• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 Abstract
 Introduction
 Review of related literature
 Method
 Results
 Discussion
 Appendices
 References
 Biographical sketch














Group Title: effects of oral reading rate and reinforcement on reading comprehension
Title: The effects of oral reading rate and reinforcement on reading comprehension
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00099092/00001
 Material Information
Title: The effects of oral reading rate and reinforcement on reading comprehension
Physical Description: viii, 163 leaves : ill. ; 28 cm.
Language: English
Creator: Armstrong, Stephen W ( Stephen William )
Copyright Date: 1981
 Subjects
Subject: Oral reading   ( lcsh )
Reading comprehension   ( lcsh )
Reading disability   ( lcsh )
Special Education thesis Ph. D
Dissertations, Academic -- Special Education -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis (Ph. D.)--University of Florida, 1981.
Bibliography: Bibliography: leaves 157-162.
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Stephen W. Armstrong.
 Record Information
Bibliographic ID: UF00099092
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 - 000294372
oclc - 07712279
notis - ABS0697

Downloads

This item has the following downloads:

PDF ( 5 MBs ) ( PDF )


Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
        Page iii
    Table of Contents
        Page iv
        Page v
        Page vi
    Abstract
        Page vii
        Page viii
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
    Review of related literature
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
    Method
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
    Results
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
    Discussion
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
    Appendices
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
    References
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
    Biographical sketch
        Page 163
        Page 164
        Page 165
        Page 166
Full Text














THE EFFECTS OF ORAL READING RATE AND
REINFORCEMENT ON READING COMPREHENSION






BY



STEPHEN W. ARMSTRONG


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
















ACKNOWLEDGMENTS


Acknowledgments are due to each of my committee members who all

played an active role in the development of this investigation. Bill

Wolking was the first member to become involved and helped in a

number of areas. He helped me to conceptualize the problem.

Secondly, he helped to edit the original proposal. Third, it was

through his training in the difficult topic of verbal behavior that

I was able to construct the sequence of paradigms that make up the

verbal episode called reading comprehension. Finally, he was helpful

in providing suggestions based upon his knowledge of the literature

and research background.

Cecil Mercer, chairman of my committee, contributed more and more

in terms of his outstanding writing skills and knowledge of the litera-

ture. Also, his leadership and continuing support have been instru-

mental in making this end result possible. Bob Algozzine and Rex

Schmid have both been highly supportive in general and especially

concerning the conceptual viewpoint taken throughout the dissertation

itself. A special thanks goes to Rex for his training in writing

style and technique and to Bob for his understanding of research

above and beyond theoretical diversity. Jim Johnston has been the

most influential in terms of operant research methodology and the

need for a conservative approach to the science of human behavior.











Thanks also go to Leila Cantara whose contribution includes

typing under adverse time constraints and doing many of the "little"

things that make any dissertation possible when the author lives

400 miles away. The support provided by my wife, Bobbi, throughout

the entire doctoral program must certainly be acknowledged. She has

worked both at her job and at home not fully understanding but accept-

ing my unusual work hours, time lines, and points of stress. Also,

my parents have provided considerable support along the way whenever

it was needed.

Finally, thanks go out to the many people who helped in the north

Georgia area. Dave Henderson was instrumental in locating subjects.

Beth Grindle worked tirelessly on the tedious job of extracting data

from the tapes and no amount of money could repay that debt. My

bosses, Jewel Wade and Chris Sharpe, have done all that they could

to allow me the time to finish this dissertation and still meet my

responsibilities as an employee.

















TABLE OF CONTENTS



ACKNOWLEDGMENTS. . . . . . . . . .. .ii

ABSTRACT . . . . . . . .... . . . .vii

CHAPTER I

INTRODUCTION . . . . . . . . 1

Comprehension . . . . . . . . . . 2
Reading Rate (Frequency) . .... ......... 2
Oral Reading Rate and Comprehension . . . . .. 3
Operant Analysis . . . . .... ... . . 4
Statement of the Problem ..... . . . . . 4
Questions Under Investigation . ......... 5
Delimitations . . . . . . . ... .. . . 5

CHAPTER II

REVIEW OF RELATED LITERATURE . . . . . . . 6

The Nature of Reading Comprehension. . . . . ... 6
Reading Comprehension as Skills . . . . ... 7
Reading Comprehension as Memory ... . . . 8
Reading Comprehension as Verbal Behavior . . . 8
Textual Behavior. ... . . . . . . .. 9
The Establishing Operation . . . . . . . 10
The Mand . . . . . . . . . . . 10
Prompts, Intraverbals, and Audience Effects . . .. 11
Three Types of Extensions of Verbal Behavior . .. 11
Implications . . . . . . . . . . 14
The Relationship of Reading Comprehension with Selected
Subskills . . . ........ .. . . 15
Listening and Reading Comprehension . . . . .. 16
Decoding Ability (Oral Reading Rate) and Comprehension. 16
Comprehension of Silent and Oral Reading ...... 18
Operant Approaches to Reading. ............ . 19
Reading as Operant Behavior . . . . . . 19
Operant Analysis of Reading Comprehension . . . 20
Providing Reinforcers for the Answer Itself. . . . 20
Comprehension Questions as Discriminative Stimuli for
Intraverbals ............. .. .... 21













Reinforcement of Textual Responses and Comprehension
Answers . . . . . . . . . . . 21
Prerequisite Skill Training . . . . . . 24
The Establishing Operation .. . . ..... 24
Summary . . . . . . . . . . . .26

CHAPTER III

METHOD . . . . . . . . . . . . 27

Subjects......... . . . . . . . 27
Materials and Apparatus . . 6. . . . ...28
Reading Materials . . . . . . . . . .28
Apparatus . . . . . . . . . . . 29
Comprehension Questions . . . . . . . .29
Reinforcing Stimuli . . . . . . . . .29
Setting . . . . . . . . . . . . .29
Procedure ....... ....................... 30
Pilot Studies . . . . . . . . . . 30
Experimental Procedure .. . . . ...... 30
Experimental Design . . . . . . . . 34
Dependent Variables ................ 35
Data Collection . . . . . . . . . . 35
Data Analysis . . . . . . . .... 35

CHAPTER IV

RESULTS . . . . . . . . .. .. . . . 36

The effects of contingent Pennies Upon Comprehension
Answer Percentage Correct . . ............ 37
The Effects of the Pennies Contingency Upon Response
Latency for Three Levels of Reading Materials . ..... 40
The Mean and Range of Words Said Correctly Per Minute for
Each Comprehension Answer Percentage Correct Across All
Reading Levels . . . . . . . . . . . 42
Mean and Range of Words Said Incorrectly Per Minute for
Each Comprehension Answer Percentage Correct Across All
Three Reading Levels . . . . . . . . .50
Mean Comprehension Answer Latency as a Function of Words
Read Correctly Per Minute Across All Three.Reading Levels.
Mean Comprehension Answer Latency as a Function of Words
Said Incorrectly Per Minute . . . . . . . 58


CHAPTER V

DISCUSSION . . . . . .... . . . . 73

Contingencies of Reinforcement and Comprehension . . 73










Percentage Correct Comprehension Answers . . ... 73
Mean Comprehension Answer Latency . . . . .. 74
The Relationship Between Comprehension and Oral Reading
Rate . . . . . . . .. .. . . . . 75
Mean and Range of Words Read Correctly Per Minute by
Comprehension Answer Percentage Correct . . ... 75
The Mean and Range of Words Read Incorrectly Per Minute
for Each Comprehension Answer Percentage Correct . 76
The Mean Comprehension Answer Response Latency and
Words Read Correctly Per Minute ........... 76
Mean Comprehension Answer Latency and Errors Per Minute 76
Comparison of the Present Study with the Previous Research
on the Relationship Between Oral Reading Rate and Reading
Comprehension Performance . . . . . . .... .77
Implications . . . . . . . .. .. . .. 78
Limitations . . . . . . . .. .. . . .. .79
Future Research . . . . . . . . ... . 79

APPENDIX A STATE OF GEORGIA REGULATIONS AND PROCEDURES
FOR SPECIFIC LEARNING DISABILITIES . . .. 81

B GUIDELINES FOR COMPREHENSION QUESTIONS . . 85

C DEFINITION OF TERMS . . . . ... 87

D NOTATION OF DESIGN ELEMENTS . . . . 89

E OBSERVER DIRECTIONS AND SAMPLE SCORING SHEET 91

F RAW DATA . . . . . . . . ... .94

G FIGURES DISPLAYING THE RELATIONSHIP BETWEEN
PENNIES CONTINGENT ON READING COMPREHENSION
PERFORMANCE . . . . . . . . 130

H FIGURES DISPLAYING THE RELATIONSHIP BETWEEN
PENNIES CONTINGENT ON READING COMPREHENSION
LATENCY . . . . . . . . ... .138

I THE RELATIONSHIP OF WORDS PER MINUTE CORRECT
AND ERROR WITH READING COMPREHENSION ANSWER
PERCENTAGE CORRECT FOR EACH MATERIAL DIFFICULTY
LEVEL BY SUBJECT . . . . . .... 144

REFERENCES . . . . . . . ... ...... 157

BIOGRAPHICAL SKETCH . . . . . . . .... 163

















Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of
the Requirements for the Degree of Doctor of Philosophy



THE EFFECTS OF ORAL READING RATE AND
REINFORCEMENT ON READING COMPREHENSION

By

Stephen W. Armstrong
March, 1981
Chairman: Cecil D. Mercer
Major Department: Special Education

The present study of six learning disabled students was composed

of two parts: (a) an investigation of the effects of contingencies of

reinforcement upon comprehension answer percentage correct and latency

across three levels of materials and (b) an investigation of the

correlations of words read orally correct per minute, words read

orally incorrect per minute, comprehensive answer percentage correct,

and mean comprehension answer latency.

The results for part one are mixed. The potential reinforcer

employed was pennies and they failed to show clear reinforcing effects

in most cases for either percent correct or latency. There was some

indication of the pennies condition increasing comprehension per-

centage when the students were at an instructional material level.

The results for part two were straightforward. There was a

strong positive correlation for all six subjects between words read

orally correct and comprehension answer percentage correct. All











subjects showed a strong negative relationship between words read

orally correct and mean comprehension answer latency and a strong

positive correlation between words read incorrectly per minute and

mean comprehension answer latency.

An additional finding was that these subjects, in grades three

and four, could comprehend material in which their oral reading

performances were considerably lower than the guidelines suggested

by the literature. None of the six subjects met the oral reading

criteria commonly cited in the literature while most subjects were

able to achieve 100 percent comprehension of factual material from

the reading passage.

















CHAPTER I

INTRODUCTION



Illiteracy is one of today's major educational crises. Recently

a Harris Survey found 18 million Americans functionally illiterate

(Haring, 1978). Some students appear to read (orally) well enough,

but do not comprehend what they read. These children are referred

to as "word callers" and often show a general lack of ability to

comprehend written or spoken language (Hansen & Eaton, 1978; Lerner,

1976).

Unfortunately, the literature contains little in the way of

experimental data on how to improve comprehension (Van Etten, 1978).

Some studies have demonstrated the effectiveness of feedback and

reinforcement for correct answers (Hansen & Lovitt, 1976; Klann,

1972; Lovitt & Hansen, 1976a; Proe & Wade, 1973). None of these

studies have adequately controlled for the effects of the level of

difficulty of material, student proficiency, and comprehension

question type. Problems in reading are identified more frequently

for learning disabled students than other deficit areas (McLoed, &

Crump, 1978). Therefore a further understanding of the relationship

between reading skills and the remediation of these skills could have

wide application for teachers of the learning disabled. The rationale











of this study is based upon a brief discussion of reading comprehen-

sion and three other related areas. These areas are reading rate,

the relationship between reading rate and comprehension, and operant

analysis as a method of experimental inquiry.


Comprehension

Most reading research has focused upon aspects of decoding with

relatively little attention given to reading comprehension. One very

good reason for this is the "fuzzy nature of the definition of

reading comprehension" (Van Etten, 1978, p. 42).

While less experimental analysis of reading comprehension has

occurred, considerable effort has been made to define its critical

components. These components vary in number from four to six and

typically range in type from basic recognition and recall of factual

material to problem solving and critical judgment (Van Etten, 1978).

Interventions designed for all levels of comprehension may lead to

differential effects in each of the areas of recall, sequence, and

interpretation (Hansen & Lovitt, 1976).


Reading Rate (Frequency)

A frequent concern of teachers is determining when mastery has

been achieved. A relatively recent method has been proposed that

combines a time and accuracy criterion. This criterion is rate or

frequency of both number of movements (words) correct and number

of errors.











Authors have suggested average rate criteria and formulae for

determining skill mastery. Alper, Nowlin, Lemoninre, Perine, and

Bettencourt (1974) suggested an oral reading rate of 100-200 words

per minute correct with less than two errors per minute. Starlin

(1971) also recommended 100-120 words per minute correct and two or

less errors. Haughton (1977) proposed a frequency of 200-400 words

per minute correct. The empirical basis for these figures is weak

as evidenced by the variation in criteria. A method used to measure

mastery in a particular level of material is the celebration or

increase in rate on the next level or skill (Haughton, 1972). Lovitt

(1977) suggests teachers assess the level of performance of students

achieving successfully. All of these techniques are supported weakly

or unsupported at all by research data.


Oral Reading Rate and Comprehension


A relationship between reading proficiency (oral reading rate)

and the correct answering of comprehension questions would appear to

be logical. Hansen and Lovitt (1976) found that interventions directed

towards oral reading led to increases in comprehension of orally and

silently read material. Lovitt and Hansen (1976a) utilized contingent

skipped and drilling to improve oral reading rate and comprehension

concurrently. Jenkins, Barksdale, and Clinton (1978) found only

minimal generalization from intervention directed toward oral reading

rate to reading comprehension. These divergent findings need to be

clarified by further research.











Operant Analysis


Operant analysis or applied behavior analysis is a system involv-

ing five basic elements: replicable teaching procedures, individual

analysis, experimental control, direct measurement, and daily measure-

ment (Lovitt, 1976). Replicable teaching procedures include the

amount of time devoted to instruction, words being taught, and the

methods used to teach them. Individual analysis focuses on the effects

of instruction on each subject's rate of learning. Experimental

control stresses the determination of a functional relationship

between independent and dependent variables by repeated opportunities

to view the changes in behavior as a result of an environmental manipu-

lation (Johnston & Pennypacker, in press). Direct measurement

procedures involve obtaining data on actual behavior that is taught.

Daily measurement leads to greater confidence in an effect, since the

changes are observed over several days. The likelihood of obtaining

atypical data is thus reduced (Haring, 1978). Thus operant analysis

lends itself.to research which can be applied directly to classroom

instruction.



Statement of the Problem


Reading specialists have emphasized comprehension as the goal

in the reading process (Stauffer, 1970). Techniques, however, that

have been used in the treatment of poor reading comprehension are in

need of experimental verification (Van Etten, 1978). The question

of what teaching procedures are necessary in the acquisition of











comprehension is an especially critical one to all teachers of

reading. Additionally, strong empirical criteria for oral reading

proficiency is lacking. The effects of reinforcement contingency

will be examined at three levels of difficulty upon reading compre-

hension.


Questions Under Investigation


1. What is the relationship between comprehension answers and

oral reading rate?

2. What is-the relationship between contingencies of

reinforcement and comprehension performance?


Delimitations


Certain types of comprehension do not lend themselves to develop-

ing objective questions. Comprehension will be restricted to factual

recall. The areas of analysis, sequence, interpretation, synthesis,

critical reading (evaluation), and creative reading will be excluded

from study.
















CHAPTER II

REVIEW OF RELATED LITERATURE



The following review is divided into three sections. The first

section focuses on the nature of reading comprehension ending with a

conceptualization based on Verbal Behavior (Skinner, 1957). The

second section describes the relationship between reading comprehen-

sion and other skills. The final section describes operant research

in reading. The findings of both general and reading comprehension

research are organized according to a Verbal Behavior (Skinner, 1957)

framework.



The Nature of Reading Comprehension


Readingcomprehension has typically been conceptualized as a

multi-level skill. Smith (1969), utilizing the Taxonomy of Educational

Objectives (Bloom, Engelhart, Furst, Hill, & Krathwohl, 1956),

determined seven categories of reading comprehension. These are

memory, translation, interpretation, application, analysis, synthesis,

and evaluation. Later she reported that this list was too cumbersome

and proposed the four levels of literal, interpretative, critical, and

creative reading comprehension. Literal comprehension is recognizing

and understanding the direct ideas of the author. Interpretation











refers to discerning implied meanings in the text. Critical reading

involves forming generalizations, drawing conclusions, making compari-

sons, analyzing and applying ideas from the text. Creative reading

includes developing new thoughts, fresh ideas, and imaginative

insights from reading (Smith, 1972).

Miller (1976) detailed four types of reading comprehension. These

are (a) recalling information from long-term memory; (b) elaborating

and abstracting: (c) interpreting; and (d) naming. Rystrom (1970)

gave a hierarchy of reading comprehension which involves vocabulary,

syntax, item recall, item sequence, inference, and evaluation.

Hansen and Lovitt (1976) utilized recall, sequence, and interpretation

in a study of comprehension of orally and silently read material.

Guszak (1969) attempted to determine what types of reading

comprehension were actually being used by elementary reading teachers

with their pupils. Findings indicated that 56.9 percent of the

questions involved immediate recall of factual information. Less

than 30 percent of the questions were concerned with higher thinking

processes such as evaluation, judgment, translation, conjecture, and

explanation.


Reading Comprehension as Skills

Smith (1969) makes the point that reading comprehension cannot

be taught, since it is too broad a category. Instead, specific

skills must be identified. Fareed (1971) offered a list of compre-

hension skills. These are "(1) noting clearly stated main facts and

important details; (2) grasping the main ideas; (3) following the

sequence of events or steps; (4) drawing inferences and reaching











conclusions; (5) organizing ideas and relationships; (6) applying

what is read to solve problems and verify statements; and (7) evaluat-

ing material for bias, relevancy, and consistency" (Lerner, 1976,

p. 311).


Reading Comprehension as Memory

Comprehension of written material is made up primarily of memory

and what is frequently called thinking processes. Higher levels of

comprehension, such as interpretive, critical, and creative reading,

are closely related to thinking processes (Lerner, 1976). These

levels, however, are not as important to reading teachers as recall

and recognition (Guszak, 1969). Poulton and Brown (1967) made the

case that reading comprehension would be "more correctly described

as a measure of memory" (p. 219). Some implications for reading

comprehension are readily apparent. The difficulty of recalling

information for comprehension questions will depend upon the length

of the passage. This factor needs to be controlled in the development

of reading materials for investigation.


Reading Comprehension as Verbal Behavior

Reading comprehension questions and answers, like other forms of

what we call language, have traditionally been analyzed in terms of

their form of "meaning." Another approach, functional analysis, has

been advocated by B. F. Skinner in the book Verbal Behavior (1957).

This method involves accounting for the relevant conditions and

variables of which the behavior is a function. When these have been

identified for both the speaker and the listener we have a full











account of verbal episode. Interaction of these conditions and

variables, as well as the roles of the speaker and listener result

in multiple causation. Skinner's definition of verbal behavior

includes two elements: (a) that it is behavior reinforced through

the mediation of other persons and (b) the action of the mediating

individual must have been specifically trained to provide such rein-

forcement.

While Verbal Behavior was written to be an all inclusive account

of this subject matter, we need only borrow certain concepts, rela-

tionships, and extensions to deal with oral reading, comprehension

questions and their answers. An introductory text by Peterson (1978)

will be used to provide definite features for these parts of verbal

behavior.


Textual Behavior

The defining characteristics of textual behavior are that the

response is vocal, it is controlled by a prior stimulus that is the

response produce (end result) of writing behavior (such as a book)

and that there is point-to-point correspondence between the stimulus

and the response. Basically, point-to-point correspondence refers

to discriminative stimuli and responses in which each discriminative

component controls a single response component regardless of the

mode of verbal behavior. If we say "cow" after hearing the word

"cow" or reading it there is point-to-point correspondence in both

cases. If we hear or read "cow" and say "hereford" there is no point-

to-point correspondence.











The Establishing Operation

An establishing operation is a difficult concept to define.

It precedes the response and increases the effectiveness of a

particular stimulus change as a reinforcer. Since both discrimina-

tive stimuli and establishing operations exert antecedent control

(or change the probability of an upcoming response) it is important

to contrast them. An example would be if a teacher states that each

correct comprehension answer will result in a nickel earned (and

nickels are currently reinforcing) then the reinforcing value of

the book or printed story is increased. Reinforcement does not

occur for comprehension answers unless they are correct and follow-

ing the appropriate question. Comprehension questions serve as

discriminative stimuli for correct comprehension answers if they

have been present when a given class or response has been reinforced.

The statement about nickels serves as an establishing operation and

the questions as discriminative stimuli.


The Mand

A mand is a verbal response whose form is controlled by an

establishing operation. The comprehensive questions from the view

of the teacher are mands. Teachers will have either read the story

or listened to the student read it prior to the questions and form

verbal behavior (questions) accordingly. Thus the story serves as an

establishing operation for the teacher's verbal behavior of asking

certain questions. Teachers are characteristically reinforced by

student answers which are controlled thematically by the story and

questions (are correct).











Prompts, Intraverbals, and Audience Effects

The comprehension questions serve as prompts to the student

by providing supplementary stimuli and allowing the teacher to

identify the response that the speaker is likely to emit. Compre-

hension answers (to the student) are generally intraverbals. They

are verbal responses in which the prior controlling variable is a

verbal stimulus (the story) with no point-to-point correspondence

between the stimulus (the question) and response (the answer). The

control is thematic. There is no point-to-point correspondence

between the question and answer. The teacher and the school setting

can serve as the audience effecting control over the response form

likely to be emitted (proper grammar) and a listener (the teacher)

in whose presence verbal behavior is typically reinforced.

Three Types of Extensions of Verbal Behavior

New situations may have certain commonalities to previous ones

and these common features may be either relevant or irrelevant. The

degree to which the new novel stimulus shares relevant features

determines whether the situation involves generic, metaphorical, or

metonymical extension.

Generic extension involves a previously learned response with a

novel stimulus that has all the relevant features of the stimulus

which previously controlled the response along with some that are

irrelevant. Relevance refers to stimuli that serve a definitive

function for a given concept of category. The comprehension question

contains the relevant features from the story (e.g., "What happened to

Lillian at the end of the story?" for the response (previously a











textual one) "She signed a contract to perform with the circus").

The stimulus is novel in terms of the form (a question) and the word

order. The response is one that is previously learned (as a textual)

and all the relevant features are present completing the defining

characteristics of generic extension of the intraverbal. This is

the case in questions requesting factual recall, putting events in

sequence, or the literal level of reading comprehension.

Metaphorical extension employs a previously learned response and

novel stimulus, but the novel stimulus has less than all the relevant

features of the previous controlling stimuli. This would relate to

questions requiring interpretative, inferential, or application

answers.

Metonymical extension requires a previously learned response

and a novel stimulus with none of the relevant stimulus features of

the stimuli previously controlling the response. These questions

would roughly correspond to creative and evaluative comprehension

questions and answers.

In Figure 1 individual A is the teacher and B is the student.

The sequence of events begins with the teacher saying that he/she

will give a nickel for each correct comprehension answer. This

serves as an establishing operation in that it increases the effec-

tiveness of the book as a reinforcer. The child then responds verbally

or reads the story (represented within the box). After the child

finishes reading the story, the teacher asks the child a comprehensive

question. This question is a mand from the teacher's point of view

and a discriminative stimulus to the student. The student listens to







Teacher: Establishing
Operation



Child: Textual


Rv "I will give you a nickel for each correct
comprehension answer."

Sr-D
Lillian . -- Rv "Lillian


Teacher: Mand


Child: Intraverbal
and Audience
Effect


Rv "What happened to Lillian . ."


SO I
Prompt Rv
"Sh


Sr-D

t


R (gives nickel)

1
-. Sr


Rv verbal response
SD discriminative stimulus
Sr reinforcing stimulus
Sr-D stimulus that is both discriminative and reinforcing

quotation marks vocal verbal response
words within boxes printed verbal response products (book)
parenthesis responses that are non-verbal
indicates series of events within one individual
S- indicates response of one individual, A or B, has effects upon other individual
Figure 1

Account of an Episode of Verbal Behavior Involving Oral Reading and
Comprehension Questions and Answers











the question which leads to intraverbals from the story relating

thematically (or in theme) to the question.

The student responds verbally with an answer which if correct

will characteristically reinforce the teacher. The teacher then

performs the non-verbal response of giving the student a nickel which

has been found to be reinforcing to the student. The teacher and

classroom setting exert the effect of the audience by determining the

form of verbal response.


Implications

When a breakdown occurs and a child fails to answer a comprehen-

sion question correctly, the verbal behavior model may be used to

locate the source of the problem. Each component can be analyzed

systematically from the establishing operation to the reinforcer.

The establishing operation in the example is somewhat artificial;

however, in the more common situation a teacher would tell the child

to read the story and be ready to answer comprehension questions.

The effectiveness of the establishing operation would depend upon

the child's history of reinforcement under similar circumstances.

If the child has been reinforced for answering comprehension questions,

then the book would take on reinforcing properties.

The textual response (reading from the book) would depend on

the child's ability to convert written verbal behavior to spoken

verbal behavior. When there is exact or near exact point-to-point

correspondence between the stimulus (the book) and the verbal

response (reading),the textual behavior is not the source of the

problem.











The student next is faced with the comprehension question

which should serve as discriminative stimuli for intraverbals (or

words thematically related) from the story. The comprehension

questions themselves may be inadequate as sources of discriminative

stimuli for the proper intraverbals and may need to be supplemented.

Some students may be able to answer certain types of comprehen-

sion questions and fail to do so with others. Sometimes a child

answers correctly factual recall, sequence, or literal level

questions (generic extension), but cannot answer (a) interpretative,

inferential, or application questions (metaphorical extension) or

(b) creative or evaluative comprehension questions metonymicall

extension). In this case the student's verbal behavior history is

deficient in terms of answering questions or providing intraverbals

when the discriminative stimuli are weak.

Finally the teacher's verbal or non-verbal response may or may

not serve as a reinforcer increasing the future probability of the

entire episode and especially the particular verbal response, the

comprehension answer.


The Relationship of Reading Comprehension
with Selected Subskills


Reading comprehension is dependent upon certain subskills for its

development. The connection between reading comprehension and (a)

listening comprehension, (b) oral reading proficiency (as measured by

rate), and (c) silent and orally read material will be discussed in

the following sections.











Listening and Reading Comprehension

Reading and listening are both modes of receiving information.

There are many similarities between them. Research studies (Vogel,

1975, 1977) have shown a high correlation between listening and

reading skills. Also, instruction in listening comprehension often

results in improvement in reading comprehension (Lerner, 1976).

Some differences between reading and listening are (1)

listening (language or speech) is typically acquired first; (2)

the process of speech acquisition is more gradual; (3) speech

acquisition is under the control of stronger reinforcers; and (4)

speech is typically taught in a one-to-one setting (Staats & Staats,

1962).

Since speech is learned before reading and there appears to

be a strong relationship between the two skills, it is necessary

to control for the level of speech development in subjects under-

going an experimental analysis of reading comprehension. Intra-

subject replication (Sidman, 1960) is an excellent way to control

for these effects, since the same subject receives the various

manipulations.


Decoding Ability (Oral Reading Rate) and Comprehension

A very common problem with teachers of reading is determining

reading ability in order to place children in a reader that is appro-

priate for them. This is also a problem or variable to control when

researching reading comprehension since proficiency in a reader will

certainly influence comprehension of the material in that reader.

Hansen and Eaton (1978) reported that training decoding skills often











leads to a "simultaneous improvement in comprehension" (p. 62).

The relationship between decoding ability and level of material is

critical when reading comprehension is investigated.

Lovitt and Hansen (1976b)described four ways (in order of

preference) in which a reader may be matched with a book and

suggested one that they consider more accurate. The most simple

method is to place a child in the reader that corresponds to his

or her age or grade. This is also the least accurate, for ability

is often unrelated to age or grade. Another method is to use achieve-

ment test scores to assign readers. The weakness here is that per-

formances on an achievement test reflect actual classroom performance

only indirectly. The third method is to use the placement tests that

are a part of many basal series. This method is certainly an

improvement over the previous choices, since the tests are related

to the material used later. However, the tests are generally only

given once and may not accurately reflect the pupil's true ability.

The fourth method, the informal reading inventory (IRI), typically

assesses two skills (word recognition and comprehension) and percentage

scores lead to three competency levels (independent, instructional, or

frustrational) (Lovitt & Hansen, 1976b).

Lovitt and Hansen (1976b)devised a method that combines the

IRI with techniques of applied behavior analysis. This combination

led to direct measurement across many levels (IRI) and frequent

measurement (applied behavior analysis). One of the measures is oral

reading rate (words per minute correct and errors per minute). They

placed students in the highest reader where their average correct











rate was between 45-65 words per minute and average error rate was

between 4-8 words per minute. They also used a comprehension score

between 50-75 percent. These criteria are for the instructional

level of performance.

Several authors have suggested set numerical criteria and

formulae for determining mastery of materials in order to progress

to the next level of difficulty. White and Haring (1976) and Hansen

and Eaton (1978) have proposed formulae for determining mastery.

These usually involve some ratio with tool skills. Starlin (1971),

Haughton (1972), and Alper, Nowlin, Lemonine, Perine, and Bettencourt

(1974) have offered set numerical criteria. These range from 100 to

400 words per minute correct and two or less errors. Wolking (1973),

as well as Starlin and Starlin (1973) found that there is considerable

variation between proficiency rates of adult and child performances.

Wolking found the proficiency rate of words per minute correct varied

between 50 for a high achieving first grader and 198 for an adult.

Wolking's data came from students at different grades who were selected

as being high and low readers. Words in text were read by the high

third grade readers at 120-132 words per minute correct and zero

errors per minute. Although the empirical basis for these rates

is somewhat lacking (Criswell, 1978), there does appear to be a

direct relationship between oral reading rate and reading mastery.


Comprehension of Silent and Oral Reading

Hansen and Lovitt (1976) found that when intervention was directed

upon oral reading and its comprehension, silently read material was











also influenced. They also found that comprehension of orally read

material was higher than silently read material (prior to interven-

vention). Comprehension, then, may be studied using oral or silent

reading. Although reading comprehension typically involves silent

reading, oral reading is more readily measured directly. Results

may be generalized to silent as well as oral reading performance.


Operant Approaches to Reading


A series of behavioral procedures resulting from the pioneering

work of B. F. Skinner have been grouped together to form a system

called Applied Behavior Analysis. Analysis on the individual level

is probably the recognizable feature of Applied Behavior Analysis (ABA).

This technique is closely related to the concept of experimental control.

Other research systems often report group data typically making compari-

sons between experimental and control groups (Haring, 1978). In ABA

the patterns of fluctuation in individual responding are readily

available for visual inspection and subsequent control. Because

this type of control is feasible, statistical procedures are unnecessary.

The systematic removal of sources of instability in the data leads to

stable responding. When stable responding has been achieved, the

effects of an intervention can be viewed with greater confidence

(Sidman, 1960).


Reading as Operant Behavior

Arthur Staats and his associates first utilized the techniques of

applied behavior analysis to explore the relationship between reading











behaviors and reinforcement (Raygor, Wark, & Warren, 1966; Staats,

Minke, Finley, Wolf, & Brooks, 1964; Staats & Staats, 1962). This

work led to the development of techniques and apparatus for the

investigation of reading behavior, as well as results indicating

that reading (like other operant behavior) is influenced by its

consequences.

Other early researchers analyzed reading behavior (Haring &

Hauck, 1969; Lovitt, Eaton, Kirkwood, & Pelander, 1971; Wolf, Giles,

& Hall, 1968). The early works, both in the laboratory and the field,

have focused on reinforcement as the independent variable and oral

reading as the dependent one.


Operant Analysis of Reading Comprehension

Recently ABA investigators have published studies in which reading

comprehension is a dependent variable of primary interest (Hansen &

Lovitt, 1976; Hauck, Metcalfe, & Bennet, 1978; Jenkins, Barksdale, &

Clinton, 1978; Knapczyk & Livingston, 1974; Lahey, McNees, & Brown,

1973; Lovitt.& Hansen, 1976a;Roberts & Smith, 1980). Each of these

studies can be categorized according to the paradigm(s) outlined in

Figure 1 with the addition of prerequisite skill training.


Providing Reinforcers for the Answer Itself

This approach comes in at the very end of the verbal episode and

merely provides reinforcement for end product or comprehension answer.

For this to work the subject (according to the account of the verbal

episode) must have successfully completed all the other paradigms.

This technique is simple and straightforward and was the first one











utilized in a study by Lahey, McNees, and Brown (1973). Lahey et al.

found pennies and praise to increase the percent of correct compre-

hension answers for two sixth graders who scored at grade level at

decoding and two years below grade level in comprehension.


Comprehension Questions as Discriminative Stimuli for Intraverbals

There can be considerable variation in how comprehension questions

are worded. Whether these questions provide sufficient stimuli to

evoke intraverbals (and comprehension answers) will determine the

student's ability to respond correctly regardless of a reinforcement

contingency or any other facet of the episode. Two investigations

have dealt with this topic. Knapczk and Livingston (1974) prompted

students to ask questions when they encountered unfamiliar words or

directions in the comprehension questions. The students comprehension

performances increased as they were able to provide themselves with

more and more stimuli leading to intraverbals from the story and

correct comprehension answers. Hansen and Lovitt (1976) had students

drill on the comprehension answers missed which should help them to

sort the stimuli that evoke the proper intraverbals from those that

lead to incorrect ones.


Reinforcement of Textual Responses and Comprehension Answers

Lovitt and Hansen (1976a) directed their investigation at two para-

digms concurrently (see Table 1). They employed contingent skipping and

drilling to improve oral reading rate correct and error along with com-

prehension answer percentage. The subjects were seven boys aged 8-12

categorized as learning disabled. When a subject met criteria for








Table 1

Review of Selected Studies on Reading Rate and Reading Comprehension


Author(s) Subjects Measures Treatment Results


Lovitt & Hansen
(1976a)


7 LD boys
BA 8-12


-words read per
minute correct
and error.
-percentage of
correct answers
on comprehension
questions


Contingency of skipping
reading material based
on reading performance.
Drilled on words read
incorrectly and questions
answered incorrectly.


Oral reading rate
improved during the
skip and drill
contingency.


Limitations: The effects of skipping and drill cannot be analyzed separately.
The changes from intervention phase to final baseline phase were
minimal. The baseline phases were short.


Jenkins, Barksdale,
& Clinton (1978)


3 LD students
CA= 11


-oral reading
rate correct.
-answers to
recall questions.


-used money contingent on
correct answers to com-
prehension questions.
-noncontingent money
condition.


-Comprehension
improved under
money contingency.
-Words read correctly
only increased
slightly under money
contingency.


Limitations: The major limitation is that all three students began at very high
level of reading correct words.










Table 1--Continued


Author(s) Subjects Measures Treatment Results

Roberts & Smith 8 LD boys -oral reading rate Manupulated contingencies Targeted reading
(1980) CA 10-12 correct to effect reading per- scores improved in
Reading grade -oral reading rate formance, e.g., tokens, each intervention
levels 1.2- incorrect Used modeling and condition, compre-
2.4 -comprehension instructions to improve hension improved
answer percentage reading performance. less than oral
correct reading rate correct.
Limitations: Since the interventions were mingled it is impossible to determine
the respective effects of each treatment. The variability scores of
the subjects was considerable and limits definitive conclusions.











oral reading and comprehension, a predetermined number of pages were

skipped. Both oral reading and comprehension increased under this

contingency.


Prerequisite Skill Training

Hauck et al. (1978) investigated a 10 year old reading student

who displayed excellent decoding ability, but could not comprehend

what he read. A complicated procedure was instituted which involved

task analyzing reading comprehension into its component skills and

training each component which lead to increased comprehension.


The Establishing Operation

Jenkins et al. (1978) investigated what was called generalization

across behaviors, settings, and time of interventions (see Table 1).

This study found that increasing oral reading rate led to only minimal

gains in reading comprehension. This "generalization" is more likely

to be the effects of reinforcing oral reading serving as an establish-

ing operation (or making books more reinforcing) which would account

for increases in many reading behaviors. The effects would be stronger

in the same book versus a new one and any number of factors could

influence their magnitude. Hansen and Lovitt (1976) found that with

students whose oral reading increased by direct intervention, reading

comprehension also increased without intervention.

Roberts and Smith (1980) worked with eight learning disabled

students who had oral reading and comprehension deficiencies. They

found that when their intervention package was directed at one target

behavior (rate correct, error rate, or comprehension answer percentage










correct) that the targeted behavior increased with only mild

increases in non-targeted behaviors. However, as noted in Table 1,

the treatment package involved contingencies that were presumably

reinforcing and this would lead to the reading material becoming

established as reinforcing in its own right and causing the increases

in non-targeted behaviors.

In all of these cases the experimenters thought they were

investigating the relationship between oral reading and comprehension.

Their subjects were considered deficient in both behaviors. They felt

that oral reading or texting was a necessary verbal behavior for

responding correctly to comprehension questions. So, if they increased

oral reading, the comprehension should increase as well. This all

follows the example (Figure 1) of a verbal episode that leads to a

correct reading comprehension answer.

The problem, as noted earlier, is that we have very little in

the way of empirical data to determine at what level oral reading

rate should be for the subject to be considered proficient (and thus

be capable of correct comprehension responses). The "deficient"

oral readers in these studies may be below some arbitrary criteria

but be reading well enough to comprehend the material (see Table 1

for specific criticisms of the study). Therefore, when reinforcement

is directed to oral reading rate this increases and the same occurs

when reinforcement is for comprehension answers with only minimal

increases in the non-targeted behavior due to the establishing

operation effects.











Summary


Although several conceptualizations of reading comprehension

have been developed the functional viewpoint extracted from

Skinner's Verbal Behavior (1957) appears to hold much promise.

Certain skills such as listening and decoding are prerequisite to

reading comprehension. However, we have little research to deter-

mine the level of decoding or oral reading sufficient for reading

comprehension. It is also important that findings for the easier

material to measure oral reading can be extrapolated to silent

reading which occurs more frequently. Finally, certain paradigms

of the model for the entire verbal episode called reading compre-

hension have been investigated. It has been found that these

changes lead to increased comprehension or the predictable effects

when the subject's performance breaks down at that part of the

episode.

















CHAPTER III

METHOD



In Chapter III the methods and procedures of the study are

presented. Included are a description of the subjects, materials,

apparatus, settings, procedures, and experimental design. The study

is composed of four major phases:

1. Identification of eligible subjects.

2. Identification of eligible subjects meeting criteria

for discrepancy between oral reading and comprehension.

3. The selection of three reading material levels.

4. Applications of reinforcement procedures to increase

reading comprehension performance.



Subjects


The six subjects were third and fourth grade students from a

rural area of North Georgia. The five males and one female were

selected initially from the students currently being served in the

specific learning disabilities class (see Appendix A, Georgia Depart-

ment of Education Program for Exceptional Children: Rules and

Regulations, 1978). The next qualification was that each student











be placed in the special education program primarily due to dis-

crepancies in reading performance and ability. A final requirement

was that each subject was neither a non-reader nor had too high a

reading level in spite of eligibility criteria from the state.

This final requirement is important due to the need to locate three

separate reading levels for data collection processes.


Materials and Apparatus


Reading Materials

The Barnell-Loft Specific Skills Reading Series (Boning, 1976)

is used as stimulus materials. The readers consist of one page stories

at a single reading level. The selection of the actual reader depends

upon the individual subject's oral reading rate in those readers. A

reader in which the oral reading performance falls within each of

the following ranges was selected:

1. The high reading levels were determined by using the

least difficult reader.

2. The low reading levels were selected by having the child

read progressively higher material until respondent

behavior (e.g., voice breaks) interfered with reading

behavior.

3. Middle readers were located in which the words per

minute correct and errors per minute had the least

overlap with the low and high readers.

Three days of baseline data in which the subject's performance fell

within these ranges were considered sufficient to initiate the study

using these materials.











Apparatus

Subjects' reading performances and comprehension performances,

along with the experimenter's orally presented comprehension questions,

were recorded by a cassette recorder (Sony Model TC-205). These

cassettes provided a permanent record for the major analysis and

further analysis if this is deemed necessary at a later date.


Comprehension Questions

Comprehension questions were developed with five recall questions

per story. Questions included the who, what, when, and where variety

as employed by Jenkins, Barksdale, and McClinton (1978) (see Appendix

B). Answers came directly from the story content.


Reinforcing Stimuli

Pennies were chosen as the reinforcer. This generalized rein-

forcer was employed by Lahey, McNees, and Brown (1973) and is parti-

cularly convenient for this experimental purpose. Pennies were placed

in a clear plastic container in view of the subject. A penny was dis-

pensed after every correct response (FRI) to a comprehension question.


Setting


The procedure was carried out in the learning disabilities class-

room in a partitioned corner (approximately four feet square) away

from the main activity centers of the classroom. A small table was

set in the corner upon which was placed the cassette recorder and the

reading material being used at the time. Observer and student sat

diagonally across from each other at the table. The partitions











blocked the student's view of the rest of the class except when a

student sharpened his/her pencil at a nearby pencil sharpener.


Procedure


Pilot Studies

Pilot study one. A pilot study was performed during August,

1977, in order to explore reading performance characteristics. The

Durrell Analysis of Reading Difficulty (Durrell, 1955) was administered

to 14 middle school students. All 14 students showed a high correlation

between oral reading rate (correct and error) and percent of correct

comprehension questions. This led to an interest in whether comprehen-

sion would be limited by material difficulty under optimal (highly

reinforced) conditions.

Pilot study two. In order to determine the stability of the

vocabulary in the Barnell-Loft reading materials and see whether the

proposed ranges of oral reading performance can be maintained, a

second pilot.study was carried out during May of 1979. Datawere taken

under baseline conditions with two subjects for five days at three

levels each (see Table 1). The observer was unable to mark reading

errors, comprehension latencies, or evaluate the correctness of the

answers during the performance due to a lack of time. This led to

scoring being done from audiotape.


Experimental Procedure

The subjects read three stories (one at each level) and answered

five comprehension questions each for a total of 15 questions. The

















Table 2

Summary of Data from Pilot Study Two


Grade Correct and Error Correct and Error
Level of Material Median Performance Range
Per Minute Per Minute


first 111/0 102-125/0-6

U third 78/4 75- 85/2-5

S sixth 62/6 51- 75/6-8


first 98/2 77-121/0-3
-p
U second 86/3 71- 97/1-9

i third 66/4 31- 76/0-6











stories were all read orally and rates, correct and error, recorded.

The subject then read the story silently. The comprehension questions

were read orally by the experimenter and answered orally by the

subjects.

A detailed summary of the proposed experimental procedure follows:

1. The student was screened in order to determine whether he

or she met the criteria for overall reading skill and was currently

enrolled in a class for the learning disabled.

2. Samples of reading and comprehension performance were taken

to locate reading materials which correspond with the three levels

of oral reading performance for each subject.

3. Performances on reading and comprehension began and continued

for a minimum of five days, in all phases, where possible.

4. The first phase of the actual study involved collecting

baseline data on all three levels of material.

5. The second phase employed contingent pennies for correct

comprehension answers for each level of material at staggered time

intervals (see Figure 2). The schedule was one correct response to

one penny received. The teacher announced (prior to performance)

that pennies may be earned when this phase is in effect.

6. The third phase involved a return to baseline conditions.

In summary, subjects were located through the screening pro-

cedure. Baseline conditions were initiated and followed by the

contingent pennies condition for correct comprehension answers.

Baseline conditions followed as suggested by Bailey (1977) employing

a moving treatment multiple baseline design.








Sl -___
Compre. --nn L
Answers enn __
Material
E


- I -


D


S2


D




E



3 - I


E -




M-- -


S4 ____

E




D


-- J


DM




D


D








E --


Figure 2
Experimental Design


- - -[ -











Experimental Design


The design is a moving treatment, multiple baseline (Bailey, 1977)

with elements of the multi-element baseline (Sidman, 1960) contrasting

the effects of reinforcement contingencies on comprehension per-

formance with three levels of reading materials (see Figure 2).

Academic behaviors are often irreversible and the moving treatment

baseline design is recommended for this type of responding (Bailey,

1977). Sequence effects may be examined across subjects by the

use of all six possible arrangements. The likelihood that improvement

in comprehension for one material will spill-over to another is high,

since the movements may constitute a response class. For this reason,

additional discriminative stimuli concerning the availability of

pennies will be used to specify which performances and the particular

materials being reinforced.

Reinforcement is only one and not the variable of primary interest

in this study. The effects of reinforcement have demonstrated

generality. .The main interest is in the effects of material difficulty

upon responding. Within each subject there were several opportunities

to observe the differential effects of reinforcement for the levels of

material difficulty. Interventions for each material correspond in

time with a baseline and an intervention period for the other materials.

The effects of reinforcement may be compared with two other cases of

responding at the two other levels of material difficulty (see

Appendix C for an explanation of the design notation system).











Dependent and Independent Variables

Reading comprehension is the class of the dependent variable

included in this study. The first dimension of interest is type

and number of responses to comprehension questions (correct and

error). The second dimension of interest is response latency.

Response latency is defined as the time from the end of the compre-

hension question presentation to the beginning of the answer. The

independent variable is oral reading rate (correct and error).


Data Collection

The data were collected during daily experimental sessions of 20

to 30 minutes in length. The subject's responses were recorded on

the experimenter's record sheet from audiotape. Both frequency of

oral reading, correct and error, as well as comprehension response

latency were recorded to the nearest tenth of one second.


Data Analysis

Visual inspection was the primary method of data analysis. Each

of the dependent variables was plotted as a function of time (consecu-

tive school days). Percentage correct comprehension answers was

plotted. Oral reading data and mean comprehension answer latencies

were plotted against a logrithmic scale. Oral reading data were also

plotted as a function of both comprehension answer percentage correct

and mean latencies. Aids were employed to summarize the data as need

was indicated.















CHAPTER IV

RESULTS



The primary form of analysis is visual inspection of graphic

material. Six types of data are examined for each of the six

subjects. The first two are directed at answering experimental

Question One concerning the effects of contingencies of reinforce-

ment upon comprehension. The final four forms of data explore

Question Two concerning the relationship between oral reading and

comprehension. The types are as follows:

(a) a multiple baseline design across reading material

levels for consecutive school days with presentation

of pennies as the independent variable and comprehen-

sion answer percentage as the dependent one (Appendix G);

(b) the same arrangement except the dependent variable is

response latency (Appendi.x G);

(c) the mean and range of words read correctly per minute

for each comprehension answer percentage correct across

all three reading levels;

(d) the mean and range of words read incorrectly per minute

for each comprehension answer percentage correct across

all three reading levels;











(e) mean comprehension answer latency as a function of

words read correctly per minute across all three

levels; and

(f) mean comprehension answer latency as a function of

words read incorrectly per minute across all three

levels.

Each subject's data were analyzed individually, except in instances

where subjects' performances are similar enough that nothing is lost

by summarizing across subjects.



The Effects of Contingent Pennies Upon Comorehension
Answer Percentage Correct


For Subject 1, in the middle level material, pennies appear to

display a relationship with comprehension answers. A major problem

with Subject 1 was her frequent absences. Due to this problem her

baseline and treatment phases had fewer data points that would be

expected given the number of school days involved (see Figure 3 Appendix G).

Truancy had been a chronic problem for this subject and she was almost

dropped from the study during the baseline stage. During the inter-

vention stage her attendance improved considerably from three days

during the first 15 day period to eight days during the second.

The one demonstration of this relationship between pennies and

comprehension percentage is found in the intervention phase compared

to the initial baseline responding. The comprehension percentages go

from a median of zero and range from zero to 20 (baseline) to a median

of 20 and range of zero to 60 percent correct (treatment). In both











the easy and difficult material there is a lack of independence

with treatment effects apparently carrying over from one contingency

and material to the other. In the initial baseline in the easy

material having two data points make any conclusions impossible.

The final baselines for the middle and difficult material are

obviously short and do not allow for any conclusions to be drawn.

Data from Subject 2 was difficult to interpret since he frequently

scored 100 percent correct during his baseline phases (see Figure 4

Appendix G). For this reason the only clear-cut effects of the pennies

condition are seen in the difficult level reading material. Here in

comparing the initial baseline to the intervention phase both the

median and range improved. The median went from 60 percent correct to

100 percent correct while the range went from 40 to 100 percent to 60

to 100 percent correct. The final baseline phase showed a slight drop

in median percent correct to 90 percent correct.

Subject 3's data showed an increase in comprehension percentages

for the difficult level material for the pennies phase over both the

initial baseline and the final baseline (see Figure 5 Appendix G). The

median for the initial baseline is 45 percent correct while the inter-

vention median is 60 percent correct and the final intervention is 50

percent correct. Data for both the easy and middle level reading

material are inconclusive.

Data for Subject 4 indicate that the pennies condition increased

reading comprehension percentage correct for the easy material (see

Figure 6 Appendix G), although the final baseline responding is a contin-

uation of the higher level percentage from the previous pennies phase. The











percentages correct go grom a median of 80 percent correct (initial

baseline) to 100 percent correct and the range goes from 60 to 100

percent correct (initial baseline) to consistently 100 percent

correct. There are signs of a lack of independence for both the

middle and difficult material which could account for the responding

in these materials. The data from the middle level material show

signs of the pennies condition being effective in terms of an increase

in median percentage correct of zero (initial baseline) to 20 (inter-

vention); however, the ranges overlap too much to consider these

effects orderly.

Subject 5's data for the easy material showed orderly increases

for the pennies contingency over the initial baseline both in a higher

median (100 percent versus 80 percent) and range (80 to 100 percent

versus 40 to 100 percent) (see Figure 7 Appendix G). The final baseline

evidenced a return to the higher variability of the initial baseline.

For both of the other material levels the pennies condition appeared to

depress responding.

In the case of Subject 6, the pennies condition appeared to be

effective for all three levels of material although the degree to which

the data are convincing leaves open alternative explanations (see Figure

8 Appendix G). With the data from the difficult material the median for

the pennies condition increases over the initial baseline (zero to 20 per-

cent correct), but the range of the initial baseline data encompasses

that of the pennies condition data. The final baseline is of no help

since it virtually mirrors the data from the pennies condition. For

the middle material the initial baseline is unstable making it difficult











to draw definite conclusions. However, the increasing trend at day

25 and 26 would appear to indicate some effects of the contingency

although they are delayed. The data for the easymaterial tend to

show evidence of effects of the pennies condition with an increase

in the median from 90 percent in the initial baseline to 100 percent

in the treatment phase. Again, the ranges overlap and the difference

in the median may be due to the intervention phase being shorter than

the baseline phase. A similar sequence of data points could be

extracted from the initial baseline reducing confidence in the treat-

ment effects.



The Effects of the Pennies Contingency Upon Response
Latency for Three Levels of Reading Materials


Subject l's mean response latency showed increased variability

for both the easy and middle level materials during the pennies condi-

tion, while the responses in the difficult material evidenced wide

variability throughout both conditions. There were no major trends

other than the increased ranges for the easy and middle level

materials (see Figure 9 Appendix H).

In Subject 2, the mean response latency for the difficult and

easy materials had an accelerating trend during the pennies phase

over the initial and final baselines. The middle level materials'

mean response latencies appeared to be unaffected by the pennies condi-

tion (see Figure 10 Appendix H).

For Subject 3 the pennies condition had little or no effect upon

the mean response latency although there is some evidence of a










decreasing trend for the middle materials. The presence of only three

data points in this condition, however, makes interpretation

hazardous (see Figure 11 Appendix H).

Subject 4's pattern of mean response latencies for the easy material

during the pennies condition was elevated as compared to the initial

baseline's decreasing trend and the final baseline's lower level of

responding (see Figure 12 Appendix H). The mean latencies for the

difficult material showed a high degree of variability through all

three phases to the extent of making any comparison's unwarranted.

For the middle material the main effect of the pennies condition is

increased variability. There is also an accelerating trend in the

pennies condition versus the initial baseline. The final baseline

looks like a continuation of the pennies condition.

For Subject 5 the effects of the pennies condition upon mean

response latency are only evident in the difficult material. Here

the responding during the pennies condition shows increased variability

and an accelerating trend over both baselines (see Figure 13 Appendix H).

In Subject 6, there are only two notable effects of the pennies

condition upon latency in the middle and difficult level materials.

In the middle materials the range and variability increase in the

pennies condition over the initial baseline. The final baseline's

three data points continue this trend. For the difficult material

there is an accelerating trend during the pennies condition over the

initial baseline condition which continues through the final baseline

(see Figure 14 Appendix H).











For the remaining relationships the data for the three reading

levels are combined. This seems logical since although the words per

minute correct and error for the three readers were distinct in

terms of median they overlapped in range. Thus they formed more of

a continuum than disparate points (see Table 2).



The Mean and Range of Words Said Correctly Per Minute for
Each Comprehension Answer Percentage
Correct Across All Reading Levels


In Subject 1 the median words read correctly remains around 50

until comprehension answer percentage correct exceeds 40 percent.

Then, it increases as comprehension increases to about 65 to 70 words

read correctly per minute from the levels of 60 to 100 percentage

correct comprehension answers (see Figure 15). The range also shows

a consistent decline as correct comprehension answer percentage increases.

This, however, is due to the subject's high number of days with zero

to 20 percent comprehension.

SubjectTwo only had one day's data in which his comprehension

answer percentage dropped below 60 percent. The vast majority of his

responses were correct leading to the large range for words said

correctly in the 100 percent correct comprehension category (see

Figure 16). Nevertheless, the median words said correctly increases

sharply from the seven days data where comprehension was at 60 percent

correct to the 12 days in which it was at 80 percent. The medians go

from 26 words per minute correct to 38. The 46 days where comprehension

is at 100 percent had a median of 41 words per minute correct. This

median change in rate represents almost a 40 percent increase during

the 60 versus 100 percent comprehension performance.












Table 3

Median and Range of Words Said Correctly and Incorrectly Per Minute


SUBJECT BOOK E BOOK M BOOK D

Correct Error Correct Error Correct Error
Median Range Median Range Median Range Median Range Median Range Median Range


1 60 41-90 10 5-22 53 30-67 10 8-18 44.5 31-54 16 11-29

2 54 40-85 1 0- 5 35.5 21-65 4 1- 9 31 15-39 4 3- 6

3 59 45-88 2.5 0- 6 37 32-53 4 2- 6 25.5 20-43 7 4- 8

4 78 38-96 4 0-10 50.5 25-82 5.5 2-10 27 18-33 7 5-11

5 47.5 37-76 6 2-17 38 16-54 12 4-24 36.5 20-46 12.5 7-25

6 101 72-138 5 0-10 70 46-107 11 5-28 64.5 44-83 12 5-31





































0 10 20 30 40 50 60 70 80 90 100

PERCENTAGE CORRECT COMPREHENSION ANSWERS

Figure 15

The Hedian and Range of Words Said Correctly for Each Level of Reading
Comprehension Answer Percentage Correct for Subject 1









































The Median and Range


PERCENTAGE CORRECT COMPREHENSION ANSWERS
Figure 16

of Words Said Correctly for Each Level of Reading Comprehension
Answer Percentage Correct for Subject 2











The third subject's relationship between words said correctly

and comprehension performance was less consistent due to the data

at the 40 percent comprehension level. Otherwise the words per

minute correct remain in the middle 20s until the sharp increase

to 38.5 at 80 percent correct and 54.5 at the 100 percent comprehen-

sion levels. The median rate here more than doubles between 60 per-

centage correct comprehension and 100 percent (see Figure 17).

Subject 4 demonstrated a near perfect positive relationship between

words said correctly and comprehension answer performance. Medians

remained in the middle to upper 20s as comprehension ranged between

zero and 40 percent correct. At 60 percent comprehension the median

increased to 41.5 words per minute correct and 55 words per minute

was the median at 80 percent comprehension. Finally, the median words

per minute correct reached 65.5 at 100 percent comprehension (see

Figure 18). This increase is about a ten word-per-minute increase

per 20 percent comprehension improvement at the 40, 60, 80, and 100

percent comprehension levels.

Subject 5's data showed a similar pattern as Subject 4's smaller

increases per comprehension improvement. For each of the 20 percent

comprehension level improvements this subject increased three to four

words read correctly per minute except between 60 and 80 percent

comprehension where the change was only one word-per-minute (see

Figure 19).

In the case of Subject 6, the data indicate the only major incon-

sistency in the relationship between words said correctly and compre-

hension answer percentage. This is found where the subject's median

















S100-
z 90-
E BO *

so
a 70 *



0 10 20 30 40 50 0 70 0 1040


Answer Percentae Correct for Subject 3

2r = .62421

p = .00001

10

0 10 20 30 40 50 sO 70 80 90 100

PERCENTAGE CORRECT COMPREHENSION ANSWERS
Figure 17

The Median and Range of Words Said Correctly for Each Level of Reading Comprehension
Answer Percentage Correct for Subject 3














:UU -,I_ _


i-
Sloo
z so
S 80 *
100

so 7
". 60
S50 -
S 40




o 20 r = .73917

o p = .00001


10 -

0 10 20 30 40 50 60 70 0O 90 100

PERCENTAGE CORRECT COMPREHENSION ANSWERS

Figure 18


The Median and Range of Words Said Correctly for Each Level of Peading Comprehension
Answer Percentage Correct for Subject 4


















S100
90
2 80
a 70
60
50


0






p = .00005
10


0 10 20 30 40 50 60 70 80 90 100

PERCENTAGE CORRECT COMPREHENSION ANSWERS

Figure 19

The Median and Range of Words Said Correctly for EAch Level of Reading Comprehension
Answer Percentage Correct for Subject 5











words said correctly is actually higher for 20 percent correct

comprehension (72) then it is for 40 percent comprehension (65).

Additionally, the one data point at the 25 percent correct compre-

hension level is higher than that of the median 40 percent figure.

From 40 percent comprehension to 100 percent comprehension the rela-

tionship is near perfect with nine to 20 word per minute increases

for each 20 percent improvement. The data indicate two separate dis-

tributions with zero to 25 percent comprehension representing one and

40 to 100 percent comprehension as the second (see Figure 20).


Mean and Range of Words Said Incorrectly Per Minute for
Each Comprehension Answer Percentage Correct
Across all Three Reading Levels


Again with Subject 1, her comprehension answer percentage correct

was very low with most of the data falling in the zero to 20 percent

correct category. For all other categories there were three or less

days of data. This may explain the absence of a simple negative rela-

tionship between errors per minute and comprehension performance.

Discounting the six data points at 60 and 80 percent comprehension,

the remaining errors per minute show a consistent decrease from 13 to

7.5 as the comprehension answer percentage increases (see Figure 21).

In Subject 2 (with the majority of his data in the 100 percent

category) the median words said incorrectly drops from five at 60

percent comprehension to four at 80 percent comprehension and 3.5 at

100 percent correct comprehension (see Figure 22). Again the ranges

are probably more indicative of the majority of the subject's data being

located in the 100 percent comprehension category than more meaningful

conditions.










200-v


r = .72788

p = .00001


0 10 20 30 40 50 80 70 80 90 100


PERCENTAGE CORRECT COMPREHENSION ANSWERS

Figure 20

Words Said Correctly for Each Level of Reading
Answer Percentage Correct for Subject 6


I-
S100-

S80



8 so
S 70

S *
IL 60






o
w 40 -
cc
30 *
3
()
Cn 20

0


The Niedian and Range of


Comprehension












40 -

30-


20







7- *
6
0 59
M r = -.18734
cc 4
0
Sp = .13343

0
~ 2
0
c-





0 10 20 30 '. 50 60 70 80 90 100


PERCENTAGE COQ7E T, IywPREHENSION ANSWERS

Finiure :'1


The Miedian and Range of Words Said Incorrectly for Each Level of Reading Comprehension
Answer Percentage Correct for Subject 1






































0 10 20 30 40 50 60 70 80 90 100


PERCENTAGE CORRECT COMPREHENSION ANSWERS

Figure 22

The Median and Range of Words Said Incorrectly for Each Level of Reading comprehension
Answer Percentage Correct for Subject 2











Subject 3's data evidence a negative relationship between

errors-per-minute and correct comprehension answer percentage.

This trend is consistent with only minimal variability and levels

off at 80 percent correct comprehension (see Figure 23). Except

for the one data point at 33.3 percent the data show a drop of

between .5 errors per minute and 2.5 errors per minute for each 20

percent increase up to 80 percent comprehension where they level off

at three.

In the case of Subject 4 the negative relationship is direct

except for the three data points at 40 percent correct comprehension.

All other 20 percent comprehension intervals have five or more data

points. Measures of central tendency are obviously more useful with

a large sample of data. Discounting the 40 percent data the median

errors per minute drop from .5 to 1.5 per 20 percent interval until

they reach 80 percent correct where they again level off this time

at five (see Figure 24).

Subject 5's mean errors per minute show a definite negative rela-

tionship to comprehension answer percentage with the major changes

coming at 60, 80, and 100 percent comprehension levels. For zero

and 20 percent comprehension the errors per minute remain roughly the

same. Then, for 40 percent the errors drop to roughly the same as

the errors per minute for 60 percent comprehension. After 60 percent

the median errors per minute drop three at 80 percent and drop four

at 100 percent correct comprehension to end up at five (see Figure 25).

Subject 6's words said incorrectly per minute show a negative

relationship to correct comprehension percentage but this relationship














40 -

30 -


20


I.-
z 10-

0o
a 5
7





z

0 2

c r = -.56781
0

p = .00001


0 10 20 30 40 50 60 70 80 90 100


PERCENTAGE CORRECT COMPREHENSION ANSWERS

Fioure 23

The Median and Range of Uords Said Incorrectly for Each Level of Readinq Comprehension
Answer Percentage Correct for Subject 3












40

30


20-


UJ


a 9
LU



P. 7

5
UJ
Er 4
0
_z i
Q
^ 2 I r = -.51013

g p = .00003


1 -
0





0 10 20 30 40 50 60 70 80 90 100


PERCENTAGE CORRECT COMPREHENSION ANSWERS
Figure 24

The Median and Range of !lords Said incorrectly for Each Level of Reading Comprehension
Answer Percentage Correct for Subject 4



















r = -.54630
S= .00001


0 10 20 30 40 50 60 70 80 90 100

PERCENTAGE CORRECT COMPREHENSION ANSWERS
Figure 25


The Median and Range


of Words Said Incorrectly for Each Level of Reading Comprehension
Answer Percentage Correct for Subject 5


i\











is less than straightforward. The most aberrant data is again the

single point at 25 percent comprehension. The medians for words said

incorrectly per minute in the zero to 60 percent comprehension

intervals vacilate between 10 and 12.5. Then, from 60 to 80 compre-

hension percentage the mean drops from 11 to 5 and back up to 7 at

100 percent (see Figure 26). This subject's inconsistencies between

errors per minute and correct comprehension may be related to the fact

that most of his errors were caused by substituting articles synonymous

with the words in the passage and would be less likely to reduce

comprehension performance.



Mean Comprehension Answer Latency as a Function of Words
Read Correctly Per Minute Across All Three Reading Levels


In all six subjects' scatterplot of mean comprehension latency

compared to words said correctly there was a very strong negative

relationship. That is as words said correctly per minute increased,

the mean response latencies for the comprehension answers decreased

(see Figures 27-32).



Mean Comprehension Answer Latency as a Function of
Words Said Incorrectly Per Minute


Again, all six subjects' data showed the same pattern. In this

case there was a positive relationship between words said incorrectly

(or errors) per minute and mean response latency. Simply as the

frequency of errors increased the mean comprehension answer latency













40-

30


20




S 10
"5 9

0 8

UJ
7
c 4
0
Z 3


< 2 r = -.59324

c p = .00001





0 10 20 30 40 50 60 70 80 90 100


PERCENTAGE CORRECT COMPREHENSION ANSWERS
Figure 26

The Median and Range of Words Said Incorrectly for Each Level of Reading Comprehension
Answer Percentaoe Correct for Subject 6














33,00 .00 95,00 03.o0 57.00 63,00 00.o0 ,00 6o .00 07.00
. . I .. R7
"* ' * -- .- . -.---.-----,-- -_-----.------- -- ----








*.o oo







-oD
u.oo



5.00 .00
.0 0 .... .00oo















3.0 0* 3.. 0
3 -





I. 1o I I I,
.00 9 2.0oo



l' -- -L I* D 0







0 FL01rtr v l1L5 37 Oxt.lu'11 n U-t hl 0 nlQl ll VAtLU L 0



Figure 27


A Scattergram with a Horizontal Axis of Uords Said Correctly Per Mlinute

and a Vertical Axis of [lean Comprehension Answer Response Latency for Subject 1
and a Vertical Axis of rlean Comprehension Answer R~esoonse Latency for Subject 1











ILuE oOIOt" I H09I0L 09. 9 0o/00/251


19.00 27.00 3:. 00o 3.00 91,00 5.900 .00O 75.OO B1..0 91.00

10.00 10.00



9.00 -- -----0
.0o 0 9.00



7,00 .oo












s.oo .o 500
6.00 .00




..oo I ,oo



.00 .. .. .. 00




2.00 + 0 2.00





S- -9 9 9 -_


+----+----------------------------- -------------+----+----+---- ----- --
05.00 23.o0 0.oo 390.00 4 .00 .00 0.,00 o ..O 09.00 7o v 5.oo00
IoIu C 0/0009/25. 10.27-<9. 90I0L 0
osTAnisucs ..
0 CO9REL0 IIO 1 l1- 7BB '. S0U0.0LL -22930 500 0 .SI lFlC0N/ 0000- --
0 ST[5 ERR OF E0T .391B0 0ITE0r0E0 0A) .9000 ... s1, r00n0 Or 0 .00I
o0 lCIIC0AL0 b- .0000-
0 000LTElB ALUE9S -& EXCLUIEIi UALUO S- 0 9h5100 IO YUEs 0



Figure 28



A Scattergram with a Horizontal Axis of Words Said Correctly Per Minute and a

Vertical Axis of Iean Comprehension Answer Response Latency for Subject 2













L t ak







9.UO
lo.oo




7.00



.oo0














5,oo a


+ _.00
_.oo






?.oo


_---_ .------------.----,- -,.--------- ....
0.00 .oo 00 0. .00 O 00.o ).00 100.00
oUIooPoOO. 03.05.00. 0.00,0


0500oIo0.00 OooO'- oIH SOIFTrAO 0
00000,0o0.0 00 00000IY 511,.00 000tl 01
Ll l I -.oooSS SlI*U ,.o0.5o


L.tL.Ul U0t gI LS- 0


Figure 29


A Scattergram with a Horizontal Axis of Words Said Correctly Per Minute and a

Vertical Axis of Mean Comprehension Answer Response Latency for Subject 3


0 51 LL I0 L -

0 OLOI ISOI 00LIL50




24.00 3.00 .00 .0 .000 0.00 6,oo 0 .00 00 0o.0oo 0o. 6.00
----------t ----*---------------- ---- -----1---- ------- -----------






.. . . .00o

la.oo


6.1o


4.oo


I

]l


I

I



2 _00 - a -- -


.oo
. . I . .


a as


12-


OOOJI
-,6400O

*,9.0000


nrtlus untuth











4IL 00ME (CREa0rION ATE 30/09/23.I
SCArreI0554 OF 0 (HOUNI L00T54C
-2.00 00.00 30.00 46.00 90.0 _'.00 +0.00 7a. 0 6-.00 -4.00




.00 + -.- ------------ ---------





4 .00 .00.
-. .. .













0.00 + 4 0.00










0 00









0.00 1. 5 .0 .. +,. 4 3 a I.00U o







Fi ure 30


A Scattergram with a Horizontal Axis of Words Said Correctly Per Minute and a
Vertical Axis of Mean Comprehension Answer Response Latency for Subject 4
04 .i0









- - - - - - -.. . .. ..
0 oob --0 4 I0 .0 oo








ost--~-t ~---t ----- .-- -----






Fioure 30

A Scattergram with a Horizontal Axis of Words Said Correctly Per Minute and a
Vertical Axis of Wean Comprehension Answer Response Latency for Subject 4










1sL


1T0E-52 001*00 Aa0 ACoOSSI o0u3. -0 -

---- -- .- ---- --- -- -- A-- -------




i,00 o.oo



.o +o

0.00 3 A0.00










..00.. A:oo



.00 4 4.00
.oo 4.0oo





I _ I_

.00 E5 I 2 I s00
T I 9 2..









1 i ou ? i L.. .. --- E. .








Figure 31
----------------



A 0I.1.5.ALLT IOA (0- -.9250' I. SIOl I, 300o lbOl5 Z01 0979 0 ,0003- -
0 S1t. (rS 33 3SF .,S7o 1 -00c0rr SIll 31,01* 0f A -,30oo[
o 510,5333IlL0 L A ,00001 si O1L LI -Fl .00- 5 ot 0... l l OF 3 .0003,

0 LXoLIIEL I 00LI00- 0 M*05(I.2 V2213 0


Figure 31


A Scattergram with a Horizontal Axis of Words Said Correctly Per Minute and a

Vertical Axis of Mean Comprehension Answer Response Latency for Subject 5












ISUb 8

\LE ONILME (0C0 0ID0 1T9E 0 600/0/2. 0.

0.00 5. .00 6..00 09.00 09.00 99.00 0o.oo 110.00 20.00 00.oo
--------------------------- -------- -------+-- ------ ------.
10,00 +






9,00 .00















..000




9.00 0 0 .00


.0 00
8s a



























S .00
0 000
..,oo
:.oo '
m --
*0
1 .. . . o


.000 0



S I .-----------




O-i T 1I CS.' 0 15,
0 .ll Sl- 0 0 l.0006


---- -- 0--0----- --.---
71.0 ,oo 0 *roo v.ao


___ __--, ----.----_--_ ----,----,-------------
09.00.1 00. IJ000 0

30200e15 00000 0.6 00

9,211(' 000106 III .1rt .0010'l
,l1, L 0 l ,11 "- .00000


Figure 32



A Scattergram with a Horizontal Axis of Words Said Correctly Per Minute and a

Vertical Axis of Mean Comprehension Answer Response Latency for Subject 6







66



increased. The relationship here varied from a relatively weak one

in Subjects 1 and 2 to a strong one for the remaining four subjects

(see Figures 33-38).














Fld r N* IAm rtlr n II.AL 8 0/09/2B t .1


7.00 I I10.00 .00 23.00 00 1 0.0 43 .00
------ -- -. -- ------- ---- . ------










-0~




o.oo 2
7.00

























0 ,00 0 00










01.00 1. 0 0 00 O oo
0,00 0,00 13,00_ .. 17.00 1 >00 ,0000 09.0300o o, 000 .0 .0 0 .0. 10



0 i , ..1. .0 2 SINIFIC CE
I
ID






























05 S.O01I



!I~LII* 11 118 IO I*l 01' STH TLN Ll


Figure 33




A Scattergram with a Horizontal Axis of Words Said Incorrectly Per Minute and a

Vertical Axis of Mean Comprehension Answer Response Latency for Subject 1


1100




IDL)O




s(ll)




800




I~L)




DO




LO~




0(1


ly)















Ill, NUN01 I0000lla0 0l09 OlO0/..23.


1o.00 3.00 .0(1 7.00 .00 l .01, 13.00 I.0 1o7.00 I. o00






9.00 -



0.00 -


7.o0






1
6.00




5.00




r.00









3.00
o
I
I

t.o

I



I ,




I o


+ - 7- .00



+ 6.00



S ,oo






















0





0 92. . .. . ... .
--~----~--~. -- .---*---r- co





6,oo ~ ~ R/Y/I ? ,oo .oo l oo i o so s Co oo


.5.. C

ostaIIsrlcs,,
0 CO 1-EL1] O r)-
I S1 l 1 ST -


.0J,4 IN0IChLFI (0 ) 1.09711 SllN LhFD OF A
.IOOUl El01U ) -o J 00. Sll0, c,1i,, 1 I -

66, 0 >I IIIII l VALUEFS- 0 MIbSING V0LI* t -


Finure 34


_ .0f033.
.0^/iUl


A Scattergram with a Horizontal Axis of I'ords Said Incorrectly Per NIinute and a

Vertical Axis of Mean Comprehension Answer Response Latency for Subject 2


-----------------I------
2.00 1.00O











IL 000E ICRE o ,T TO 9I.15 9O/O l2.
50900055099 LI CA'OOI II LA
,0o .00 5. 00 .0 I9.00 1 .00 I0oo ,o L.OO I.0'uo 10.4o
.4-----------. -----------------_ ___- ___-- .___-_-- -___ --____-_ _+.----- .--- ----- ----- ----- --------
lo-.. I oI uo



7.00 I 0.0




a I .oo







7.00 I 9 0.00

















90 02











ISUb R 0009/2J, 13.05.19. PAuE o
o I



2.. S. -III


1T---aTICS.. -I- ----- - _-- -r

0C0 ,LoI o.9- .10 SoS .40 1 SIUtI0ICANCC 0 Do0O6


0 OI LoD Vo LES 9 LOLIOI II. VLUES- o 0'. I0 vatIS 0


Figure 35



A Scattergram with a Horizontal Axis of Words Said Incorrectly Per i'inuta rad a

Vertical Axis of Mean Comprehension Answer Response Latency for Subject 3
i















FILE NONMlL ICrFATlON [AT1E BO/9/23.
scArltl oaM or Iouws LAT lckob1 s thk

1 0 1.-- 0---1 ------t------ ----l --- 7 .... *--. Ll l---.------------. .--- -- .---*-
0000.00



0.00 + --



0.00 B.00
iosoo











.00 .oo



0.00 + ..00








.oo00 + ..00













1.00 1.2.00
00oo











------------ --------------------------- ---- --------------- ----------- ---
0 2.00 40.00 .00 0.00 10.00 12.00 .00 1.00 0.00 o0.00

STATISTICSS.
o.oC






11 X




1505 110 0. 5

0 C00f00REL0 TIN 0 ( >)- .37000 S0U0 E0I .13 4 SI ,NIFI0 1CAN00 E R -- -00270 -
0520 71 (52 2.02200 12211110 (01 .I2.'010 5l, II.), I r 0 0
0 1IoItlI0,L, A.ooo0:. bls,11F 101 -.l .'00 II. lh 00 iI .01,000

o 1002 Il 0nUIs 0 55 20C1,1Il VaLItC- 0 5l1SSNb V2 ,UfS -


Figure 36



A Scattergram with a Horizontal Axis of Words Said Incorrectly Per Minute and a

Vertical Axis of Mean Comprehension Answer Response Latency for Subject 4











1SL

Sc TTarr h"m of (h1l + 4 AT
4.01U 0.00 2.00 16.00 20.00 24,00 20.00 32.00 36.00 40.0o
o10.00 4 0 o.o










7.oo aoo0
2 2
2 2- i ,





.+00 00







5.00 4 I- --











.OO o ..O
5 .00 5.00










.00 -- . I- -0 00







20 0 oo 0-0 70 ExC 0 0L1 E- -I 1 f -oo 0









Figure 37
-- ^- -- t-^S :; ''" f ~~ J- t --t- ~ -" - ~ ;.~~~i- ; --=- ----- --- 1 0


2 .01 10.0 014 1.0 2;.CO ZA.00 +.0.O 34,00 ls.00 4.00
1SL._ /0/23, I0I.J ii;_ IftE __

0 COhELAl ON l*1- .4743 *intllkel' ,2j04 ZIr.NIFICftN CE .0OCO?






Figure 37


A Scattergram with a Horizontal Axis of Words Said Incorrectly Per Minute and a

Vertical Axis of Mean Comprehension Answer Response Latency for Subject 5













C0 1 0rhlhA OF
2.00 6.00 10.00 14.00 i,0 200 00 2.00 30,00o I .oo a 0.0O
--- ------ ------ --------t ----l---t---- ---- ---------------
10,00 0 ____10.00



I.oa o.a_
9.00 9.00





0.00 9 .0 9.00





0.00 + + 6.00




.ao -- - .o





--0



9.00 .+ 9.00

3.00 ++< 0 3.00oo








S+-------- +---- +--------------------. ------ -- ---- -----_+-- ..+_...---- .-._ --- -.. -...-
0 o 00 9.00 12.00 .16,00 0.000 .2 6 00o_ 3 0.0 3.0o 00.00

1 -r*El- --CE
o l Fl0l L 1r N IR9 .003941 LtRE 0 .1 55 0 NF1 AC I -0 .00+12
0 >1, F0 I) LS0' -. I01 191+11 1+ -Ir t.2111r, lllI,0II10 0W i


0 1O1011 .11E 5.ICl2 X lldF li V0l ES- 0 MI9+1N VALUE0 0 0

Figure 38



A Scattergram with a Horizontal Axis of Words Said Incorrectly Per Minute and a

Vertical Axis of fean Comprehension Answer Response Latency for Subject 6
















CHAPTER V

DISCUSSION



The discussion chapter will be organized around the two questions

under investigation: (1) Is there a relationship between contingencies

of reinforcement and comprehensive performance? and (2) What is the

relationship between comprehension answers and oral reading rate?



Contingencies of Reinforcement and Comprehension


The answer to this question will be investigated by reviewing

the effects of a condition of contingent pennies upon two dependent

variables. The pennies condition responding will be compared to

initial and final baseline performances in terms of percentage of

correct comprehension answers and mean comprehension response latency

across the three material levels.


Percentage Correct Comprehension Answers

For five of the six subjects the pennies condition was effective

for only one of the material levels. In the case of Subject 6 there

were indications that the pennies were reinforcing for all three

levels, although the responding wasn't as orderly as would be necessary











to establish a strong relationship. The reading rates in which

the pennies showed a relationship corresponded with Lovitt and

Hansen's (1976b) suggested instructional levels for placing students

in readers. These guidelines were to locate a reader in which the

student read at a rate of 45-65 words per minute correct and 4-8 errors

per minute. The reading materials for which an effect was noted had

median reading rates correct or error which corresponded with these

suggested ranges and in one case both rate correct and error rate fit.

In the case of Subject 6, the reading performances were almost univer-

sally too high in rate correct and too high in errors to meet the

guidelines. This may in part account for his data showing inconclusive

yet suggestive relationships between the pennies condition and compre-

hension answer percentage correct.

A secondary explanation for the effects noted in the results

chapter has to do with the possibility that pennies are not an effec-

tive reinforcer for these subjects and perhaps the population in

general. Lahey (1980) replicated his original work recently using

nickels and has suggested that inflation may have led to a reduction

in the pervasiveness of pennies as reinforcers.


Mean Comprehensive Answer Latency

The two effects that were found in the pennies condition were

increased variability over the baseline phase and accelerating trends

in the pennies condition versus the baseline condition. Alba (1975)

noted the tendency for response latency to increase with incorrect

responses over correct ones. Therefore, the higher the percentage of

correct comprehension responses the lower the latencies expected.











This would account for the occasions when comprehension percentage

increased and response latency decreased. However, there seems to

be no relationship demonstrated between increased comprehension

percent and decreased latencies. There are also no apparent relation-

ships between changes in the two measures of any kind.


The Relationship Between Comprehension and
Oral Reading Rate


Four forms of data were analyzed to answer this question includ-

ing: (1) the mean and range of words read correctly per minute for

each comprehension answer percentage correct, (2) the mean and

range of words read incorrectly per minute for each comprehension

answer percentage correct, (3) the mean comprehension answer latency

as a function of words read correctly per minute, and (4) the mean

comprehension answer latency as a function of words read incorrectly

per minute.


Mean and Range of Words Read Correctly Per Minute by
Comprehension Answer Percentage Correct

In all six cases there was a strong positive relationship between

the median number of words read correctly and comprehension. Also in

many cases the ranges showed little overlap. This consistent finding

represents a very orderly correlation and strongly suggests that a

manipulation of oral reading rate correct would lead to a subsequent

increase in comprehension. A possible method for demonstrating further

the relation would be to examine the data by material levels. In

examining the figures in Appendix I it is apparent that different










reading levels were obtained (although there is some overlap of the

middle level with the other two levels). Moreover, rates and compre-

hension answers percentage correct were higher for easy material and

lower for difficult material (see Appendix I). The reading rates

correct and error formed more of a continuum than separate distri-

butions. For that reason these variables were combined for the

charting with comprehension answer percentage and latency.


The Mean and Range of Words Read Incorrectly Per Minute
for Each Comprehension Answer Percentage Correct

Errors per minute showed a negative relationship with comprehension

answer percentage correct. This relationship showed greater variability

in many of the subjects, however, in all subjects the relationship was

present. Subject 6 again showed the most variability to the extent

that the relationship could barely be discerned.


The Mean Comprehension Answer Response Latency and
Words Read Correctly Per Minute

In all six subjects there was a negative relationship between words

read correctly per minute and mean response latency. That is, as

rate correct increased, the response latency decreased.


Mean Comprehension Answer Latency and Errors Per Minute

In all six subjects there was a positive relationship between

errors per minute and latencies. That is as the number of errors

increased the latencies also increased. The relationship ranged

between a very strong one and a less convincing one.











Comparison of the Present Study With the Previous Research
on the Relationship Between Oral Reading Rate
and Reading Comprehension Performance


Only three previous investigations address directly the problem

of the relationship between oral reading rate and comprehension. These

are (1) Lovitt and Hansen (1976b), (2) Jenkins et al. (1978), and (3)

Roberts and Smith (1980). The results of each of these studies will

be compared with those of the present one. Each of these investigations

have been reviewed previously and will only be dealt with as they pertain

to the present study's results.

Lovitt and Hansen (1976a) used a contingency that involved improved

oral reading rate and comprehension concurrently. This procedure

appeared quite effective, but failed to address the question of whether

an improvement in oral reading would lead to an indirect increase in

comprehension. The present study suggests that this would occur

within certain ranges of words per minute correct and error.

Jenkins, Barksdale, and Clinton (1978) dealt more directly with

this transfer from one facet of reading (oral reading rate correct and

error) to another (comprehension answer percentage correct). They,

however, employed subjects who were already (prior to the study) reading

at levels that the present study suggest are nearly twice that

necessary for 100 percent comprehension. The minimal gains they

found in comprehension when oral reading rate was increased would only

be expected to be minimal.

Roberts and Smith (1980) also investigated this "transfer" effect

from oral reading rate improvement to reading comprehension percentage

improvement. The subjects in their study were reading at lower rates










correct and higher error rates than the children in the Jenkins et

al. (1978) study. The oral reading intervention should (according

to the present study) also lead to an increase in comprehension per-

formance. Unfortunately, experimental control was not demonstrated

and in fact the comprehension performances showed a relatively con-

sistent increase across all six phases of the investigation and in

many cases the highest comprehension percentages were found in base-

line phases.



Implications


The data from this study do not provide much conclusive evidence

concerning the effects of contingencies of reinforcement upon compre-

hension performance other than it appears that pennies are not the

pervasive reinforcer found in the Lahey et al. study. The question

concerning the relationship between oral reading rate correct and

error and reading comprehension has more convincing data. The rela-

tionship is apparently a moderately strong one as demonstrated in the

following conditions: (a) as oral reading rate increases reading

comprehension increases; (b) as the number of errors per minute increase

the reading comprehension decreases; (c) as the oral reading rate

increases the comprehension latency decreases; and, finally, (d)

as the error rate increases the comprehension latency also increases.

These findings are consistent across all six subjects. Thus, both oral

reading rate correct and error provide evidence of an empirical rela-

tionship with reading comprehension answer percentage correct and

latency. The strongest relationship is with oral reading rate correct










and both comprehension answer percentage correct and latencies.

With the finding precision teachers can be more confident that in

teaching oral reading skills they are not doing so at the expense

of comprehension. The rates that have been suggested as necessary

for proficiency seem to be well above those found correlating with

high levels of comprehension performance. Also there appear to be

individual variations in proficiency as measured by comprehension

performance.

Previous proponents of precision teaching have probably

suggested these higher rates be attained in a reader before pro-

gressing to the next level reader to foster success in the second

level reader. This is an important consideration. Instructional

times frequently becomes a factor with these students, however, some

combination of these two concerns is needed.



Limitations


The major limitation of this study is the age group of the

subjects. Wolking (1973), as well as Starlin and Starlin (1973),

have found that different oral reading rates are important at

different ages.



Future Research


The present study needs to be replicated systematically in the

following order of importance: (1) using the same procedures and age

level subjects by a different investigator; (2) the same procedures







80



need to be explored with different age groups of subjects; and

(3) parameters for the other paradigms in Figure 1 need to be

investigated to discover how to identify them as problem areas or

rule them out as the source of the breakdown in the sequence of

events called reading comprehension.
















APPENDIX A

STATE OF GEORGIA REGULATIONS AND PROCEDURES
FOR SPECIFIC LEARNING DISABILITIES



Specific Learning Disabilities

1. Definition

"Specific Learning Disability" means a disorder in one or more of

the basic psychological processes involved in understanding or

in using language, spoken or written, which may manifest itself

in an imperfect ability to listen, think, speak, read, write,

spell, or to do mathematical calculations. The term includes

such conditions as perceptual handicaps, brain injury, minimal

brain dysfunction, dyslexia and developmental aphasia. The term

does not include children who have learning problems which are

primarily the result of visual, hearing, or motor handicaps, of

mental retardation, of emotional disturbance, or of environmental,

cultural, or economic disadvantage.

2. Eligibility and Placement

A. Evaluation Team

As a minimum, the following shall constitute the Learning

Disabilities evaluation team.

1. A teacher certified in Specific Learning Disabilities.

If one is not available within the system, another person











qualified to conduct diagnostic evaluations relevant

to Learning Disabilities should be designated.

2. A qualified psychological examiner.

3. The child's regular teacher. If the child does not

have a regular teacher or is less than school age, a

person qualified by the state to teach a child of

equivalent age should be designated.

B. Evaluation Components

The following shall be included as a minimum in each initial

evaluation of a child identified as having a Specific Learning

Disability.

1. An audiometric and visual screening preceding the assess-

ments to determine adequacy of sensory acuity for

subsequent testing with appropriate accommodations or

follow-up if hearing and/or vision are found inadequate.

2. An individual evaluation conducted by a qualified psycho-

logical examiner.

3. To substantiate the identification of the deficit areas)

in Section C, Number 2 below, a minimum of two evaluation

instruments must be administered individually be trained

Learning Disabilities personnel or a specified member

of the evaluation team. Both formal and informal assess-

ment measures should be utilized.

4. An assessment of language skills administered by an appro-

priate specified member of the evaluation team.











5. Written samples of the student's relevant classroom

work and a statement of the student's academic and

social performance by the regular teacherss.

6. A written observation report of the student's relevant

performance in the regular classroom setting by an

evaluation team member other than the regular classroom

teacher.

C. Eligibility Criteria

1. When provided with learning experiences appropriate for the

student's age and ability, it is demonstrated that he or

she is not achieving commensurate with his or her age and

ability levels in one or more of the areas in Item 2 of

this section.

2. Based on the results of assessments required in Section B,

Evaluation Components, a severe discrepancy is demonstrated

between ability and achievement in one or more of the

following areas.

(a) Oral expression (e) Reading comprehension

(b) Listening comprehension (f) Mathematics calculation

(c) Written expression (g) Mathematics reasoning

(d) Basic reading skill

3. In determining expectancy, the following formula should be

utilized: 2CA + MA 5.2 = Expectancy
3

The team should consider expectancy on an individual basis

for each child, in light of the information from the

psychological evaluation on cognitive functioning, age of

the child, and significant school experiences.











4. In determining severe discrepancy between ability and

achievement, the following is to be utilized.

(a) 1st grade by performance on appropriate evaluative

measures and adaptive behavior in the classroom

(b) 2nd grade 1 year or more below expectancy

(c) 3rd and 4th grade 1 years or more below expectancy

(d) 5th and 6th grade 2 years or more below expectancy

(e) 7th and 8th grade 3 years or more below expectancy

(f) 9th grade and beyond 4 years or more below expectancy

5. It is determined that the severe discrepancy between ability

and achievement is not primarily the result of

(a) a visual, hearing, or motor handicap;

(b) mental retardation;

(c) emotional disturbance;

(d) environmental, cultural, or economic disadvantage.
















APPENDIX B

GUIDELINES FOR COMPREHENSION QUESTIONS



Who Questions

Who ? Example: "jumped"
(verb or verb phrase)

Who ? Example: "thinks there may be life on
(short phrase) other planets?"

What Questions

What happened during ? Example: "the lightning struck?"
(verb phrase)

What happened to ______? Example: "the crowd?"
(noun or noun phrase)

What did/does the do? Example: "elephant"
(noun or noun phrase)

When Questions

When did the story take place? (if the answer is provided directly in
the story)

When did/does __________ ___________?
(noun or noun phrase) (verb or verb phrase)

Example: "the flying saucers" "land?"

Where Questions

Where did the story take place? (if the answer is provided directly in
the story)

Where did/does ____________ ______________
(noun or noun phrase) (verb or verb phrase)

Example: "the gold" "fall?"






86



Where was/is ?
(noun or noun phrase) (verb or verb phrase)

Example: "the tomb" "found?"















APPENDIX C

DEFINITION OF TERMS



Criterion

A criterion is an aim or goal. A criterion is expressed in

terms of desired frequency of performance for a specific skill.


Frequency

Frequency is the number of cycles (responses) per unit of time

(minute). Frequency is synonymous, in this sense, with rate.


Level of Difficulty of Material

Level of difficulty of material is defined by the frequency at

which the student demonstrates performance of the material. Levels

of difficulty range from slow frequency of performance with many errors

(very difficult material) to fast frequency of performance with no

errors (easy material).


Comprehension

Comprehension is the correct answering of comprehension questions.

These questions include recall of factual material, sequence interpre-

tation and the determination of the main idea.


Response Class

A response class is a set of responses that have at least one

characteristic in common. These characteristics may involve the







88



movements themselves (all are eye-blinks) or the effects of these

movements on the environment (we may open a door by pushing against

it with out hand, foot, knee, or body but the result is the same on

the environment) (Whaley & Malott, 1971).
















APPENDIX D

NOTATION OF DESIGN ELEMENTS


This system, devised by Johnston and Pennypacker (1980) is

a graphic shorthand to communicate design elements. It allows more

flexibility than traditional arrangements.

The following symbols are used to denote the major features of

research designs:


A dotted line indicates measurement, although
the independent variable is not in contact with
the behavior.

A vertical line extending upward indicates the
beginning of a new experimental condition and
one extending downward indicates a return to
an old one.

A brace encloses the subject, behavior, and
setting.

Example:

Subject 1

Reading token
reinf. _I
Class

Variables which remain constant are not relabeled
for either the brace items or experimental
elements.


L







90



Subject 1

Reading tutoring
s token
Class reinf.



Subject 2








Thus, Subject 2 is also reading in class and merely
undergoes the tutoring first and token reinforcement
second.
















APPENDIX E

OBSERVER DIRECTIONS AND SAMPLE SCORING SHEET



Examiner Directions for Recording Reading and Comprehension Performances


1. Spend a few minutes each day establishing rapport with the student

in order to encourage optimal performances.

2. Turn to each designated story and explain to the child that he or she

will first read the story orally (as fast and as accurately as he or

she can), read it silently, and then answer questions about the story.

3. Start the recorder.

4. Place the book in front of the child face down and say "Ready, Go"

(as you turn it over).

5. After the child has finished reading the story orally, turn the

recorder off while it is read silently.

6. Start the recorder and begin asking the comprehension questions.

Be sure to record both the comprehension questions (read by the

examiner) and the answers (spoken by the student).

7. Repeat this procedure for each story (normally six stories per child,

two stories per book).

8. Whenever possible use a single side of a cassette for a day's per-

formance and always retain the same child's reading on a single

cassette.

9. Be sure to label each side of each cassette with child's name and

the date (month and day).







92



10. Turn in the cassettes as they are completed or as soon as

possible.

11. Try to develop as much consistency and predictability as possible

from day to day in procedure (instructions, seating arrangement,

etc.).

12. Do not allow students (subjects or others) to disrupt reading

sessions.

13. Try to avoid interfering with the normal operation of classes.

14. Establish and maintain good rapport with the teachers, principal,

and other school personnel.




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