi ?




C0 Ca- 1- ow o P- n
.I. z 2 4 I..: : 8 4-
o ..I > :t i 4 s a .i r I 2 3 4 : : i. : a:: s -


1 9 i. a 4* 5 : i .


| e $ ".1.0 0 DO. LW t ; -1 .0 o: XL a
j ;- : a 2 :: ': ::.7. : 9 D -< *
v, L L 0 ..04. U.1. -- | o .




~ ~ 7l4 JjI 4- "8 :: n -..00- A-
S4. 4 2. 3
I.;, art 0 o ... .. i:' .* j |.r_ e : *<** ____ - **' -4





I I .. ) 4 7 i; |a 4 a .

,. ..F -



O 1 24 at. 4
Lot P--, o 03n.".- -^- -- .0- -- -.- -o --f -









? a .3. 1 *
7^ rL ^ ^^ ^ ur c*'dCL. IIIt
g ^ ? r I ;S- *C L. .I ** ^L _;;


I

















r O JE C T s o r '" .. .".. -
rojES i Pennypn~or Edlngor Golger Johnson ORPP inc :A:u: 'e RTE F.~tf (t .,) -
Pp4** ",- c 'swere IP ort lac-U- Fotr m i Wa n u
VIE No. 2 LEAD POd T,,


S r ;a!. 'r u2 t: K t -.- n i- O :.rk-t h r '. :t a o ,4 .- c *> -tr Ce,. e -. : T tt-S ......- I s; a% C. i :
el*yi:s-j 45:84 "?P *L. W iI 'nu .-,t*' 4 4 t.LI Lw C..: !* '1 WA n. 4nv. SVg 6151 - v trt LU r I*9 9- uS' -7 a ut e *e OU r I oC t. lll Wj r, 3'1 P. t- r "
rrt t:4' 1az-1 Ld, s- + S LT I:t:?,o ti la t-r-t. T 5 ir r- v S ui T: e C I

1 '. 2 C.5 = :' : :1: ::p .-; ,: :: : :: : : ,,: :: a:pA ":: 1 ::4 : ,:$:: ::, ." :. .: :'|':., : : :, c',' -
~---r--O f~L~i~ iLiW-IC~ ryl C_~ I.*YIY-l1[LM


::.: :: :: : ::. : :


::5 :83:-


:: L: ::.


S.: :: :: .: -:.: .::.: .....:







SAT
or O c ::: -r .aC -:


^^^t :T ::: : :: ::t: ::
::' .t : ": -'_:_ .:e


0 ::2 3-- :


, :. .i:-. :'-. :: .

;:r: *': ::r: : r: : r


j~j~4 :.~:


:3: ::y:


:-f ` ::: :7:: ::: : 4:


".7:-: .6* t--. .i: .':... .... ... -






It ::: :-. :4: ::-:: :y"
r^ :|: : :e:: : :: :<:: : ;: : :: : :: : :: : --


,.1 jtr : ): ,jhj


41! :: :: : : 5: 4 !7:: -5:: "
..r W^r-f T -. ::c1J: j 7


w : KI: ja: : r o I r
.. ::...: ,t; :,'' m: :: 4 7. :: "::. ..
i o *: 5: :: : **:-.: ***.: ,:::: af- *. ....: -:to.!;.: :" -: :0 v: :- : .y : ::" ::: *:__. c j-


S :: ::,: :: : :: : : : : : :: : W: : : :::::: :3:" : :; : :: ::: :,:: :,' : -
,,y ::r ***- *-' : i -. :- :: r -:. 1-- .-: : :: -'-a: ra.. ": : : ** j~





: ;: : :* : ::-: ::4: :;: :: :.' ; : :<.: : i : :$-: i3:: :<: :e: 1: :';: "; *-1 .^
:tr r to e- : 4: RI r? y 4


), *: :: ::: -: : ::. : : ,.' : : ::- _:- j _. :-| : : :;: ::; *-.:"- :f: :'-::, .y

r I *,i* --
:::: ::r : ;4w :: i; ; .
:W: f.. : y :,F: : W: : :j :4: ::: :0: 9:: ".- W; KV -M):


^ : :;T: ::r: ::: :*: :4 :: -: :7T: ::: a r : .: :r :
o: :: r: =-r: =: : : *: :s : : :.: :o: : 3: : :
t fIr |X :g ,r ai. -1' 1 i -_& Lr |. .l3 ^=
.= :::: : : t: : : :** = : a :, n : : a
0- ::: : : : : :: :.. '* ~~


:; ;: -.r : .. ; = : .... :=- a J

:* o: : : :1: -a 2 '*: I .I.; a:: :' S :": :r-" :1a' :7: Ir


1~nurvrrr=~*-as;*,~~irr


~L~L]'"U
::(::


.- :::i: ::: ::3: ::A:


--


I+ --


















IT"OXICT AzipHIM U4I
ver- --ncgo -Johnson RP Ine C.S1k L PbtJ -
MMSIRzw M7{M.N w4! C4UT~


A" wvO -w# tP :iL PO. :if IF' ya0 i*,114 I wn
?Ia41 v z ** *m F M'42p04 04. : 14 W.4rp~f


(107M .A :11 0 4 F:P :d F SAME TIAN OW-: p.
.*l w8 va* :o 11 : F :wm Ifl 114444 c C.: nwt 1 pf


Ss41 :4* %v 0 E11i r wm ilit.:1 c ,.n
ve RAME O LOWI *M 9A4 .Z=0. AZ ~ .
4 :3: :4: :4: :: :7t: :0: :9:
... A:: rt: 3 1: I~ ~ i --~I

:0: I-P. .1: .1 :4;.: :3,: -:t .'et:-.4 :

.e.w &C .4LI & :0: -... : :-i 414

o I 1 S 4 6 .0. .7 S
N'JfY1~ ,4* *S urn ( 0 yO *11 TPM
ri V .4S .I ,.d .-n4 Owl Fl" 4 3o 4TH TRY
jftO. jag I" :l*- 2*9 -.:4' F MOM "1W' C0':: b C441

rrar4 dCA -gL WO sl* 114 IF WOW44411 04Ew-: P0.9 E.
PCoVo,'o of 5 &A : :4T* IF MOM744 TIV.4 CR:: *4W1




urc: 4- 11 :Y :4~ run~I~~rarlrd
57444. :lld) :34 :&V !; F .fl T" "' I r VO




11P -0 a"E F mux9 TI-M 4444*1 vwsT dtlm70 42 r.,
M44Ae4T AREC CAS f. 41 %M 0 *%O : F 1 :
::::: ::::. :tC .: A:40 :*t









VAw :491.~n :4h. 3*X :4": ~-; 211*04 -44P,~C pmfta
:. :: : :0: :r-:: :..: :A:
/S. :J, A;: -; .










7K NI m .7r mw md 4 Wro LS ?W 4TH Wt.
Kw .r : -0 4 :41 :W :114 : :41" F MOM TWO :77 pet:0










Y4cPT.'4~ :I9" my *411 F 1421E fl4101 x MAE 1,% W WS WI
(DO-'6 :!~ 4 .4 :*I'" IAE1'* f: W 4r j4
:101; :1". t 4,, a ,~! If KOW T c~~- a~m OW, vl


IICE OFALAST :.*M :-0- Da- ILO.- IPr~lCI


:-0. ::1: :-Z: ;-3: :A :-

I2ACQ UE4.7.4 4:X: UTW


0 1 1.* 3*91
:I VEWP.)El -4 w"A 70 KT-U .,ST Z9 3 PO 114 TR




to OT :.v wo :41.4 r M= ",I wwlw4,cvwl S"O.~'

~rur,,r :44ra ,rT riI IF W-M 'w" OW,:mr"


. I- wll js :O 0 w4- :011 F NIL p11M c :: cgrd I"-


... ... .I71 U
.11' 44 C fl*1


.. .


tn~IM *iua O :r:-: L: :V:: :41, -&:: :6:: w
P4Y174)40E ".: :0:: : &:: :31: :4:- 5:: :A:: W9'
544400 -0:-. :t-: .4t:: 4:: ;4:z 4- .4
AOC47t1.14444 T-:4 :::- 14L: ORTO: 0090 001.: M. C po
W,4fbO" WJ :X-: r~t:-;M 4
9 tM.. :*97 :A:: :3:: P::.


Snt(IYj, -.A:: :4:: 4::,



P LV*4.4444 ::: ,: ... 4.. L (21. J7L -7:..:: 0
4 1 42 6- 7 i. 9
-7L-Cl ll2il


EFTMTIVE WED1 A, k:: 5:: :4:: :5:: :.6 :: 0


LM114 WEIW 0:V::.4:: :4:: : A:: A:: ::: :2:: 915
XXLPAT*34 MrIMI- NO- CM 1M OOWX V3* A4:A OC4a- FaL."
z- ft-pm 4 =








:A:- VAJ : A ::::: :41-
:A:: S; Az A..7 1


trniwt W40O. .0t ;:k: :A::: :0 4







P W4.44 .:I:. i.. .0.: .. .7 1 ..: ....' -


TriIp k4.%, o :: 5)i : A- 6 V 0 45 7
Y~aLmO)tn~n :::: N:CIID *~! rP- OI~ cwd "


Do :mKA : M :-0 .0 .4p! IF MTcm r4" CM-.-:" Orpow7 W tort
WT OTWCTP ocan"'I: -$A 4M


0 4 :A:. An :Am A A 1 : 4 r :
'~2Ls EYEt ~ awu.'w*.1C AYIK'i-LPbI'W2 :4:'lt~ :


rt


PI
















PPMC.JEC1T "'"' tnm hEa
NPAES: Pennlpacker Edingor Geiger Johnson SRP inc ,A,: M: .4
....i RN-- -M3.. --- "- --K -- --fO-- ,,i,._
US Ho 2 LEAD PF C a 35. I9 a CIa ,.
... ... sol s .

Env 0. I:.Jlr %ze marl.1 1 S








c s Te to r'
U :::F ::y ::;y :::. ::7: :: rf ::F :: f :: :5 :. : : 7:4 ::j: 7779. ::74 77: ::: ::: .
1:0 :;:i: : ;: y :;:y "-': ::v :::: 79: ;:7: tf ::, :7 : ::'f: ::,: :.,: 779: ::: ::, : ; :: -^
-:y :::F ;::: t:: ::: :: y ":-i: ::" : :: : ::: :: ::: :47 ::( : :::: ::,: "' :7: ::i --






::I: nime -io I: s
~ -:: : :;S ":I: :: ': :: : ; t: : : :: : :: : : T '. :: :







::: : :: :: : ::: : : : : : :: ,: :: : :: : :- : : : ::. ::-, y








5 : 79 ::;.: :: :i 4 7.57 : 75: ::: .-. 7 .. .- *. -a 7::: ::(: ::y: :?:y 6" '
^ ::, :;^ 74 t ** r A ,i 3 S "vr ae t-o
-* :::|: :: :: .: ::: :- i_ : :


:y :T %T::: :::p::: ioi::4 s 'a WtM> 'Jw "-f 7eW t aI--
I



- --- --- --
S ^^ ^**l~'; ~ '^'f; '- '-'**'* '-*' '*'*'*'*j~i L ~~ .'ijb;(.-'*f' '*"t- roncaf S .M wB :i,= =M ^-n '^ ")":'`' r~, ~,~ ~Z.


: ... .. ... .. .
a .: .: ,. w ., _, ::::: :.,:! .






S,. .- -. ..




0 ** -**' **. "/, n -* '


F r -


775: 775


::4: :t s:
::: p* ::If:
7:5: hit :-


.-6 t: 7 :g :Y 7 *
S :** **k I : '
S r 4" b" *': : "




S *



0 *J* 9 : 9 b 9
On ^ l B : "
S :. : .. .


"' : t I "*' 4 t: Is "
: 0* "7" ," : **'* i 9 :' "' ." :' 7: :: "


S. '; J7 ~ KK:;~:4:T7w 7 1- *.


I

























LSE No2 LEAD PENCIL F I" V t- 1 ,J,0,SH N O r C,,
LAT 6 85 J335 1Kelee CItI?
iF A OGoUP NIAMEC I) KItt lein

"' .I : ..: :.. T r, : "




l t I 'T Is -sto nTO o1 e f ar ..cn Ib -



IF I "-TfY E l. U.T A. "E.r[ L 1 -. ... ..... ...


0. : ::: : : ;:: : ::

:0: ::t: :: :: :: : :: : :: : ::r:: -: : :::









.. ST R T..5.IA..t


- ; z .'.; .: -


:



S .:- .-: ,














". : .r '. r ,-.s, ,':.



:~* **r '<. ., mi. nC.


." ..:. r





v. c







enr- ''a "cy,

. s..5 "..o ftfn,


'lTY. Gaiines'vi e t:::::.fo


:::: :::: ::: ::::

::g:: :- : *: ::4: the.

::S:: ::t: ::*: ::&: hr.

:::: ::- : :: ~: :: 7: tM


:*x :9


oak S t nts6 M :( O uMOtILftC# csaM ;: :--:: --.

: ::: ::(:: :- :: : I:: :4:: tne :4:: :*: :::: :4: :.;


LLS T.Q V'1LQut C IF v:
'I I *1I :1i2 *:5W0 4 w *m0- :A& 230

ft ::: :-o- :sM: :*o: ;: :dla :fa- :a- *m 1


off" O ::::: ::::: -: :

:A-: :::. :JQ: :R:: .42:: A


o:: :*i: : :: :::: :t*:: ::::


-..- --.. -, .,, : _






S .. ..... ... : .. -

- .:- .r





.". .. "- ... .. ..... .. -.



.








... ..:..


'i '" ". M .' :' a'. ..o C:'* .
h .. .r rc T.. ,... 0C :. .


flr~ r r m z r r-r. -- -


30 AL :SECUFITt:-C t :S CNED.N.9RfI ::::: ::::: ::::: ::::: :::::

Sk-amrm act D*>< actr "n 0'y" p IUn- osp"a"" P

:0: : ::: ::3 ::4- ::t: ::: 5::: ;t: ::=:

0: ::: :: ::2: ::5: :-: ::t: ::*: ::t: :n:

:0: ::: ::: : :::3:: ::: :: :: ::: :;

:50: ::: ::l- :::: ::*: ::B: ::*: ::;: ::9: ::S:


:0: ::t: :: ::. : ::4: ::: ::a; ::I: ::&: :::


::0: ::: ::t: : W: ::A: ml ::&: :A: ::R: ::1: :::

:0: ::t; :.: !::: ::: 1** ::*: : 6: ::li: ::: :::


:5:: :4:: :T+ :A :
:*-:- :-*-- **?** -


::::t ::' :3: ::k: to.#

::1:: ::E: : A: ::4:: (M>-


. . ... .. .. ,, .. __


::A: re


:OjASS LAtXvsxw -:::: Ypwr waHEd gHr T : :go : :if: : S: 0- xvtf

IM T 0a4-0Tr:a WrE: ::: W t ::it: ::I:: :2:: ::!: ::t: ::+:: ::7::

*V TION:CC'UB:ION, ::':: g, "s fps- '.s S- 6 Wf L ai. #

F0S3C, :::: ::::::::: wI AiVr g. wr r fr~ r ne p.

arSm Wtu .. gpW H ONOO f# -: --- --- : -4


0. .r ;.T -:.- .. .: -c ::.



-n m" -. i =i f-


., &4 *-.. .. HI. ...-4 :,.:- -*51


~_ _1 ~_i i_ :;I


Ga) m > ::::: :::: ::::: ;::7 :::..* :::: :::= :;:: -


:~: :~:: ::h: :P: :~f


~ lC~~llmCIR*-'M)*C Ca BCI* m rs


~

:: :- L- ;- :-





:~- ~ ~-:


. us Ma


1a*hLF4


::B: :.a: ::.: ::-: :.:


;:0: ::-: ::p: ::s:





87




dX
z5
U
0 04
M I






Si i j I 11: t -





80 !.! .I 0








o~~ t Ld 8
A-
94,

LLL
liIx
4O









wt II




T
31ntmlli bd SP11"MO
: iL ] 1 I

3jf~.J~AJrJJ SJ11A'l' O-.r 800;
























APPENDIX E

Performance Rates for All
Children Participating in
the Study






89













S-OOMMOOO





< ,0 o m000000














u wcnoommo
O oooo NO OOt



0 00000000000







0 N- NOO
0 000 0 CD 0 0 0 00









I
0 . .1 .0
0 00000000








LU




0 u I, C,, 0 0 i0




uco 0 (_ )0 N -'_ (_) 0 _-4
u coooooooooo
I- I . .


















n Q 0C r- il 0t LFl 0
-9






CO Oo o o










-J 00 0000 0000
ca
0000000000000



U OOC0000000






a r r f O co NO





5: 0 00~00000000
2 orio n*n.n'I
3 o; *-i *< .-i -
1C






90

9O



0000000
< o00o0000
I-- ,--4 ,_--4 ,-4 \



'--' n O ', O 0 O
a





Om LOOOOOOO

0 mON0oomOO
0- c0 "0 L '--.- O O



0


T 0-L'000000 O0


i- 0000N







u o 0ooo "
uC" 0000000






L c i I c o o






S* *. *. ..
T- C M r Vr- cw 0 n







0



< '0 mr cOa o M
I * .

I 0n

< o noc..iomoo






co CO CJ D CC) -40 C % L Ir o f-- 1-
u 000000-00000

a omoc oooomnoo






U 000000000000



&z I
I-






















< U 00000000

S * . .
Ln





















u oaoooooooaooo
U 000000000o00
CL
V)




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs