Citation
Effects of oral reading rate and inflection on comprehension and its maintenance

Material Information

Title:
Effects of oral reading rate and inflection on comprehension and its maintenance
Creator:
Tenenbaum, Henry A. ( Henry Abraham ), 1951- ( Dissertant )
Wolking, William D. ( Thesis advisor )
Pennypacker, Henry S. ( Reviewer )
Sharp, Bert L. ( Reviewer )
Turner, Edward C. ( Reviewer )
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida
Publication Date:
Copyright Date:
1983
Language:
English
Physical Description:
xiv, 191 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Cloze procedure ( jstor )
Direct instruction ( jstor )
Language comprehension ( jstor )
Learning ( jstor )
Learning disabilities ( jstor )
Mental stimulation ( jstor )
Oral reading ( jstor )
Punctuation ( jstor )
Reading comprehension ( jstor )
Reading instruction ( jstor )
Counselor Education thesis Ph. D
Dissertations, Academic -- Counselor Education -- UF
Memory ( lcsh )
Oral reading ( lcsh )
Reading comprehension ( lcsh )
City of Orlando ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Abstract:
Rigorous experimental analysis of the effect of oral reading rate on comprehension has only recently been performed. There is still controversy over which rate of oral reading maximizes comprehension and retention. Inflection and how it interacts with rate of oral reading have received very little attention. This study sought to determine how oral reading rate, when approximating functional conversational speech (150 to 200 words/minute) and inflection, impacts comprehension and maintenance of comprehension. A single subject design was developed to determine how oral reading at 150 to 200 words/minute with inflection compared with oral reading a 1; 40 to 60 words/minute (instructional rate), wit! inflection on measures of comprehension and maintenance of comprehension. High oral reading rates without inflection were also compared with low oral reading rates without inflection to determine their effects on comprehension and maintenance of comprehension. The dependent variables were a free recall task, answers to 10 comprehension questions and written responses to a Cloze procedure. These occurred immediately following reading criteria, and at three and 10 days after criteria was reached. Six subjects were used in this study; two subjects were of high school age and reading below grade level and four subjects were in the third grade reading on grade level. For this study an ABCD design was used with four subjects and a CDAB design was used for two subjects so that any effect that sequence may have had could be determined. Also, the high rate conditions were yoked to the low rate conditions to keep the number of trials equal. The results confirmed that the combination of high oral reading rate with inflection (when reading approximates conversational speech), increased both the accuracy and speed of comprehension and its maintenance more than any of the other combinations. The combination of high oral reading without inflection was found to increase comprehension and maintenance of comprehension when compared with low oral reading rate with and without inflection. Also, inflection training in both the high and low rate oral reading conditions improved comprehension. The results of Experiment 2 systematically replicated the results of Experiment 1 across reading levels, reading passages, settings, and subjects.
Thesis:
Thesis (Ph. D.)--University Florida, 1983.
Bibliography:
Includes bibliographic references (leaves 129-136).
General Note:
Typescript.
General Note:
Vita
Statement of Responsibility:
by Henry A. Tenenbaum.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
000440923 ( alephbibnum )
11272134 ( oclc )
ACK1487 ( notis )

Downloads

This item has the following downloads:


Full Text














EFFECTS OF ORAL READING RATE AND INFLECTION
ON COMPREHENSION AND ITS MAINTENANCE









BY

HENRY A. TENENBAUM


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





UNIVERSITY OF FLORIDA

1983
































Copyright 1983

by

Henry A. Tenenbaum


















ACKNOWLEDGEMENTS


A doctoral dissertation and all the steps along the way to a

doctorate require one to commit much of his energy and lifestyle

toward this goal. For me, this dissertation was greater than any

one individual. Several people have contributed directly to my

dissertation and others have given me their encouragement and per-

spective that I needed to succeed.

My acknowledgements must begin with my parents, Jack and Bertha.

I will probably never know the extent of their suffering when their

lives were threatened and when they were separated from each other

during World War II. You would think they would be embittered people

who had given up their desire to be a constructive part of society.

Yet, when they escaped the torments that existed for them in Europe

during and after the War, they stood strong and proud to become

Americans. From their pride and eagerness to begin a new life, they

provided me with the devotion, love, and their constant nudging to

go on and to become the best of what I wanted to be. One may never

be able to provide parents such as mine with all the acknowledgements

due them. But, I will continue to live my life in a manner that will

continually nake then proud.












I must also acknowledge my sister, Sara, and my brother,

Theodore. Throughout my life they have always provided me with

another view and the necessary wisdom to make the best choices

concerning my career.

Probably one of the most difficult tasks throughout a doctoral

program is maintaining a fulfilling relationship with your spouse.

My wife, Sarah, never once strayed from her dedication and love for

me. The sacrifices that she made and her understanding of what was

needed to pursue a doctorate were an inspiration to maintain my

motivation and perseverence.

My dedication and interest in the behavioral sciences were not

shaped by accident. Dr. Robert E. Anderson gave me the opportunity

to work with children and encouraged me to continually increase my

skills on the graduate level. My friends, Kay Kaldor and John

Carrier, provided me with many of the fundamental principles that

were needed to work with children successfully. I will always be

grateful for the time and patience they have shown me.

The people who have contributed more directly to my doctorate

will be remembered. As life is, people come under different con-

tingencies and leave their friends. But comfort can be had when one

knows that whenever we meet or need to be close again, the likeli-

hood of its occurrence is great.

fly closest friend and the one who influenced me the most was

my chairman, Bill Wolking. Bill is a remarkable human being who

seems to thrive on friendship and his dedication to students. Bill












has given me unselfish support and the guidance to know what an

empirical science of behavior should entail. Bill was able to not

only maintain my interest in applied behavior analysis and precision

teaching, but to widen my interests and to develop my curiosity of

the world outside of the university halls. He is a magnificent

person and I will continue to seek his leadership throughout my

career.

In any doctoral program one becomes influenced by many people.

But, Hank Pennypacker was one of those people along with B. F.

Skinner whose writings were continua'ly used as a fundamental guide

to enhance my understanding of the principles involved in a science

of behavior. To him I will be grateful for the wisdom that he has

given me and to humankind.

Special thanks go to my other committee members Ed Turner

and Bert Sharp. As a committee member one is constantly being asked,

along with everything else, to review manuscripts, attend meetings,

give counseling, and to provide leadership as it is needed. Without

Ed and Bert my doctoral work could have been a much more difficult

task.

While working towards my doctorate others outside of my

committee went above and beyond to provide me with much needed

support. These people were typically part of the Special Education

Department. To these people I give my thanks. More specifically

my gratitude goes to Dr. Cecil Mercer who many times offered and gave

his guidance. Special thanks also go to Dr. Bob Algozzine who always












provided me with the humor, encouragement, and his attendance at

the meetings needed to complete my degree.

I also wish to express my thanks to many of my friends who I

have grown close to and who never failed to ask whether or not I

needed their help in some way. These people have already completed

their doctoral degrees or are pursuing doctorates themselves. My

appreciation goes to Bonnie Engel for the many years of encourage-

ment she gave me. My appreciation goes to Leonard Weiss who kept

me interested in my studies by offering his professional respect and

by reaching out to maintain a long friendship with me. Special

thanks go to Charles Hughes who helped me to express many ideas

and, whenever he could, reminded me that I needed to graduate.

Another colleague who always made me feel in charge and helped to

maintain my motivation was Sue Peterson. To her I send very warm

thanks. To all the people who I have not mentioned, but who influenced

my life in many subtle ways, I give my appreciation.

Lastly, my love goes to all the children past and future who may

never have the opportunity to pursue a doctorate. Through the child-

ren I have met, each one helped me to better understand human behavior.

They truly were my inspiration.


















TABLE OF CONTENTS



ACKNOWLEDGEMENTS .... . . . . ... iii

ABSTRACT . . . . . . . . . . xiii

CHAPTER i INTRODUCTION ..... . . . . .

Rationale . . . . . . . . . . .

Significance .... . . . .... . . . 4

Statement of the Problem ...... ....... .. . 5

Delimitations .. . . . . .. ..... 5

Definition of Terms .... . . . ....... 6

Summary .. . . . . . . . .. 7

CHAPTER II REVIEW! OF RELATED LITERATURE . . . . 9

The Nature of Reading Comorehension . .. . . 10

Reading Comprehension as a Multifaceted Skill ... 10

Listening and Reading Comprehension . . .... . 11

Reading Comprehension as Facts . . . ... . 12

Oral vs. Silent Reading: whichh Way Maximizes
Comprehension? . . . . . . . .. 12

Theories, Strategies, and Tactics oF Reading and
Comprehension .............. .... . 13

Traditional Teaching Methods .. ... . . 13

Recent Strategies and Tactics for Teaching Reading and
Comprehension .... . . . . . . 18

Summary ........... . ..... ... 26













Dependent Measures of Reading Comprehension ..

Reading Comprehension and Memory ..

Summary . . . . .

Oral Reading, Rate, Punctuation, and Comprehension

Summary . . . . .


Verbal Behavior ....

Experimental Analysis of Behaviom

Stimulus Control ...

Comprehension Questions and Free
Stimuli for Intraverbals . .

Summary . . . . .

CHAPTER III METHOD ...

Experimental Questions . .

Experiment 1 . . ..

Experiment 2 . . . . . .

Independent Variables . . .

Apparatus . . . .

Dependent Variables ...

Subjects . ..

Design . . .

Procedure . . ..

Pretraining Period . . .

Experimental Conditions .


r and B. F. Skinner



Recall as Discriminati






















. . . . . .


Data Analysis--Dependent Variables ..












CHAPTER IV RESULTS . . . . . .

Questions 1-6 Results for Experiments 1 and 2
Free Recall Total Performance . . . . . . .

Experiment 1 Results for Question One . . . . .

Question One Results for Experiment 2 . . . . .

Condition Two Results for Experiment 1 . . . .

Question Two Results for Experiment 2 . . . . .

Question Three Results for Experiment 1 . . . . .

Question Three Results for Experiment 2 . . . . .

Question Four Results for Experiment 1 . . . .

Question Four Results for Experiment 2 . . . . .

Question Five Results for Experiment 1 . . . .

Question Five Results for Experiment 2 . . . . .

Question Six Results for Experiment 1 . . . . .

Question Six Results for Experiment 2 . . . . .


Accuracy Results . . . .

Question One Results for Experiment 1

Question One Results for Experiment 2

Question Two Results for Experiment 1

Question Two Results for Experiment 2

Question Three Results for Experiment

Question Three Results for Experiment

Question Four Results for Experiment 1

Question Four Results for Experiment 2

Question Five Results for Experiment 1

Question Five Results for Experiment 2


67

67

69

70

71

71

74

75

75

76

76

76

77


. . . . 78

. . . . 78

. . . . 78

. . . . . 80

. . . . 80

. . . 81

. . . . . 82

. . . . 82

. . . . . 83

. . . . 83

. . . 84











Question Six Results for Experiment 1 . . . ... 84

Question Six Results for Experiment 2 . . . ... 85

Free Recall--Words Per Fact ... . . . . . 85

Summary Free Recall Questions 1 6: Total Performance,
Accuracy, and Wlords Per Fact . . . . . . . 87

Ten Comprehension Questions . . ......... .. 88

Question One Results for Experiment One . . . ... 88

Question One Results for Experiment Two . . . ... 88

Question Two Results for Experiment 1 . . . ... 91

Question Two Results for Experiment 2 . . . ... 91

Question Three Results for Experiment 1 . . . ... 92

Question Three Results for Experiment 2 . . . ... 93

Question Four Results for Experiment 1 . . . ... 93

Question Four Results for Experiment 2 . . . ... 93

Question Five Results for Experiment 1 . . . ... 94

Question Five Results for Experiment 2 . . . ... 94

Question Six Results for Experiment 1 . . . ... 95

Question Six Results for Experiment 2 . . . ... 95

Latency Measure Results . . . . . . . ... 96

Summary of Results: Dependent Measures of Comprehension . 96

Results Cloze Procedure--Total Performance . . . ... 98

Question One Results .. .. .. .. .. .... 98

Question Two Results . . . . . . . . . 99

Question Three Results .. . . . . . . 101

Question Four Results . . . . . . . . . 102

Question Five Results . . . . . . . . . 103

Question Six Results .. . . . . . . . 103











Cloze Procedure Results--Accuracy . . . . . . .

Question One Results . . . . . . . .

Question Two Results . . . . . . . . .

Question Three Results . . . . . . . . .

Question Four Results . . . . . . . . .

Question Five Results . . . . . . . . .

Question Six Results . . . . . . . . .

Results Cloze Procedure--Percent Correct . . . . .

Summary of Results--Cloze Procedure . ...

Summary of Results Experiments One and Two . . . . .

CHAPTER V DISCUSSION . . .. . . . . . .

Variability . . . . . . . . . . . .

Replications . . . . . . . . . . . .

Practical Implications . . . . . . . . .

Implications for the School Psychologist . . . . .

Generalizations . . . . . . . . . . .

Suggestions for Future Research . . . . . . .

REFERENCES . . . . . . . . . . .

APPENDIX A FREQUENCY MULTIPLIERS, TOTAL PERFORMANCE, FREE
RECALL . . . . . .

B ACCURACY RATIO MULTIPLIERS, FREE RECALL .

C FREQUENCY MULTIPLIERS, TOTAL PERFORMANCE
CLOZE PROCEDURE . .....

D ACCURACY RATIO MULTIPLIERS, CLOZE PROCEDURE

E 10 SAMPLE COMPREHENSION QUESTIONS, CLOZE
PROCEDURE . . . . .


104

104

106

106

107

107

108

109

109

111

114

118

120

124

124

125

126

129












APPENDIX F POST-EXPERIMENTAL INTERVIEWS . . . . .. 165

G RAW DATA . . . . . . . . 174

BIOGRAPHICAL SKETCH .. . . . . . . .. . 190















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



EFFECTS OF ORAL READING RATE AND INFLECTION ON
COMPREHENSION AND ITS MAINTENANCE

By

Henry A. Tenenbaum

August, 1983

Chairperson: William D. Wolking
Major Department: Counselor Education

Rigorous experimental analysis of the effect of oral reading

rate on comprehension has only recently been performed. There is

still controversy over which rate of oral reading maximizes compre-

hension and retention. Inflection and how it interacts with rate

of oral reading have received very little attention.

This study sought to determine how oral reading rate, when

approximating functional conversational speech (150 to 200 words/

minute) and inflection, impacts comprehension and maintenance of

comprehension. A single subject design was developed to determine

how oral reading at 150 to 200 words/minute with inflection compared

with oral reading at 40 to 60 words/minute (instructional rate), with

inflection on measures of comprehension and maintenance of comprehen-

sion. High oral reading rates without inflection were also compared

with low oral reading rates without inflection to determine their












effects on comprehension and maintenance of comprehension. The

dependent variables were a free recall task, answers to 10 compre-

hension questions and written responses to a Cloze procedure.

These occurred immediately following reading criteria, and at three

and 10 days after criteria was reached.

Six subjects were used in this study; two subjects were of high

school age and reading below grade level and four subjects were in

the third grade reading on grade level. For this study an ABCD

design was used with four subjects and a CDAB design was used for

two subjects so that any effect that sequence may have had could be

determined. Also, the high rate conditions were yoked to the low

rate conditions to keep the number of trials equal.

The results confirmed that the combination of high oral reading

rate with inflection (when reading approximates conversational

speech), increased both the accuracy and speed of comprehension and

its maintenance more than any of the other combinations. The combina-

tion of high oral reading without inflection was found to increase

comprehension and maintenance of comprehension when compared with

low oral reading rate with and without inflection. Also, inflection

training in both the high and low rate oral reading conditions

improved comprehension. The results of Experiment 2 systematically

replicated the results of Experiment 1 across reading levels, reading

passages, settings, and subjects.


















CHAPTER I

INTRODUCTION



This is a study of how inflection and oral reading rate affect

comprehension and maintenance of comprehension. The impact that

inflection may have on comprehension and maintenance of comprehen-

sion has not received extensive investigation (Weaver, Holmes,

Curtis, & Reynolds, 1970). Many reading experts agree that reading

with the correct inflection is an important skill (Durrell, 1949;

Heilman, 1967; Spache & Spache, 1977). The importance of reading

fluently has also been considered a major variable for developing

comprehension by only a few investigators (Lovitt & Hansen, 1976;

Smith, 1972). Yet, research reported has not addressed the relation-

ship between both levels of fluency and levels of inflection on

comprehension and maintenance of comprehension.



Rationale


There is a concern that there has been a paucity of research on

how to develop and improve reading comreenension (Armstrong, 1981;

Turner & Fey, 1977; Van Etten, 1978). The idea behind this investigation

came forth as a result of discussing the concepts in verbal behavior












(Skinner, 1957) with others. In beginning an investigation in

reading comprehension, the concepts embedded in Skinner's (1957)

analysis of verbal behavior appeared to be a good guide for research

in this area.

Skinner (1957) wrote about reading comprehension as a particular

instance of verbal behavior. Within his analysis, understanding

written language is similar to listening comprehension. They are

similar in that they both need a preceding set of events for either

one of them to occur. Also, both kinds of comprehension must be

functional to maintain their occurrence (i.e., produce benefits

for the reader).

A difference between listening comprehension and reading compre-

hension is that listening comes under the control of another person's

vocal behavior while reading comes under the control of words. People

with normal hearing will probably have a greater chance to receive

reinforcement from listening than from reading. In other words,

listening appears to have a longer history of reinforcement than

does reading. Also, it is easier to have access to another person

who will engage in conversation than it is to find a person with whom

one can discuss what one has read. Another difference appears to be

the rate at which conversational speech occurs as compared to reading

rate. People have been trained to listen and understand conversations

(which include inflection) of 150 to 200 words per minute (Wolking,

13S~). The average elementary aged school student is trained to read

only 40 to 90 words per minute (Lovitt, 1982).












Earlier researchers demonstrated that receptive vocabulary

is prerequisite for comprehension of written words (Sidman, 1971;

Staats & Staats, 1962). Also, there is a body of research to

suggest that when the rate of reading approaches conversational

speed, there is an improvement in comprehension (Lovitt & Hansen,

1976) and retention (Berquam, 1981). Although inflection has been

implicated as an important variable for the development of compre-

hension, the literature contains very little on how rate and inflec-

tion affect comprehension and the maintenance of comprehension.

From the above concepts and research an empirical question

concerning rate and inflection developed. Taking Skinner's (1957)

analysis of reading and listening comprehension, perhaps reading

comprehension can be improved when the reading episode approximates

conversational behavior. In other words, can comprehension and

maintenance of comprehension improve when reading sounds similar to

the way people speak? If so, students can learn that reading has

functional value. What is said can be equivalent in function to what

is read. To answer the question two independent variables will need

to be controlled. These are oral reading rate and inflection.

To determine the impact that the independent variables have on

comprehension and maintenance of comprehension, a single subject

design was used. Single subject research has been proven to be a

successful strategy in observing and subsequently controlling for

individual variation in responding (Johnston & Pennypacker, 1980;

Sidman, 1960). Precision Teaching will be used as a way to teach












and measure inflection, reading, comprehension, and the maintenance

of comprehension (Armstrong, 1981; Berquam, 1981; Hansen, 1979;

Haring, 1978; Wolking, 1982).



Significance


Society views reading and reading comprehension as an important

measure of the effectiveness of public school instruction (Haring,

1978). The importance of reading with comprehension cannot be

underestimated (Turner & Fey, 1977). This study attempts to describe

one or more variables that may have a positive effect on the develop-

ment of comprehension and the maintenance of comprehension. Hope-

fully, this will increase the efficiency with which reading comprehension

comes under the control of teacher-environmental arrangements.

There appears to be a correlation between the illiteracy index

and public opinion. The 13th Annual Gallup Poll, funded by Phi

Delta Kappa, reported that 20 percent of the parents of children

attending public schools gave the schools a grade of D or F (Kappan,

1980). This was the highest percentage of failing grades given in

eight years. A teaching strategy that uses effective and accountable

techniques may help to improve the reading performances of students.

More than likely, when the reading performances improve significantly,

suoport for public education will increase. Familiarity with the

mechanisms that enhance reading and comprehension will give school

personnel the opportunity to take a leadership role in enhancing

the accountability of public school instruction and improving public

opinion.












Statement of the Problem


The present study investigates the effect of oral reading rate

with and without inflection has on the dependent variables, compre-

hension and maintenance of comprehension.

This study also serves to extend Lovitt and Hansen's (1976)

findings. In their study, high oral reading rate was not compared

with other rates. Also, they did not report any control for inflec-

tion. This study also systematically extends Berquam's (1981) findings

that high rate increases retention, by replication with a different

skill.



Delimitations


Comprehension has been defined by many investigators as having

several components. This study will only describe the relationship

of the independent variables with measures of comprehension and

maintenance of comprehension. The dependent variables measure

instances of recall. This limits the generality of the findings

across other components of comprehension. However, as already noted,

increased oral reading rates resulted in improved comprehension

across other components (Lovitt & Hansen, 1976).












Definition of Terms


Accuracy Ratio: This is a measure of accuracy derived from the

ratio Frequency Correct
Frequency Error
Behavior: The behavior of an organism is that portion of the

organism's interaction with its environment that is characterized

by detectable displacement in space through time of some part of the

organism and that results in a measurable change in at least one

aspect of the environment (Johnston & Pennypacker, 1980).

Celeration: This is the basic unit of measurement to describe

behavior change; change in frequency per unit time.

Comprehension: Comprehension is an instance of verbal behavior

known as an intraverbal. An intraverbal is a verbal response in

which the prior controlling variable is a verbal stimulus with no

point-to-point correspondence between the stimulus and the response.

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 equivalent to rate in this

definition.

Inflection: A change in pitch or loudness of the voice. The

change of form that words undergo to mark case, gender, number,

tense, person, mood, or voice (Merriam-Webster, 1974). In addition

to the above definition, inflection in oral reading comes under the

control of punctuation marks.











Latency: The amount of time between the occurrence of a

signal and the beginning of a movement.

Learning: A change in performance per unit time; also called

celebration.

Level of difficulty of material: Level of difficulty of

material is defined by the frequency at which the student demon-

strates performance of 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).

Readability: Readability is defined by Fry's readability formula.

This formula will provide an estimation of the grade level of reading

material.

Response Class: A response class is a set of responses that

have-at least one characteristic in common. These characteristics

may involve the 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 a hand, foot, knee, or body, but the

result is the same on the environment) (Whaley & Malott, 1971).

Standard Behavior Chart: A standard six-cycle semi-logarithmic

chart that displays frequency as movements/time and celebration as

movements/minute/week.



Summary


Experimental questions were developed to determine whether oral

reading, when it approximates conversational speech, can enhance












reading comprehension and maintenance of comprehension. More

specifically, the purpose of this investigation is to understand

the interaction of oral reading at two rates with and without

inflection as it impacts reading comprehension and maintenance of

comprehension. Precision Teaching, single subject technology, and

Skinner's Verbal Behavior paradigm, are the strategies that will be

employed to engineer the research design, data analysis, and conclu-

sions.

Special educators who are interested in the area of reading

have only recently begun to determine the efficacy of oral reading

in enhancing comprehension. Oral reading at high rates has been

shown to have a positive effect on reading and reading comprehension

(Lovitt & Hansen, 1976). Studies that have investigated the effects

of varying reading rates on retention have suggested a positive

relationship between high rates of performance and retention of facts

(Berquam, 198i). This study is a systematic replication of Lovitt

and Hansen's study (1976) and also looks at the effects of inflection

on reading comprehension and its maintenance over time.


















CHAPTER II

REVIEW OF RELATED LITERATURE



The following review is divided into six sections. The first

section focuses on the nature of reading comprehension. The second

section describes theories, strategies, and tactics that relate to

reading comprehension. The third section describes the most typical

methods of assessing the dependent measure, reading comprehension.

The fourth section refers to how oral reading, rate of oral reading,

and punctuation affect reading comprehension. Section five describes

reading comprehension as an instance of verbal behavior (Skinner,

1957). Although each main section has a summary, section six is a

summary of the previous main sections.

Only those studies and position papers that were relevant to

the experimental questions were reviewed. First, oral and/or silent

reading must be the major independent variable. Second, studies

which contained at least one of the following dependent measures of

reading comprehension were chosen: answers to comprehension questions,

Cloze procedures, and free recall. Third, studies that were performed

within a laboratory setting were sought. This was thought to be an

efficient way to provide the reader with studies that have undergone

experimental analysis. Moreover, by reviewing these studies, one












begins to set the occasion to systematically replicate known

laboratory principles of reading comprehension. So few laboratory

studies were found that research conducted in applied settings with

less than adequate control had to be included.



The Nature of Reading Comprehension


A definitive statement as to what constitutes reading comprehen-

sion and how to teach it is hard to come by. It is no wonder then

that the literature contains very little experimental data on how to

improve reading comprehension (Armstrong, 1981; Van Etten, 1978;

Vogel, 1975).


Reading Comprehension as a Multifaceted Skill

Reading comprehension has usually been conceptualized as a complex

skill. There are many definitions of this skill. Smith (1969)

outlined seven categories of reading comprehension by using the

Taxonomy of Educational Objectives (Bloom, Engelhart, Furst, Hill,

and Krathwohl, 1956). These categories are memory, translation,

interpretation, application analysis, synthesis, and evaluation.

The use of these categories has not gone without criticism.

Smith (1969) reported that the list was too large and proposed

a four level approach to reading comprehension. These were literal,

interpretive, critical, and creative reading comprehension. Miller

(1976) also believed that most definitions and components of reading

comprehension were too long. He described four types of reading











comprehension. These were (a) recalling information from long-term

memory, (b) elaborating and abstracting, (c) interpreting, and (d)

naming. In addition to these categories, listening comprehension

and word comprehension have also been included as a definition of

reading comprehension (Lerner, 1976; Vogel, 1975, 1977).


Listening and Reading Comprehension

Studies (Sidman, 1971; 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; Sidmarr, 1971).

Some differences between reading and listening are (a) listen-

ing (language or speech) is typically acquired first, (b) the process

of speech acquisition is more gradual (Staats & Staats, 1962), (c)

speech acquisition is under the control of stronger reinforcers

(Staats & Staats, 1962), and (d) speech is typically taught in a

one-to-one setting (Bijou & Baer, 1961). The use of vocal language

is usually learned before reading. Reading comprehension is strongly

influenced by spoken language (Staats & Staats, 1962). Along the

same lines matching-to-sample (visual and/or auditory equivalence

training)/ has been used as a definition of reading comprehension

(Sidman, 1971; Wolking & Greenwood, 1979). This is in line with

Skinner's (1957) understanding of comprehension. In his classic

book Verbal Behavior, Skinner (1957) defines reading comprehension

as an intraverbal. For example, "A person comprehends a text when

he can describe it in different words" (Johnson & Chase, 1981, p.

110).











Reading Comprehension as Facts

The most widely used measure of comprehension in the public

schools is the ability to remember specific facts from what has been

read (Guszak, 1969; Johnson & Chase, 1981; Rystrom, 1970). The

ability to recall facts does appear to be a fundamental prerequisite

for all other measures of comprehension. Without the ability to

recall facts, concept formation and other abstractions will be

limited (Johnson & Chase, 1981). Thus, one might propose that train-

ing a reader to recall facts should be the primary comprehension

task.


Oral vs. Silent Reading: Which Way Maximizes Comprehension?

During the early 1900's, E. L. Thorndike recommended that an

exclusive emphasis on precise pronunciation and oral reading was not

adequate for developing comprehension (Thorndike, 1917). His conclu-

sions initiated a greater awareness and concern for reading comprehen-

sion and silent reading; however, data to support his contention were

not provided. Yet, the emphasis shifted from oral reading to silent

reading within the public schools. Along with silent reading, an

emphasis on decoding words has been used with some research indicating

gains in comprehension (Rystrom, 1970). However, a new body of

research has been supporting the use of oral reading to increase

comprehension.

This new research suggests that when oral reading rate reaches

to a standard of performance, comprehension improves (Armstrong,

1981; Hansen, 1979; Lovitt & Hansen, 1976). Similarly, Samuels












(1979) demonstrated the effectiveness of using a standard within

his repeated readings paradigm was an effective way to increase

comprehension.


Summary

Reading comprehension has usually been viewed as a multifaceted

complex skill. Silent reading with an emphasis on teaching decoding

skills and word comprehension has been the accepted way to teach

comprehension. Other ways of teaching comprehension have been

related to listening comprehension. The most widely used measure

of comprehension has been recall of-facts. Since the early turn

of the century, silent reading along with teaching decoding skills

has been the predominant method of teaching reading and comprehension.

In more recent years a body of research has evolved to support the

use of oral reading at high rates to increase comprehension.



Theories, Strategies, and Tactics of
Reading and Comprehension


Many different ideas of how to teach reading and to maximize

comprehension exist. Many reading programs have been developed to

coincide with stages of child development. Theories of child develop-

ment are myriad.


Traditional Teaching Methods

Teaching reading usually begins as a sequential process. More

often, beginning readers are taught to read from a basal series












(Heilman, Blair, & Rupley, 1981). This early reading curriculum

usually involves letter naming, saying letter sounds, word-picture

recognition and word attack or phonetic analysis, respectively.

More often than not, little emphasis is given to speed or fluency

building. Also, mastery of reading skills is usually assessed over

a long period of time. Teaching in the regular classes is also

performed in large groups. Students are asked to read in a round

robin format with each child reading approximately 40-50 words in

30 to 45 minutes.

Since the early 1900's many reading methods have been developed.

One such method that existed since the 1930's is the Language

Experience Approach (LEA).

Many educators have advocated the LEA rather than a skills

approach (Hasselriis, 1982). LEA requires teachers to use printed

material in its "natural" form; then whole sentences are combined to

form stories, articles, magazines, newspapers, and books. It requires

that if instruction is focused on words, parts of words, or other

subsets of language, those fragments of language should not be

removed from their original context (Hasselriis, 1982). This method

recognizes the relationship between reading, speaking, writing, and

listening as part of an individual's personal experiences (Spache &

Spache,1977). It further requires that the students be given

extensive opportunities to read, write, listen, and speak in natural

settings rather than through contrived means. Proponents see that

a child cannot be expected to deal with ideas of language in reading

that are much further advanced than those he can speak or write.












Linguistics and reading comprehension. Reading has been

related to language development. Concerned with language and its

components, the linguistic approach has been drawn upon to guide

curriculum development, instruction, and methods of teaching reading

and comprehension. Linguistics has been defined as the study of

language. Proponents of this system attempt to explain the phonemic

and grammatical structures of language (Karlin, 1975). Two of the

more salient domains in linguistics are structural linguistics and

generative or transformational linguistics. Structural linguists

study oral language to identify its sounds; its units of meaning,

carried by words and by parts of words; and its syntax. Transforma-

tional linguists seek to find the knowledge one must have in order

to make such utterances or to understand them. They typically try

to produce grammars that describe the often unobservable knowledge

of language (Karlin, 1975).

Transformational grammarians believe that knowledge of the

grammar of a language enables one to produce it, and that meaning

is derived not from the words in a sentence, or the surface structure,

but from knowledge of the deep levels of the language. Grammar

serves as a bridge between the deeper structure of a language, in

which meaning resides, and the surface structure of the language

(Chomsky, 1965). Comprehension then is a process which occurs between

the surface and the deep structure by means of the vocabulary,

phonology, and syntax of a particular language. Therefore, compre-

hension of what is read is assessed through techniques that rely on












inferences and/or prediction of a next word within the context of

a passage.

Similar to the area of linguistics, psycholinguistics has

recently had an impact in the arena of reading comprehension.

Psycholinguists believe that reading and comprehension are an

active process. For example, people can read words without decoding

and can understand without knowing how to read words (Smith, 1971).

Psycholinguists believe that comprehension of words and word identi-

fication may be a result of comprehending the story before reading

it. Also, psycholinguists believe that readers read for meaning

rather than word identification (Smith, 1971). A technique that has

been developed to test comprehension within this model is the "Cloze"

procedure. This procedure uses an automatic word deletion process

(e.g., every nth word), whereby words are removed from a printed

passage. The pupil's task is to say the exact word that was removed

and replace it (Rankin, 1965). Proponents of this method indicate

that pupils can predict the next word on a Cloze procedure when their

prior knowledge, general understanding of the material, context

clues, and a knowledge of word usage are well developed.

Reading in context with a focus on punctuation has been described

as a way to increase comprehension (Heilman, 1981). This is consistent

with the linguist point of view. The linguists and psycholinguists

have also emphasized comprehension to be partly influenced by intona-

tion. In addition to the consonant vowel phonemes, they relate that

English uses a number of intonational phonemes. Speech consists of












a flow of words arranged in particular patterns which result in

distinctive rhythms or melodies which are unique to English

(Heilman, 1967). Reading in context along with using the correct

intonational patterns is viewed as having an impact on comprehension.

Reading as operant behavior. In 1938, B. F. Skinner published

The Behavior of Organisms. In that book he describes the differences

between operant behavior and reflexes. He described the use of fre-

quency as a basic datum of behavioral research to describe how behavior

changes under operant control. It was notuntil Azrin and Lindsley

(1956) used operant procedures with children that the power and

generalizability of Skinner's fundamental principles could be

realized. In 1957, Skinner's book Verbal Behavior conceptualized

how reading and comprehension can be defined as and come under operant

control.

The inaugural study that used principles of operant behavior to

explore the relationship between reading behavior and reinforcement

was performed by Arthur Staats and associates (Staats & Staats, 1962).

This led to the systematic development of strategies and tactics for

the investigation of reading behavior. Staats and his colleagues

(Raygnor, Wark, & Warren, 1966; Staats, Minke, Filey, Wolf, & Brooks,

1964; Staats & Staats, 1962) demonstrated that reading, like any

other operant behavior, was influenced and controlled by its conse-

quences.

Often, operant technology has focused on arranging contingencies

of reinforcement for academic behaviors. For example, the use of












"token economies" (Ayllon & Azrin, 1969) has been used to increase

academic performances across most subject areas (Ayllon & Roberts,

1974; O'Leary, Becker, Evans, & Saudergas, 1969; Wolf, Giles, &

Hall, 1968). The efficacy of using reinforcement contingently has

also been demonstrated on a national level with Project Follow

Through (Benjamin, 1981).

Reading comprehension is also a class of responses which a

science of behavior can and has investigated (Staats & Staats, 1962).

However, the study of reading as a natural science phenomenon has

only recently gained momentum (Indrisano, 1982).


Recent Strategies and Tactics for Teaching Reading and Comprehension

Recent approaches to teaching reading have focused on strategies

that have proven empirical value. Some of these are still undergoing

parametric evaluations. One particular approach which has gained in

momentum over the past decade is Direct Instruction.

Direct Instruction. Direct Instruction has emerged as an effec-

tive method to develop reading and comprehension. The beginning

sequence of this approach begins with first having the child say the

letter sound, word-picture identification, blending sounds into words,

and letter naming, respectively (Englemann & Bruner, 1974). Much of

the direct instruction approach requires the child to continually read

aloud. Comprehension at the beginning levels emphasizes visual dis-

crimination tasks, matching to sample, and visual equivalence training

(Sidman, 1971). Direct Instruction emphasizes teaching children

letter sounds before naming them. This was developed empirically as











a method to teach children that reading is similar to speaking.

An important hallmark of Direct Instruction is the interaction

between the instructor and the students. Immediate feedback is

typically provided. Also, Direct Instruction is under the philosophy

that all children can learn. Therefore, instruction does not pro-

ceed unless the pupil demonstrates that a skill or a lesson has been

learned. There are important studies published that have described

the efficacy of Direct Instruction with a program such as DISTAR

(Benjamin, 1981). These studies typically have shown how DISTAR,

compared to many other popular teaching strategies, managed to

raise the performances of low achieving students so that they were

on and above grade level (Benjamin, 1981; Englemann, 1975).

Precision Teaching and reading comprehension. Recent investiga-

tors have used the technology of Precision Teaching (PT) and reported

studies in which reading comprehension was the primary dependent

variable (Armstrong, 1981; Fleisher, Jenkins, & Pany, 1979; Hansen

& Lovitt, 1976; Jenkins, Barksdale, & Clinton, 1978; Lovitt & Hansen,

1976; Roberts & Smith, 1980; Weaver, Holmes, Curtis, & Reynolds,

1970). The findings of and the paradigms that were used in each of

these studies can be seen in Table 1. Included in Table 1 is the

only study that could be found comparing oral reading and inflection

with comprehension (Poulton & Brown, 1967).

Precision Teaching (PT) has five basic ingredients: (a) replic-

able teaching procedures, (b) individual analysis, (c) experimental

control, (d) direct measurement, and (e) daily measurement (Haring,

1978). Within this framework, 0. R. Lindsley introduced Precision












Table 1

Review of Selected Studies on Oral Reading Rate
and Comprehension


Author(s) Subjects Measures Intervention


24 Housewives


Answers to Oral and silent
questions. reading.


Results: Memory for the first 30 percent of the passage was less
after reading aloud. Remainder of the passage memory was equally
good for both conditions. The last 10 percent of the passage was
remembered reliably after reading aloud. Difficulty level of compre-
hension questions was not reported. Also, questions restrict the
total number of facts that could have been recalled.

Limitations: Readability was not controlled. Rereading passages
was not controlled. Also, rates of oral and silent reading was not
controlled. The use of punctuation and inflection was not controlled.


Weaver, Holmes, 18 volunteer
Curtis, & undergraduates
Reynolds, 1970


Cloze proc. Oral and silent
reading with
punctuation. Oral
reading without
punctuation.


Results: Reading silently with punctuation resulted in higher scores
on the Cloze questions.

Limitations: Reading rates were not controlled. The number of trials
read silently and orally could not be managed successfully. Cloze
procedure restricts that total number of facts that could have been
recalled.


Words read C
per minute s
correct and o
error. q
Percentage of a
correct r
answers on
comp. questions.


ontingency of
kipping. Drill
n words and
questions read/
answered incor-
ectly.


Poulton &
Brown, 1967


Levitt &
Hansen, 1976


7 LD boys












Table 1--Continued


Author(s) Subjects Measures Intervention


Results: Subjects oral reading rate increased during the skip and
drill phase. There was also a simultaneous increase in number of
correct answers to comprehension questions.

Limitations: Minimal changes from the skip and drill phase to the
final baseline phase. The effects of punctuation training was not
reported in the study.


3 LD students


Correct oral
reading rate.
Answers to
-comp. questions.


Money contingent on
correct answers to
comprehension
questions.


Results: Reading comprehension improved under the money contingency.
Then number of words read correctly increased also under the money
contingencies.

Limitations: Subjects began at a high level of reading correct words.
Comprehension questions restricts that total number that could have
been recalled.


36 students.
14 good readers.
22 poor readers.


Words read per
minute. Per-
cent correct
on Cloze.
Number correct
on factual and
inferential
questions.


Decoding training.
Oral reading rate
of 90 wpm. In-
struction in
reading for
understanding.


Results: Poor readers improved their number of words read correctly
with training. Good readers did not improve significantly.

Limitations: Low proficiency standard for oral reading. Did not
control punctuation and inflection. Difficulty level of comprehension
questions not reported. Comprehension questions restricts the total
number of facts that could have been recalled.


Jenkins,
Barksdale,
& Clinton,
1978


Fleisher,
Jenkins, &
Pany, 1979












Table 1--Continued


Author(s) Subjects Measures Intervention


Oral reading rate
corr. Percent
correct of comp.
questions.


Contingencies to
increase oral
reading. Model-
ing and instruc-
tion.


Results: Oral reading rate increased, but answers to comprehension
questions improved less.

Limitations: Punctuation and inflection training was not controlled.
Many variables involved with the study. Difficult to separate out
the controlling variables. Level of difficulty of the comprehension
questions was not reported.


6 LD students


Type and number
of responses to
comp. questions.
Response latency.


Contingent
pennies for
correct answers
to comprehension
questions for
each level of
material pre-
sented at
staggered inter-
vals.


Results: There was an inconclusive relationship between the contin-
gencies and the percent of comprehension questions answered correctly.
In all six cases there was a positive relationship between the median
number of words read correctly and comprehension. There was a nega-
tive relationship between words read correctly per minute and mean
response latency. In all six cases there was a positive relationship
between errors per minute and response latencies.

Limitations: Did not control for punctuation training. The rate of
oral reading was not controlled. Also used comprehension questions
that may have limited the total number of facts that could have been
recalled. The level of difficulty of the comprehension questions was
not reported.


Roberts &
Smith, 1980


8 LD boys


Armstrong,
1981












Teaching at the University of Kansas in 1964 (Haring, 1978). A

hallmark of PT is the insistence on daily charting of targeted

skills on a six cycle semilogrhithmic chart (Pennypacker, Koenig,

Lindsley, 1972). PT also incorporates the use of frequency as the

basic datum to describe behavior. Many classrooms throughout the

country use PT techniques, although great variability exists between

classrooms (Lovitt, 1977). Lovitt (1977) describes five characteris-

tics of PT that are the best practices within these classrooms (p.

175).

1. The teacher or pupil must pinpoint each behavior of a

child's program. !F one goal is to increase ability to read orally

from a certain text, a situation should be arranged to deal directly

with that behavior.

2. An aim (e.g., goal) must be determined for each identified

behavior. In order to determine an aim, the teacher or pupil must

decide the rate at which the selected behavior should occur and the

date on which that criterion should be achieved. Lines of progress

are then drawn on a chart of that activity from the intersection of

the current date and rate to the point of the projected date and rate

intersect.

3. Third, the teacher or pupil must count the number of times

the behavior occurs.

4. The teacher or pupil should chart each day the frequency

of the pinpointed behavior.












5. The teacher and pupil should evaluate the performance of

each charted behavior every day. If the correct rate is above and

the incorrect rate below the corresponding progress lines, the

current instructional technique should be continued. If, however,

the progress is not satisfactory, an instructional change should be

considered.

Direct Instruction and PT are two methods that can be used

in a coordinated way within the classroom. Both advocate direct

interventions. PT and Direct Instruction focus on observable data

to guide interventions that are necessary to have a learner reach a

mastery criterion. PT also uses frequency as its basic datum. PT

differs from direct instruction methods by its insistence on frequency

of response as being the datum of interest. Typically, Direct

Instruction uses a percent correct criterion. In other words, PT

relies on accuracy and rate to measure performance while Direct

Instruction will emphasize the number of trials to criterion or

accuracy.

The ecological approach. This approach is similar to the Language

Experience Approach and to ABA. The approach deals with the inter-

actions between the organism as a whole and its entire environment,

which includes the existence and behavior of other organisms in that

environment (Bijou & Baer, 1961). In the educational sense, ecology

is the relationship of the learners to their instructional environment

(Indrisano, 1982). The learner is viewed as a dynamic part of an

ongoing changing environment.












Automaticity theory. The tactics advocated by the proponents

of automaticity are in some ways similar to PT. This method

advocates the use of repeated readings (Samuels, 1979). "Fluent

readers automatically process information at the visual and

phonological levels and are therefore able to focus attention on

the meaning codes in the text" (Box, 1982, p. 51). Advocates of

this theory believe that when each level of processing receives

attention and assisted repeated reading to a fluent rate, selective

attention will be improved. Samuels (1979) advocates oral reading

at 85 wpm as the criterion for fluent reading. Although not proponents

of automaticity theory, Smith (1969) and Beck (1982) advocate fluency

to be at 150 to 200 wpm.

PT and automaticity theory both include fluency training and

rereading as an important tactic to increase comprehension. Consistent

with PT is the notion that fluent reading is a more functional way to

increase reading comprehension for three reasons. First, fluent

reading can increase retention. Second, fluent reading increases the

probability that the reader and/or audience will attend to what has

been read. Consequently, this may provide the reader with a greater

variety and frequency of reinforcement because of the faster pace.

Third, as this study seeks to substantiate, when oral reading approxi-

mates conversational speech, comprehension will improve. These

reasons are similar and coincide with those given by the proponents

of automaticity and PT.












Microcomputers and reading. Another approach which has recently

been revived because of advances in electronics is the use of

programmed instruction through the use of microcomputers. The use

of machines and programmed instruction are contrived methods that

can be made a valuable tool for instruction (Skinner, 1958). More-

over, through programmed instruction and teaching machines the student

can rehearse newly learned skills, receive immediate reinforcement,

and develop independent work habits. Although much more technologically

advanced than teaching machines, microcomputers can be used as a very

sophisticated learning device as Skinner had envisioned. Furthermore,

microcomputers can reinforce small increments of behavior that other-

wise may go unnoticed by the instructor. However, the use of micro-

computers have been limited because of software design. The importance

and/or the expertise to incorporate fluency building into educational

software has typically not been made part of the recently developed

software for students. However, some educators have recognized this

deficiency and have begun expressing their concerns and have suggested

alternative strategies (Eshelman, 1983; Wolking & Buss, 1983).


Summary

The nature of reading comprehension is difficult to delineate.

Several theorists have believed that reading comprehension is a

multifaceted skill. Others have actively sought to decrease the

number of skills necessary to constitute what is commonly called

reading comprehension. There is agreement as to the importance of

retaining facts as a minimum skill that is the foundation for later












types of comprehension. During the early 1900's silent reading

was considered to be the best method toteach reading and comprehen-

sion. Then along with silent reading, decoding and word comprehension

were thought to be the best ways to optimize comprehension. Some

researchers in the 1960's provided data which support the notion

that when the rate of oral reading reaches a proficiency standard,

there is a concomitant increase in comprehension. Lovitt and Hansen

(1976) provided data that indicate that although improvements were

not as great during oral reading training, reading silently at high

rates improved comprehension of materials that were read silently.

There are many theories as to how comprehension should be

conceptualized and subsequently taught. Some theorists have con-

jectured that language and comprehension are closely related. The

linguists, which include the psycholinguists, and the transformational

grammarians advocate developing comprehension through increased under-

standing of the deep structures of language. The Language Experience

Approach advocates the use of already learned language as a place to

begin instruction. Also, reading materials should consist of func-

tional value to the students. Similarly, proponents of Applied

Behavior Analysis see reading as operant behavior that comes under

the control of contingencies within the immediate environment. The

fundamentals of operant research were developed by B. F. Skinner

(1938). Skinner's (1957) conceptualization of reading and compre-

hension was presented in his book Verbal Behavior. He defined

comprehension as an intraverbal. As an intraverbal (see definition












of comprehension), the aim of teaching reading comprehension will

be to functionalize it. More recent approaches to reading compre-

hension are Direct Instruction, Precision Teaching, and Automaticity

theory. The research supporting these recent approaches is providing

evidence as to their effectiveness across many environments. More

specifically, instruction which incorporates a standard of per-

formance, direct and frequent measurement of performance, high rate

of oral reading along with practice of the same or similar stimuli

has the greatest probability of increasing reading comprehension.



Dependent Measures of Reading Comprehension


Without a standard definition of reading comprehension, strategies

and methods to measure its occurrence are difficult. Without a standard

definition, it is difficult to decide which treatment effects are to

be considered significant. In spite of this difficulty investigators

have developed ways to measure comprehension.

In general, responses to comprehension questions are the most

common format. Typically multiple choice type questions are used.

However, true/false questions, fill-in-the-blanks, and vocabulary

definitions have also been used (Hansen, 1979). Methods of assessment

that employ questions have many drawbacks. Comprehension questions

reveal only a portion of the available information that may be obtained

from textual material. Also, students are assessed on what another

person considers to be the major points made in the text.












More recently other methods have been developed to correct

the problems inherent in the use of questions to measure compre-

hension. Short-answer fill-ins have been suggested in an attempt

to facilitate recall of response rather than recognition. These

open-ended questions do not resolve the other concerns about questions

as a measure of comprehension. Cloze techniques have been suggested

as another alternative to comprehension questions. Difficulty arises

with the Cloze procedure because it does not provide evidence of the

ability to understand entire passages. Moreover, this technique

does not permit evaluation of a person's critical and inferential

comprehension skills (Hansen, 1979).

A more direct measure of comprehension was proposed by Sidman

(1971). He believed that reading comprehension occurs initially when

a beginning reader can match a word to a corresponding picture. In

this way the child indicates that written words are symbolic repre-

sentations of objects found in the environment. He also proposed

that pictures and words become equivalent to each other, because

they both can serve as discriminative stimuli for the same operant

(same class of responses), of the same auditory word. Several

studies have reported that matching spoken words to pictures and

to printed words is a sufficient prerequisite for the emergence of

reading comprehension at the word level (Sidman, 1971).

Another direct measure of comprehension is whether the child can

locate the actual object or perform the prescribed action. This

method relies on recall as a prerequisite'skill. The ability to











recall facts does appear to be a fundamental prerequisite for all

other measures of comprehension. Without the ability to recall

facts concept formation and other abstractions will be limited

(Johnson & Chase, 1981). Thus, one might propose that training a

reader to recall facts should be the first comprehension level.

A free recall task that uses frequency of facts told correctly

to criterion (e.g., 15 to 30 correct facts recalled per minute), can

significantly aid retention, type of information remembered and

provide a direct measure of its occurrence (Hansen, 1979). The

ability to recall facts to criterion might then be considered as a

tool skill for all other types of comprehension. Once proficiency

has been demonstrated, another type of comprehension should be taught.

Johnson and Chase (1981) proposed a hierarchical instructional

typology for adult learners based on Skinner's Verbal Behavior (1957).

Their suggested sequence begins with teaching cf free recall and then

moving on to more abstract methods of teaching and measuring compre-

hension. Within the instructional typology all types of comprehension

measures that use answers to literal questions, and fill-in-the-

blanks, are considered as fundamental steps in developing abstract

comprehension. Other measures of comprehension such as Sidman's

(1971) equivalence training would likewise be regarded as a funda-

mental step in developing comprehension. This:study will measure

comprehension on a fundamental level. In other words, all responses

will be considered as free recall across different stimulus events.

Since free recall is the basic building block of all later












comprehension, steps should be taken to determine the accuracy

and rate that will improve more abstract types of comprehension.

This is regarded as a step towards standardizing a definition of

comprehension.


Reading Comprehension and Memory

Although recall of facts is the most sought after measure of

comprehension, many reading programs do not employ teaching

strategies that enhance memory. Reading curricula usually rely on

one trial learning per story. Students are expected to read a

story one time and then answer comprehension questions. This does

not appear to be the most efficient teaching strategy to enhance

memorization. Ebbinghaus (1885) learned that the more exposure or

practice he had to learn a list of words, the greater was his recall

for these words.

Although practice is highly correlated with an increase of

retained material, Berquam (1981) investigated the relationship

between rate of performance and extra practice of learning nonsense

syllables. He used a paired associate task in a trigram form.

His study confirmed the results of earlier investigations, in that

level of previous learning has a very strong relation to subsequent

retention of skills. The group that worked to increase performance

required less time to relearn the nonsense trigrams than did the

group which received extra practice without fluency training.

Since Berquam's study used frequency of response data, the results

indicate that the basic learning/retention relation is found with












frequency data as well as with the accuracy measures of previous

studies. Berquam (1981) concludes from these findings, "It seems

likely that teachers could increase their students' skill levels

and subsequent retention by using short, concentrated periods of

fluency training, similar to those used by precision teachers"

(p. 73).

This study builds on Berquam's findings. In Berquam's study,

he used nonsense trigrams. For this study, reading passages will

be used. As Berquam used rate in terms of frequency, this study

will also use frequency of oral reading as a basic datum. This

sets the occasion to systematically extend Berquam's (1981) findings.


Summary

Many different kinds of measures of reading comprehension exist.

Most of the ones used are question and answer type. Problems with

questions usually center around the subjectivity of the answers.

Also, the format in which the questions are asked usually does not

contain a time component. The question answer format usually solicits

the fundamental skill of recall. Since recall is a fundamental skill

for later types of comprehension, then a format that directly assesses

facts per some unit of time appears to have great merit. Memory has

been an important issue for experimental psychologists and educators.

Studies have been performed in which practice and, more recently,

practice with fluency increase retention on later measures.












Oral Reading, Rate, Punctuation, and Comprehension


Since retention increases with practice and fluency, the rela-

tionship between oral reading rate and comprehension should be

investigated. The relationship between oral reading and comprehen-

sion has received little attention. An early investigation of the

efficacy of oral and silent reading was conducted by Poulton and

Brown (1962). In their study 24 housewives read some passages

aloud and others silently and then answered questions. Time

allowed for reading silently was matched for time reading aloud.

The researchers concluded that memory for the first 30 percent of

the passage was less after reading aloud. For most of the remainder

of the passage memory was equally good. The last 10 percent of the

passage was remembered better after reading aloud.

To improve oral reading Lovitt and Hansen (1976) used a high

rate of oral reading along with word drill made up of words taken

from the pupil's reading passage that were read incorrectly. They

also reported a simultaneous increase in comprehension. Their pro-

cedure also generalized to silently read material. Also, when allowed

to read the passage silently after they were trained to read it

correctly orally, their silent reading also improved. The results

from this study may also be generalized to silent reading as well.

Oral reading appears to improve comprehension, but the rate at which

oral reading maximizes comprehension has recently begun to receive

attention.












There is little empirical data at this point to determine

which rates of reading are needed to maximize comprehension

(Buss, 1982). A wide range of proficiency rates for academic

skills are reported in the literature (Mercer, Mercer, & Evans,

1982). Haughton (1982) reported that oral reading reaches competent

levels in the 250 to 400 words correct per minute range, but in the

first three grades most children receive instruction in the 50 to

70 words per minute range. Smith (1971) reported that "normal"

readers call words at 200 wpm. After reviewing the reading per-

formance of over 3,000 second and third graders, Kunzelman (1973)

reported that the students designated as top readers by their

teachers read orally between 150 and 210 correct words per minute

with two or less errors at a particular grade level is considered

instructional level (Starlin & Starlin, 1973).

There appears to be a great deal of discrepancy of what consti-

tutes proficient reading. However, W'olking (1973) reported that,

by sixth grade, the median child has attained approximately 95

percent of adult accuracy and 65 percent of adult speed in the

performance of basic academic skills. Wolking's (1973) data appear

to translate into an easy formula by which a proficiency can be

judged on a criterion referenced basis. However, this does not

settle the question of which rate maximizes comprehension. Reading

at or near conversational adult speech has been regarded as a

possible goal for oral reading rate that may maximize comprehension

(Wolking, 1982). This also appears to be in line with Skinner's












understanding of verbal behavior (1957). Moreover, in the Wolking

findings, comparing children with adults to determine rates appears

to approximate the development of a reading rate that resembles

conversational speech. Conversational speech has been clocked at

150 to 200 wpm (Wolking, 1982). Although this review does not

report any studies that used a proficiency standard of 150 to 200

wpm, some researchers have used high rates of oral reading to

determine its effect on comprehension.

In their study, Fleisher, Jenkins, and Pany (1979) trained 36

fourth and fifth grade students to read at 90 wpm to see the effects

it would have on comprehension. They had two groups of children.

The first group were good readers while the second group were poor

readers. The investigators concluded that there was no significant

change in any of the three measures of reading comprehension between

the groups.

In a more recent study Roberts and Smith (1980) demonstrated the

efficacy of improving reading fluency by using high rates of oral

reading with eight learning disabled boys. However, comprehension

improved less than oral reading correct did. In support of Roberts

and Smith's (1980) study, Armstrong (1981) reported a positive

correlation to exist between high rate of oral reading and compre-

hension.

The impact that punctuation has on oral reading should not be

overlooked (Weaver, Holmes, Curtis, & Reynolds, 1970; Wolking, 1982).

One study that was significant to this research was found which












directly addressed the question of punctuation and its relationship

with comprehension. In this study 18 volunteer undergraduate

students were assigned randomly to one of six conditions: oral

reading with punctuation, oral reading unpunctuated; reading final

word in sentence, punctuated; reading words in sentences unpunctuated

(Weaver et al., 1970). A Cloze procedure was used to assess compre-

hension. The investigators concluded that reading silently with

punctuation resulted in higher scores on the Cloze procedure. They

concluded that the imposing of the verbal production has an inhibit-

ing effect upon comprehension (p. 82).


Summary

In summary, investigators studying the relationships between

reading and reading comprehension have reported results which suggest

that oral reading at high rates can improve comprehension. Except

for Lovitt and Hansen (1976) and Armstrong's (1981) studies, most

results have not been replicated. The rates at which oral reading

will optimize comprehension and retention have not been reported in

the studies reviewed. Studies which sought to determine the effects

of oral reading rate did not use proficiency standards that were

advocated by precision teachers. Experts in the area of Precision

Teaching report a rate of 200 wpm to be the proficiency aim for

reading aloud. Many of the studies cited did not control for the

possible effects that punctuation training may have had on oral

reading and comprehension. A study that did investigate the

relationship between reading and use of punctuation had design












difficulties. In that study, they could not prohibit rereading

of the silent reading phase. The effects of overt and covert

reinforcement have been well documented. Yet, many of the studies

cited in this review do not report the use of contingencies that were

operating during the study. This investigation will use a minimum

of intermittent praise as the only source of reinforcement. Also,

some studies which used group designs did not adequately control

for receptive vocabulary differences in their subjects.



Verbal Behavior


Skinner's (1957) conceptualization of verbal behavior can help

to improve the strategies and methods that are presently being used

to research and to teach reading comprehension. Skinner (1957)

made the case in his book Verbal Behavior that all instances of

verbal behavior such as reading, writing, answering questions, are

defined by their function. As noted earlier, comprehension is an

intraverbal. This concept uses the idea that comprehension like

other kinds of verbal behavior is part of a dynamic system that is

functional in some regard. For example, if a teacher has established

herself as either possibly rewarding and/or aversive, she can elicit

responses from a student thatare functional to both the student and

the teacher. The response by the student would be functional

because he/she would be either seeking a positive reinforcer or

possibly avoiding a punishment. The response by the student then

is a stimulus for the teacher to behave in some fashion. Thus, the












student's response sets the occasion for some teacher behavior.

Viewed in this light, intraverbals such as free recall can be

taught for some functional value. For instance, when a child learns

to provide answers at a rate of 20 per minute, he/she will probably

move on to a new story and/or understand similar stories at a faster

rate. Since the occurrence of a high rate behavior typically

implies that some reinforcement contingency is operating (Ferster

& Skinner, 1957), it can be said then that high oral reading rate

has some functional value. Since recalling facts at 20 per/minute

may increase understanding a similar story at a faster rate, then

recalling facts at a high rate will take on a positive functional

value.


Experimental Analysis of Behavior and B. F. Skinner

Since this study uses B. F. Skinner's conceptualization of

reading comprehension as an instance of verbal behavior, a discussion

of his contributions and related research are included.

Behavioral practices resulting from the pioneering work of

Skinner have become the system called Applied Behavior Analysis (ABA).

Analysis on the individual level is probably the most salient charac-

teristic of ABA. This strategy is closely related to the concept of

experimental control (Johnston & Pennypacker, 1980). Other research

methods often report group data that compare the means of experi-

mental and control groups (Johnston & Pennypacker, 1980). By using

single subject designs the variation in individual responding is

readily available for visual inspection and possible subsequent












control (Johnston & Pennypacker, 1980; Sidman, 1960). Since

control of variation is feasible, statistical procedures are

often unnecessary. Systematic removal of sources of variability

in the data leads to stable responding (Sidman, 1960). When stable

responding has been achieved, systematic replication of the effects

of an intervention can be reported with greater confidence

(Johnston & Pennypacker, 1980; Sidman, 1960).

Johnston and Pennypacker (1980) have developed a definition of

behavior that includes the properties of a physical science as they

relate to behavior (see Appendix A). Verbal behavior (Skinner, 1957)

is a class of behavior within this generic definition of behavior.

Peterson (1978), based on Skinner's book Verbal Behavior (1957),

defines the verbal episode as having the following features.

1. It is established and maintained by reinforcement.

2. The reinforcement is mediated by another person.

3. The other person's action that results in the rein-

forcement must have been specifically trained in order

to reinforce speakers.

Peterson then goes on to list features which are irrelevant to a

definition of verbal behavior (p. 9). These are as follows:

1. The topography of the behavior: which muscles are used

in making the response.

2. Dynamic characteristics of the response: speed,

intensity, repetition.

3. Verbal or non-verbal stimulus.












4. Stimulus mode: auditory, visual, gustatory, etc.

5. Reinforcement features: conditioned, unconditioned,

type of schedule.

Many people involved with reading research have made the point

that reading, like speaking and listening, should be considered as a

language process (Goodman, 1976; Mcleod & Crump, 1978; Peaster, 1976;

Ryan & Semmel, 1969; Staats & Staats, 1962). Skinner (1957) defined

reading in terms of verbal behavior which is explicitly different from

language. He referred to reading as textual behavior. It is con-

trolled by a prior stimulus that is the response product (end

result) of writing behavior (such as a book) and there is point-to-

point correspondence between the stimulus and the response. 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. For example,

copying a sentence word for word from a blackboard will have a point-

to-point correspondence between what was on the blackboard and what

was copied.

As with speaking, reading comes under the control of an audience.

The audience is a type of controlling variable (Peterson, 1978). The

audience is usually a listener in the presence of whom verbal behavior

is typically reinforced. Moreover, it controls a group of response

fcrms. The teacher and the school setting can serve as the audience

effecting control over the response form likely to be emitted. For

example, in a high school Spanish class, the teacher may only wish to











respond to questions and answers given in Spanish by the students.

In this way, Spanish is strengthened and English, in that particular

class, is weakened. Thus, the teacher defines the occasion for only

certain types of stimuli to be reinforced. This also holds for

reading and comprehension. For example, when a teacher responds

only to words read correctly, the teacher sets the occasion to be

an audience that will reinforce only correct word calling. In other

words, the audience is developing or has developed stimulus control

over a behavior or a class of behaviors. Closely related to this

concept of audience is the notion of stimulus control.


Stimulus Control

Verbal behavior develops as the audience mediates reinforcement.

Stimulus control is a function of discrimination training (Whaley &

Malott, 1971). The discrimination training procedure consists of

reinforcing a response in the presence of one stimulus and extinguish-

ing it in the presence of other classes of stimuli. When the response

is more likely to occur in the presence of the discriminative stimulus

than in the presence of the nonreinforced stimulu, stimulus control

has been established. Transfer of stimulus control has been demon-

strated through fading and errorless discrimination training (Moore &

Goldiamond, 1964; Sidman & Stoddard, 1967; Terrace, 1963; Touchette,

1968).

Skinner (1957) regarded reading as a type of stimulus-response

relation (stimulus control) in which the controlling stimuli are

visual words--written or printed text. In a mediated transfer study,












Sidman (1971) sought to discover if saying a word in the presence

of the visual equivalent of the word would suffice to establish

reading comprehension. Also, would teaching auditory-visual word

matching suffice for oral reading to emerge? He reasoned that

children normally understand words they hear before they learn to

read with comprehension (p. 6). He used receptive auditory visual

training for reading comprehension. By bringing receptive language

under discriminative control of words and pictures (he used pennies

and a bell contingent on correct response), Sidman was able to

demonstrate that retarded children could match sound to word and then

word to picture without instruction. Simply, he taught the retarded

child the word symbol association by saying the word in the presence

of the written word. The child was then able to read the word and

point to the correct picture that was associated with the word. This

phenomenon has been replicated with normal preschool children (Wolking

& Greenwood, 1978).


Comprehension Questions and Free Recall as
Discriminative Stimuli for Intraverbals

The episode that is usually known as comprehension can be inter-

preted as an instance of verbal behavior. More specifically, compre-

hension in Skinner's metaphor of Verbal Behavior is called an intraverbal.

An intraverbal is a verbal response in which the prior controlling

variable is a verbal stimulus (e.g., a discriminative stimulus

situation such as reading) with no point-to-point correspondence

(i.e., a thematic response instead of an exact duplication of what












has been seen or heard), between the stimulus (the question) and

response (the answer). In other words, there is no point-to-point

correspondence between the question and the answer. In this study,

comprehension is defined as an intraverbal. The maintenance of

comprehension is simply an attempt to evoke elementary intraverbals

(Johnson & Chase, 1981) over longer time periods.

Comprehension questions can be designed to evoke intraverbals as

the mediated transfer studies (Sidman, 1971; Wolking & Greenwood,

1978) have demonstrated. Any questions or procedures that require

the subject to perform recall behavior engage that subject in per-

forming intraverbals (Johnson & Chase, 1981). However, if compre-

hension questions are not designed carefully, basic intraverbals

(the foundation for extended intraverbals and concept formation)

can be disrupted (Johnson & Chase, 1981).

Two investigators 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 compre-

hension answers missed. This should help them to discriminate the

stimuli that evoke the proper intraverbals from those that lead to

incorrect ones. The reader is encouraged to read Peterson's (1978)

introductory text to Verbal Behavior for a more detailed account of

other types of verbal behavior.












The concepts of Verbal Behavior (Skinner, 1957) may be readily

applied to the professional working with school aged children. As

Johnson and Chase (1981) have pointed out, an entire instructional

sequence can be mapped out for each individual student. This will

include a functional approach to teaching skills from the funda-

mental to the abstract level. Moreover, by engineering functional

assessment strategies, interventions can become more precise and

capable of evolving into a practical solution for many of the

difficulties that may burden school children (Grimes, 1981).



Summary


Several definitions of reading comprehension exist. This, in

part, has contributed to the difficulties involved with the develop-

ment of a standard measure of reading comprehension. Major theorists

have promoted their own definitions and measurement schemes congruent

with their theory. This has made it difficult to replicate and

extend reported findings. Based on Verbal Behavior (Skinner, 1957),

comprehension is defined as an intraverbal. This is a functional

definition which can be used as a standard definition. Along with

frequency, the basic datum of behavioral research, this can set the

occasion for researchers and practitioners to make direct comparisons

of other research that finds its way into the professional literature.

This is thought possible because frequency uses a universal dimensional

property of behavior known as countability (Johnston & Pennypacker,

1980). Also, frequency uses an absolute and standard unit of











measurement to express countability. Along with frequency as a

basic datum, conceptualizing comprehension as a functional behavior

can help to pinpoint exactly the conditions, topography, and the

outcome of the episode known as comprehension. This is considered

desirable because each of the dynamic parts of comprehension can be

analyzed and understood in relationship with the entire episode.

One common type of intraverbal which appears to be a prerequisite

for all other types of intraverbals is recall. Free recall or facts

told per minute appears to have much promise as a measure of an

elementary intraverbal. Free recall and other types of recall type

situations rely heavily on memory. Several studies indicate that

relearning skills to criterion accelerate as a function of the number

of learning trials needed to reach criterion during initial learning.

More recently, Berquam (1981) demonstrated that daily practice with

an emphasis on high rate performance decreases the number of trials

needed to relearn a skill compared with practice without fluency

building.

During the turn of the century, E. L. Thorndike (1917) spear-

headed the movement towards silent reading. His contentions prevailed

until the early 1960's and 1970's. Since then, use of oral reading has

been demonstrated as being an efficient method of increasing compre-

hension (Lovitt & Hansen, 1976). More recently, Samuels (1979)

demonstrated that repeated readings along with a fluency criterion

increase comprehension. The minimal rate range of oral reading at

which comprehension will be maximized has not yet been established.








46



Most comprehension studies have not controlled for the effects

that inflection may have had on comprehension and the maintenance

of comprehension. Similarly, studies that manipulated punctuation

did not control reading rate and/or the number of trials (resulting

in a practice effect). Inflection provides the reader and listener

with auditory cues that may aid in the comprehension of written

materials for normal hearing persons. The strategies and tactics

designed to view individual variation as it is needed in this study

have been derived from applied behavior analysis and its related

area, precision teaching.

















CHAPTER III

METHOD



Experimental Questions


This study addressed six questions as follows:

Questions 1-3: How does high rate of oral reading without inflection

(Condition HRNI) influence comprehension and mainten-

ance of comprehension compared with the following

conditions:

Condition LRNI: Low rate of oral reading without


Questions 4-5:


inflection?

Condition HRWI: High rate of oral reading with

inflection?

Condition LRWI: Low rate of oral reading with

inflection?

How does high rate of oral reading with inflection

(Condition HRWI) influence comprehension and mainten-

ance of comprehension compared with the following

conditions:

Condition LRWI: Low rate of oral reading with

inflection?












Condition LRNI: Low rate of oral reading without

inflection?

Question 6: How does low rate of oral reading without inflec-

tion influence comprehension and maintenance of

comprehension compared with low rate of oral reading

with inflection?

The format of the experimental questions lend themselves to

making a summary statement about which conditions had the greatest

influence on comprehension and maintenance of comprehension. This

can be found in the results chapter.



Experiment 1


Experiment 1 began in November, 1981, and the last day of data

collection was in February of 1982. The study was done to probe the

procedures and the design to see if they were adequate for observing

the suspected relationships between rate, inflection, comprehension,

and maintenance. Also, it seemed particularly important to test

ways to sequence the occurrences of the dependent measures. The

experimental questions listed above were used. Since the first

experiment yielded data that were systematic and orderly, it was

decided to continue this line of research. Experiment 2 contained

the necessary modifications that were developed as a result of having

performed experiment one. The findings of experiment one are included

in the results section.











Experiment 2


This study was designed to

experiment by using subjects of

Two other changes were made. A

dependent measures and a design

was changed from an ABCD design

of finding sequence effects.


systematically replicate the first

different age, sex, and skill level.

Cloze procedure was added to the

change was also made. The design

to one which increased the probability


Independent Variables


There were two independent variables. These were (1) reading

rate and (2) vocal inflection. Two rates of oral reading were used.

First, a proficiency reading rate of 150 to 200 words per minute

with two or less errors (Starlin & Starlin, 1973) was used. Secondly,

instructional rate or reading at 30 to 60 wpm with two or less

errors (Starlin & Starlin, 1973) was used. Vocal inflection was

defined as those audible qualities of oral reading that come under

the control of punctuation marks and make oral reading approximate

conversational speech.



Apparatus


During Experiment 1 a General Electric tape recorder (model

number 3-5311) was used to record oral reading with and without

inflection. A Casio MeIldy Alarm wrist watch was used for timing.

It has a stop watch function that records time to a tenth of a second.












The watch provides an auditory signal at 60 second inter-als. This

made it possible to make reliable measurements of frequency,

latency, and duration.

Experiment 2 used the same apparatus to serve the same

functions as in Experiment 1. The tape recording permitted the

experimenter to check and verify the onset of the independent

variables. The cassette recording helped the experimenter to check

the accuracy and reliability of recording oral reading with or with-

out inflection. The recordings were also used to record the verbal

responses to the different comprehension measures.



Dependent Variables


There were three dependent variables. They were (1) free recall

of facts, (2) answers to 10 comprehension questions, and (3) written

answers to a Cloze procedure (every fifth word was deleted). Compre-

hension and maintenance of comprehension were measured in exactly

the same way. Comprehension was defined as the score received on

each of the dependent measures immediately following criteria for

oral reading for each condition. Maintenance of comprehension was

defined by the three scores on each of the dependent variables after

intervals of three and 10 calendar days. Each subject received

three different measures of comprehension in each of the four condi-

tions, at three different times. This yielded a total of 36 measures

of the dependent variable that each of the four subjects received.











The dependent measures were taken in the following order:

free recall, ten comprehension questions, and then the Cloze

procedure. 'Jith this format the comprehension questions and the

Cloze procedure would not offer any cues or give answers to the

subject during the free recall episode. Since the Cloze procedure

used the entire reading story with every fifth word deleted, it

would have been possible for the subjects to recall facts that

otherwise would not have been recalled. To prevent this, the Cloze

procedure was always given last.

The ten comprehension questions consisted of the "what, when,

where, why, and how" variety. Whenever the textbooks listed compre-

hension questions for a story, they were included as part of the ten

questions. Questions other than the ones taken from the textbook

were developed by the experimenter. The Cloze procedure was written

with a word processor on the TRS 80 Model 1 microcomputer. The Okidata

Microline 82A dot matrix printer was used to print the text.

Each of the dependent variables was measured by using universal

dimensional quantities of behavior (Johnston & Pennypacker, 1980)

These are countability, latency, and duration. When behavior occurred

often within a particular response class, frequency was employed

When changes in the frequency, latency, and duration of behavior

occurred over time relative to the independent variables, a ratio

multiplier was used to describe the changes. For other measures, a

ratio index of change was used.











Subjects


For Experiment 1, two subjects from Gainesville Job Corps

were selected: a black student and a white student, ages 18 and

16 years, respectively. The 18 year old was on a sixth grade

reading level. The 16 year old was found to be on a second grade

reading level. These reading grade levels were determined by a

modified informal reading inventory taken prior to the experiment.

All the research was conducted within the subjects' regular reading

class at eight o'clock in the morning located at the Gainesville Job

Corps. The Open Court Basic Reader.was used with the 16 year old

(Open Court Basic Readers, 1972) and the SRA Reading Kit entitled

"Black Like Me" (Science Research Associates, 1976) was used with

the 18 year old.

For Experiment 2, four white females in the third grade were

selected from three different classrooms located within a school in

Alachua County, Florida. Subjects were selected randomly from a list

of children who, in addition to the above, met the following two

criteria: they must have been between the ages of eight and nine

years of age. They also had to have been reading on the third grade

level, which was determined by their completed work in the Ginn

Reading Series (Ginn, 1976). Prior to Experiment 2, the Inventory

of Basic Skills (Brigance, 1980) was used to assess each subject's

reading level. The purposes for using informal assessments have been

documented elsewhere (Alpar, Nowlin, Lemoine, Perine, & Bettencourt,

1974; Brigance, 1980; Lovitt & Hansen, 1976). The results indicated












that all subjects were on grade level. Reading materials for these

students were selected from a reading series which was grade appro-

priate. The students read entire stories from the Open Court

Reading Series (Open Court, 1972). Table 2 lists all the subjects

and describes the materials they each read in each condition.

Fry's (1972) readability formula was applied to increase the

probability that the stories from the reading series were grade

appropriate for the subjects. All materials used for these experi-

ments were on or near each subject's instructional reading level.

An attempt was made to insure that different reading passages would

appear at different conditions across subjects. For example, the

story in condition HRNI for subject three would also appear in

different conditions for subjects four through six. This was thought

to decrease the probability of experimental effects being attributable

to interest levels that subjects may have had for one story over

another.

The above differences stated between Experiment One and

Experiment Two subjects are in part the elements that comprised

a systematic replication of experiment one.



Design


A single subject design (Sidman, 1960) was used so that an

analysis at the individual level could be performed. This strategy

for conducting research has been well developed and documented

(Baer, Wolf, & Risley, 1968; Baily, 1977; Johnston & Pennypacker,












Table 2

Table of Subjects' Condition, Sequence, Reading
Passages, Number of Trails to Criteria, and Total Time in Minutes



Experiment 1 Subjects


Subject 1: 18 year old black male; Gr. Level 6

Readings from: SRA School House Kit Black Like Me.
Phase Sequence ABCD

Story Trials Time

Phase A: The Skeleton's Song. 8 78 min.
Phase B: The Burning Hand. 8 101' 29"
Phase C: In Jail. 11 85' 48"
Phase D: The Beauty of Me. 11 123' 33"

Subject 2: 16 year old white male; Gr. Level 2

Readings from: Open Court Basic Readers Level 2.
Phase Sequence ABCD

Phase A: The Camel's Nose. 12 55' 3"
Phase B: The Sphinx. 12 65' 48"
Phase C: Hot and Cold From One Mouth. 18 57' 16"
Phase D: The Monkeys. 18 63' 15"

Experiment 2 Subjects

Subject 3: 8 year old white female; Gr. Level 3

Readings from: Open Court Basic Readers A Trip Around the World.
Phase Sequence ABCD

Phase A: The Blind Men and The Elephant. 4 12' 8"
Phase B: The General and the Arrows. 4 21' 9"
Phase C: Antarctica. 7 26' 25"
Phase D: Cortez and Montezuma. 7 66' 24"











Subject 4: 8 year old white female; Gr. Level 3

Readings from: Open Court Basic Readers A Trip Around the World.
Phase Sequence CDAB

Story Trials Time

Phase C: Cortez and Montezuma. 8 48' 54"
Phase D: Antarctica. 8 64' 42"
Phase A: The General and the Arrows. 4 16' 54"
Phase B: The Blind Men and the Elephant. 4 33' 48"

Subject 5: 8 year old white female; Gr. Level 3

Readings from: Open Court Basic Readers A Trip Around the World.
Phase Sequence ABCD

Phase A: How the World Looks. 5 15' 18"
Phase B: Copernicus and Galileo. 5 34' 26"
Phase C: The Blind Men and the Elephant. 4 10' 25"
Phase D: The General and the Arrows. 4 21' 2"

Subject 6: 8 year old white female; Gr. Level 3

Readings from: Open Court Basic Readers A Trip Around the World.
Phase Sequence CDAB

Phase C: The Melting Pot. 7 47' 42"
Phase D: Cortez and Montezuma. 7 61' 5"
Phase A: How the World Looks. 5 22' 1"
Phase B: Antarctica. 5 29' 36"












1981; Sidman, 1960). The strategy allows for an interactive approach

between the experimenter and the independent variable. This is made

possible because N=1 designs allow for the identification of experi-

mental sources of intraindividual and interindividual variability

while the data are being generated by the subject (Pennypacker, 1982).

Also, this strategy permits the use of ratio comparisons that can be

made within experimental conditions and across experimental condi-

tions.

Since this study required that the independent and dependent

variables occur in a predetermined sequence, a design was used so

that if sequence and practice effects occurred, they could be

observed (Johnston & Pennypacker, 1980). This study involved four

conditions with two different sequences. Condition HRNI required

that the subjects read at a high rate without inflection. Condition

LRNI required oral reading at low rate and without inflection. Condi-

tion HRWI consisted of reading orally at a high rate and with correct

inflection. Reading at low rate with inflection comprised Condition

LRWI. The sequence of conditions for Subjects one, two, three, and

five was HRNI, LRNI, HRIII, LRWI. For Subjects four and six the

conditions were ordered HRWI, LRWI, HRNI, LRNI. This design also

made it possible to replicate the results across subjects and across

sequences. (See Figure 1.)

For the LRNI and LRWI conditions, the number of opportunities

to read a story was determined by the number of times stories were

read in HRNI and HRWI conditions, respectively. For example, the








57



















11


"--- ,











-j 5
C C l C' C'








D ::! -I 0 a'- '-
O 'i 1- 0- = 2

;= 3| *1 ^[



- z -. -


Sn;





1~C =! 1 i


0z

s!









!"~ ^i i-'

yl UC rS C
^ _O -I C -'--












number of times a subject read a story to meet reading rate criteria

in condition HRNI was the same number of times that the subject

could read a story in LRNI. In this way, practice effects were

equal so that the experimental effects of rate on the dependent

measures could be detected.



Procedure


The experimenter met with the two subjects in Experiment 1

every day for half hour each beginning at 8 a.m. The sessions took

place at a desk in the regular classroom at Job Corps. The experi-

menter met with the four subjects in Experiment two every instructional

day. Each subject was seen for one half hour between 8 and 10 a.m.

each day during the regular school week. The experimental sessions

took place behind the Media Center of the library at the school.


Pretraining Period

High rate for oral reading was defined as 150 to 200 wpm with

two or less errors. Reading a story at 30 to 60 words per minute

with two or less errors was considered the low rate for each subject.

To teach the idea that reading can be as fluent as conversational

speech, fluency training with an easier prerequisite skill was con-

ducted for all subjects. This is also known as tool skill training

(Haughton & Binder, 1982). Subjects were asked to say the letters of

the alphabet for one minute. Two to four of these timings were con-

ducted. Also, subjects were asked to think/say "Star Wars is one of












my favorite movies" (Wolking, 1982). This was also done to shorten

the number of trials needed to reach proficiency criteria (Haughton

& Binder, 1982). Tool skill development was done prior to reading

a story and at times just before rereading a story. For subjects

who displayed little growth in oral reading, their reading rate

criteria for conditions A and C were then reduced to 100 correct

words per minute with two or less errors. Little growth was defined

to mean those subjects who were accelerating less than 10 percent per

session and their rates were lower than 100 words per minute on each

reread for two or three days. This occurred for three of the six

subjects.

During oral reading, when a subject misread a word, the experi-

menter said the word aloud to the subject so that fluency would not

be disrupted. Furthermore, misread words were corrected and recorded

for later practice by the subject. These words were practiced every

day until they were read to a criterion of 40 to 60 per minute with

two or less errors.


Experimental Conditions

The high rate no inflection condition (HRNI) required the subjects

to reread a story orally and without inflection until 150 to 200 wpm

with two or less errors was reached. At times when the error rate

went above two per minute the experimenter modeled reading without

inflection. The modeling procedure was as follows:

(a) The experimenter asked the subject to read quickly and not

to pay attention to the punctuation marks.












(b) The experimenter then said that reading should sound like

the Federal Express commercial that they had seen on T.V. so that

their reading was fast and monotone and with equal emphasis on all

words.

(c) The experimenter then began reading a passage in a monotone

voice at a rate equal to 150 to 200 words per minute.

When criterion was achieved, measures of comprehension were taken

before starting the next condition. The number of trials to criteria

was recorded. This number was then used in a yoked manner in which

rereading in the LRNI condition only occurred for the same number of

trials that was required to reach criteria in the HRNI condition.

A new story for each subsequent condition was used. A measure of

the maintenance of comprehension of the story read in the HRNI

condition was then taken three and 10 calendar days from after the

end of HRNI condition. During the interval between the three and 10

days of the maintenance of comprehension measures for one condition

the next successive condition was in progress.

The LRNI condition required subjects to reread a story orally

without inflection at a rate of 30 to 60 wpm with two or less errors.

Modeling and verbal instructions were given similar to the instruc-

tions for the HRNI condition, except the subjects were asked to

read slowly. The experimenter modeled this by reading in a monotone

voice with equal emphasis on each word, at a rate of one word per

second, so that the 60 wpm criteria with two or less errors would

not be exceeded. When the requirements in the LRNI condition had

been met, measures of comprehension were then taken. Maintenance












of comprehension was assessed for tie LRNI condition three and 10

calendar days from the end of the condition.

Prior to the HRWI condition punctuation and inflection training

was provided. The punctuation marks that were taught and drilled

were the period, comma, quotation marks, exclamation mark, and the

question mark. First, naming punctuation marks for one minute was

a daily task prior to reading. Then the experimenter told the

subjects the function of each punctuation mark. Soon after, when

the experimenter pointed to each of the punctuation marks, the

subject told the examiner the function of each mark. This continued

until the subject was able to say the function of each punctuation

mark with 100 percent accuracy. Once this was completed, the subject

then pointed to each mark and then told the examiner the function

for each one until 100 percent accuracy was obtained. A probe sheet

with eight isolated phrases was used to practice oral reading at

high rates with correct inflection. The experimenter read these

phrases quickly and with the correct inflection. The subjects then

were asked to read these phrases as if they were speaking with some-

one. At first, daily one minute timings were used. Later, the

subjects reread these phrases as many times as they could in two

one-minute timings. Reading a phrase with the correct inflection uas

counted as one correct movement. If the subject inflected the

phrase correctly, but made an error in word calling the phrase was

still counted as correct.












In the HRWI and LRWI conditions an error was recorded when

incorrect inflection occurred during oral reading. The HRWI

condition involved rereading a story until criteria were reached

with correct inflection. When criteria were met, measures of com-

prehension were then taken. The maintenance of comprehension measure

for the HRWI condition story occurred three and 10 calendar days

after criteria had been met.

The LRWI condition required subjects to reread a story for the

same number of trials it took to reach criteria in the HRWI condi-

tion. When criteria were met, measures of comprehension followed.

Maintenance of comprehension was measured three and 10 calendar

days after the subject reached criterion.



Data Analysis--Dependent Variables


All performance data except for the answers to the 10 comprehen-

sion questions and percent correct on the Cloze procedure were

charted on a Standard Celeration Chart--a six cycle semiTogarithmic

chart (Pennypacker, Koenig, & Lindsley, 1972). (See Figure 2.)

Oral reading was plotted as words read correct and incorrect per

minute. Free recall required two levels of analyses. First, fre-

quency of correct and incorrect facts per minute recalled were plotted.

Second, words read per minute were divided by the total number of

facts recalled. This was done so that a measure of words read per

fact could be charted.





























O. O











. . . . . . .. . . . . .. .
I--.













O














A 0








- -, /
0 .. '9 -

O CC






a_ -=----
__-IL__ 0~i71-~
\ ~ i----- *-e _____
A ,'- --I






C-




-~.. ;zc P--- ____



-- C --. -= .-------
C-- ------- C
n9fNIJ 3d t=lO


4-,




C
ro
U




di c S~

'. I r LL
- C -o
widi



-4-

o,





z


C



C
Z

L)











For the ten comprehension questions, the latency to respond

to each of the ten questions and the count correct and incorrect

were plotted.

The Cloze procedure required two levels of analyses. First,

percent correct was recorded as the datum. Second, words written

correctly and incorrectly per minute were also plotted. To obtain

a rate correct per minute the total amount of seconds to complete

the Cloze test was divided into the number correct. This number

was then multiplied by 60 to get rate correct per minute. This was

also done for obtaining incorrects per minute.

The dependent measures of comprehension, maintenance after

three, and maintenance after 10 calendar days were compared. A

frequency multiplier was used to compare one condition with another

condition for measures of total performance (total performance =

rate correct + rate incorrect). An accuracy multiplier ratio was

used to describe accuracy ratio differences between conditions. A

total number correct for the 10 comprehension questions was used

to describe differences between conditions. A ratio index of change

was used to describe differences on the latency measures. The

measures for the Cloze procedure were the same for free recall.

In addition, a percent correct was used to describe differences

between conditions. (See Figure 3.)

A gain score for each of the independent variable conditions

was tabulated. This was done by counting the number of times the

dependent measures of a particular condition showed improvement over

the other conditions within the corresponding dependent measure.




















+ +

S 1 2 3 4 5 6 7 8 9 10

Instructional Days





A = Measures of comprehension immediately following criterion
performance for each phase.

+ = Measures of maintenance occurring 3 and 10 instructional
days following the end of each phase.











Figure 3

Sequence of Dependent Measures
(Comprehension and Maintenance)








66



For example, if the accuracy ratio for facts recalled per minute in

the HRWI condition for the three days after criteria measure was

greater than the number of facts recalled in the LRWI, HRNI, and

LRNI conditions for the same corresponding dependent measure, then

the HRWI condition would have a gain score of three for that dependent

measure. Then all of the gain scores are added together and pre-

sented as total number of gains for a particular condition on a

particular dependent measure.



















CHAPTER IV

RESULTS



This chapter is designed to present data as they relate to each

of the experimental questions. The data are presented in the order

they were gathered while doing the research. Therefore, the data

generated from Experiment 1 are presented first. The results of

Experiment 1 consist of the free recall measures (total performance,

accuracy, and words per fact), count correct on the comprehension

measures and the median latencies, respectively. The Cloze procedure

was used only in Experiment 2. The results of the Cloze procedure

include total performance, accuracy and percent correct, respectively.

Since the Cloze procedure occurred after the free recall measures and

answers to the 10 comprehension questions, it is the last dependent

variable presented in this chapter. For Experiment 1 there was a

total of five measures that occurred three different times for two

subjects during each of the four conditions. Therefore, a total of

120 measures for Experiment 1 was taken. Experiment 2 had a total

of eight measures that occurred three different times for each of the

four subjects during each of the four conditions. Therefore, there

is a total of 354 measures for Experiment 2.












Six questions in a comparison format were asked in this study.

Since there were four conditions, a 3 by 2 by 1 matrix is presented

as an aid to the reader when following the results of the comparisons.

(See Table 3.)



Questions 1-6 Results for Experiments 1 and 2
Free Recall Total Performance


Question One: (A comparison between two reading rates) How

does high rate of oral reading without inflection (HRWI) influence

comprehension and maintenance of comprehension compared with low rate

of oral reading without inflection (LRNI)?


Experiment 1 Results for Question One

Figure 3 shows the total performance (speed) for facts recalled

per minute for all six subjects.

Criteria performance--free recall. Subjects in Experiment 1

were D.D. and J.J. For these subjects the HRNI condition produces

a faster rate for each measure of the dependent variable than the

LRNI condition for facts recalled per minute. For D.D., HRNI

condition is faster for each of the maintenance measures. For

J.J. the HRNI condition produces a faster performance compared with

the LRNI condition. The table provided in Appendix A displays how

many times faster or slower a condition is when compared to another

condition. This number is called a frequency multiplier index. A

number with an "X" indicates that the referent condition is faster.

A number with a division sign means that the referent condition is
















Table 3

Graphic Display of Experimental Questions






Conditions


HRNI LRNI HRWI LRWI

Quest Quest Quest
1 2 3



Quest
4



Quest Quest
5 6


= High rate without inflection

= Low rate without inflection

= High rate with inflection

= Low rate with inflection


HRNI

LRNI

HRWI

LRWI












slower. From the table provided in Appendix B, it can be seen that

initially for 0.0. the HRNI condition is 1.25 times faster than the

LRNI condition. For J.J. the HRNI condition initially is 2.14

times faster than the LRNI condition.

Three day performance--free recall. Maintenance at three

days shows the HRNI condition to be 1.21 times faster than the LRNI

condition for subject D.D. For subject J.J. the HRNI condition is

2 1/2 times faster than the LRNI condition.

Ten day performance--free recall. Maintenance at ten days for

D.D. shows the HRNI condition to be 2.0 times faster than the LRNI

condition. For J.J. maintenance at 10 days the HRNI condition is

2 1/2 times faster than the LRNI condition.


Question One Results for Experiment 2

The subjects in this experiment were S.B., E.H., A.P., and

B.S.

Criteria performance--free recall. For these subjects the

HRNI condition initially produces a faster rate of recalling facts

than the LRNI condition. The frequency multipliers (see Appendix C)

for the HRNI condition are times X1.44, X1.27, X2.67, and XI.71

faster than the LRNI condition.

Three day performance--free recall. Three subjects under the

HRNI condition at the three day maintenance measures produce faster

rates compared with the LRNI condition. The HRNI condition influences

three subjects to produce slower rates, while the HRNI condition

influences E.H. to produce faster rates. The LRNI condition produces

a slower rate of responding for three subjects.












Ten day performance--free recall. It can be seen that for two

subjects, the HRNI condition at the ten day maintenance measure

influenced a faster response rate of recalling facts than the LRNI

condition.


Condition Two Results for Experiment 1

Question Two: (A comparison between inflection conditions)

How does high rate of oral reading without inflection (HRNI) influence

comprehension and maintenance of comprehension when compared with high

rate of oral reading with inflection (HRWI)?

Criteria performance--free recall. For both subjects the HRWI

condition initially produces a rate of recalling facts that is quicker

than facts recalled under the HRNI condition. From Appendix B it

can be seen that Subject D.D. under the HRWI condition is initially

1.4 times faster than his performance under the HRNI condition. And

the HRWI condition influences J.J. to be a 1.13 times faster than his

performance under the HRNI condition. (See Figure 4.)

Three day performance--free recall. For both subjects at

maintenance after three days the HRWI condition produces a faster

rate of recalling facts than the rates reported under the HRNI

condition.

Ten day performance--free recall. The HRLI condition produces

a rate for D.D. during the maintenance measure at 10 days that is

faster than those produced by the HRNI condition. However, J.J.

produces a faster rate under the HRNI condition than he does under

the HRWI condition.









72


















_-____: :T --- -- . e
N



-,____________,. _- Z ... o,





-_-. .-




.. '_--- __ [ ___- -_ ..._~._ -_ .-_- ~.. :_
1















-3-nN[ I d3d .. Nn0. =
--U i


1' ------ C









-J I- -


3IflNIN ~3d L.NflO












Question Two Results for Experiment 2

For S.B. and E.H. initially, the HRWI condition produces rates

of recalling facts that are 1.07 and 1.21 times faster than facts

recalled under the HRNI condition, respectively. From Figure 3 it

can be seen that initially, there are no differences between the

HRWI and HRNI for Subject A.P. Under the HRNI condition Subject

B.S. initially produces a rate that is 1.71 times faster than her

rate generated under the HRWI condition.

Three day performance--free recall. Maintenance at three days

shows two subjects, A.P. and B.S., under the HRII condition producing

a degree of performance that is faster than the HRNI condition. For

Subject S.B. the HR'WI and HRNI conditions produce a similar rate of

performance. Subject E.H., when the HRNI condition is in effect,

yields a rate of recalling facts that is faster than facts recalled

under the HRWI condition.

Ten day performance--free recall. The speed of performance for

maintenance at 10 days shows that for two of the subjects, S.B. and

E.H., produced faster rates under the HRWI condition than they did

under the HRNI condition. For B.S. rate of performance under the

HRWI and HRNI conditions are the same. The HRNI condition for

Subject A.P. produces a rate at ten days that was faster than her

rate under the HRII condition.


Question Three Results for Experiment 1

Question three: (A rate versus inflection comparison) How

does high rate of oral reading without inflection (HRNI) influence












comprehension and maintenance of comprehension when compared with

low rate of oral reading with inflection (LRWI)?

From Figure 4 it can be seen that under HRNI both subjects,

initially and after three days, are able to recall facts faster

than those under the LRWI condition for the same measures. At the

ten day measure, the HRNI condition for Subject J.J. produces a

faster rate. The HRWI and HRNI conditions yield similar rates for

Subject D.D.


Question Three Results for Experiment 2

Initially, the four subjects produced faster rates under the

HRNI condition compared with the LRWI condition for facts recalled.

From Appendix B it can be seen that performances under the HRNI

condition are 1.63, 1.56, 1.33, and 1.71 times faster, respectively,

than their performances under the LRWI condition.

Three day performance--free recall. For maintenance after three

days three subjects under the HRNI condition produced rates that are

faster than their rates under the LRWI condition.

Ten day performance--free recall. At the ten day maintenance

measures, the HRNI condition produces rates for two subjects that

are about 1.5 to 2.0 times faster than their rates under the LRWI

condition. For E.H. and B.S. the LRWI condition produces a faster

rate of recalling facts.












Question Four Results for Experiment 1

Question four: (A high and low rate comparison, with inflection)

How does high rate of oral reading with inflection (HRWI) influence

comprehension and maintenance of comprehension compared with low

rate of oral reading with inflection (LRWI).

Criteria performance--free recall. By comparing the LRWI condi-

tion with the HRWI condition Subjects D.D. and J.J. initially produce

faster recalls per minute under the HRWI condition. From the table

in Appendix B it can be seen that D.D.'s performance under the HRWI

condition initially is 1.62 times faster than his comparable measure

under the LRWI condition. For J.J., the HRWI condition produces a

rate that is b.7 times faster than his LRWI condition rate.

Three day performance--free recall. The HRWI condition for both

subjects produces a faster rate of recalling facts compared with the

LRWI condition at the three day measures of maintenance.

Ten day performance--free recall. For Subject D.D. the HRWI

produces a faster rate when compared with the LRWI condition at

the 10 day measures of maintenance. For Subject J.J. the HRWI

condition produces a frequency multiplier that is 1.25 times faster

than the LRWI condition.


Question Four Results for Experiment 2

For three subjects initially the HRWI condition produces a faster

rate of recalling facts per minute compared with the LRWI condition.

Subject B.S. under the HRWI condition produces the same rate as the

LRWI condition.


I












Three day performance--free recall. For three subjects,

maintenance after three days under the HRWI condition produces a

faster rate of recalling facts than under the LRWI condition.

Ten day performance--free recall. Three subjects produce a

faster rate of recalling facts under the HRWI condition compared

with the LRWI condition at the ten day measures of maintenance.


Question Five Results for Experiment 1

Question five: (A comparison between the presumed best and

worst conditions) How does high rate of oral reading with inflection

(HRWI) influence comprehension and maintenance of comprehension

compared with low rate of oral reading without inflection (LRNI)?

Subjects D.D. and J.J., under the HRWI condition, yield fre-

quency multipliers that range from a times 1.25 to a times 2.42

faster than rates that are produced under the LRNI condition, across

all measures.


Question Five Results for Experiment 2

For three subjects, the HRWI condition produces rates that are

faster across all measures in comparison with the LRNI condition.

Initially Subjects S.B., E.H., and A.P., under the HRWI condition

produce frequency multipliers that are 1.56, 1.55, and 2.5 times

faster than the LRNI condition.


Question Six Results for Experiment 1

Question six: (A comparison between inflection and no inflection

at low rates) How does low rate of oral reading without inflection












(LRNI) influence comprehension and maintenance of comprehension

compared with low rate of oral reading with inflection (LRWI)?

Criteria performance--free recall. Initially, for Subjects

D.D. and J.J., the low rate with inflection condition increases

their rate for recalling facts per minute.

Three day performance--free recall. This trend maintains for

J.J. at the three day maintenance measure. Subject D.D. produces

identical rates under conditions B and D at the three day maintenance

measure.

Ten day performance--free recall. For Subject D.D. the LRWI

condition produces faster rates when compared with the LRNI condition.

For Subject J.J. there is no difference between the speed of recalling

facts produced by either condition.


Question Six Results for Experiment 2

Initially, S.B. and E.H. under the LRNI condition produce faster

rates when compared with the LRWI condition. Subject A.P. produces

a faster rate of responding under the LRWI condition. Subject B.S.

produces the same speed of performance regardless of condition.

Three day performance--free recall. For three subjects, the

LRUI condition produces faster rates of performance.

Ten day performance--free recall. The trends are very similar

to those seen at the three day measure. That is, three subjects at

the ten day measure of maintenance produce faster rates under the

LRWI condition compared with the LRNI condition.











Accuracy Results


The following results are presented in the same format as the

performance data.


Question One Results for Experiment 1

Question one: (A comparison between two reading rates) How

does high rate of oral reading without inflection (HRNI) influence

comprehension and maintenance of comprehension compared with low rate

of oral reading without inflection (LRNI)? Figure 5 depicts the

accuracy ratios for all subjects.

Criteria recall--accuracy. From the table in Appendix C it

can be seen that, for Subjects D.D. and J.J. under the high rate

without inflection condition (HRNI), they initially produce accuracy

ratios that are 1.3 and 2.5 times more accurate compared with the

low rate without inflection condition (LRNI).

Three day recall--accuracy. For Subject J.J. the HRNI condi-

tion continues to produce an accuracy multiplier that indicates a

more accurate performance compared with the LRNI condition. This

trend does not maintain for Subject D.D.

Ten day recall--accuracy. Subject D.D. produces more accurate

responding under the HRNI condition at the 10 day measures of mainten-

ance compared with the LRNI condition.


Question One Results for Experiment 2

Initially, all subjects have a greater accuracy multiplier index

under the high rate without inflection condition (HRNI) compared with

the low rate without inflection condition (LRNI). From Appendix C,






























:i0
-s - . -









_ ---_ 1 .. . .. -
/r \




U --- -:.- ~ --- "-_ --




_:----__ -: -- .















-0-



_. _o 1
32*INI* --v+ =_ 4
'-' -----~-

~ ~ Th- -~ ~ _____

-C, -- ., i-


U
L

U



Y '~
-
u




'





U
a s i
u Ll O

y .-.
-








S'^











these subjects produce accuracy multiplier indices under the HRNI

condition that range from a times 1.44 to 3.2.

Three day recall--accuracy. For the three day maintenance

measure, three subjects produce more accurate responding under the

HRNI condition compared to the LRNI condition.

Ten day recall--accuracy. Two subjects under the HRNI condi-

tion produce more accurate responding compared with the LRNI condition.

For one subject there is no difference.


Question Two Results for Experiment 1

Question two: (A comparison between inflection conditions)

How does high rate of oral reading without inflection (HRNI) influence

comprehension and maintenance of comprehension compared with high

rate of oral reading with inflection (HRWI condition)?

From Figure 5 it can be seen that responding for Subjects D.D.

and J.J., under the HRWI condition, yields a greater accuracy ratio

compared with the HRNI condition. This trend is seen across all

three temporal measures of accuracy.


Question Two Results for Experiment 2

For Subjects S.B. and E.H. initially the HRWI condition produces

a greater accuracy ratio compared with the HRNI condition. For A.P.

there is no difference. Subject B.S. responds under the HRWI condi-

tion by producing a better accuracy ratio compared with the HRNI

condition.












Three day recall--accuracy. For maintenance at three days,

Subjects A.P. and B.S. under the HRWI condition produce a greater

accuracy ratio compared with the HRNI condition. There is no

difference observed for Subject S.B. Subject E.H. under the HRWI

condition produces a greater accuracy ratio compared with the HRNI

condition.

Ten day recall--accuracy. For two subjects, the HRWI condition

produces more accurate responding at the ten day measures of main-

tenance compared with the HRNI condition. The same measure for Subject

S.B. reveals a no difference effect. Figure 5 shows that for E.H.,

a steeper decelerating slope is present under the HRWI condition when

compared to the HRNI condition.


Question Three Results for Experiment 1

Question three: (A rate versus inflection comparison) How does

high rate of oral reading without inflection (HRNI) influence compre-

hension and maintenance of comprehension compared with low rate of

oral reading with inflection (LRWI)?

Criteria recall--accuracy. Subject D.D. initially responds

under the LRWI condition by producing an accuracy multiplier that

is 2.0 times greater compared with the HRNI condition. For Subject

J.J., initially the HRNI condition produces an accuracy ratio that

is 3.75 times greater than the one produced by the LRWI condition.

Three day recall--accuracy. Maintenance after three days for

D.D. is similar to the trends found under the initial measurement.

At three days, Subject J.J. under the HRNI condition produces an











accuracy multiplier that indicates a more accurate performance

compared with the LRNI condition.

Ten day recall--accuracy. At the ten day maintenance measure,

Subject D.D. under the HRNI condition produces an accuracy multiplier

of 3.25 compared with the LRWI condition. For J.J. a reverse trend

occurs. That is, the HRNI condition produces less accurate per-

formance compared with LRWI condition.


Question Three Results for Experiment 2

From Figure 5 it can be seen that initially more accurate per-

formance is produced by all subjects under the HRNI condition. The

accuracy multipliers range from a times 1.33 to a times 1.71.

Three day recall--accuracy. For three subjects, the HRNI

condition produces more accurate performances compared with the

LRWI condition at the three day measures of maintenance.

Ten day recall--accuracy. The ten day measure describes two

subjects under the HRNI condition as yielding more accurate performances.

For B.S. there is no difference, while for A.P. the LRWI condition

results in more accurate responding compared with the HRNI condition.


Question Four Results for Experiment 1

Question four: (A high and low rate comparison, with inflection)

How does high rate of oral reading with inflection (HRWI) influence

comprehension compared with low rate of oral reading with inflection

(LRWI)?











For both subjects, Figure 5 depicts high rate of oral reading

with inflection (HRWI condition) to produce more accurate respond-

ing across all temporal measures compared with the LRWI condition.


Question Four Results for Experiment 2

As seen in Figure 5, three subjects under the HRWI condition

initially produce more accurate performances compared to the LRWI

condition. For Subject B.S. there is no difference.

Three day recall--accuracy. Accuracy at three days under the

HRWI condition produces more accurate performances for three subjects.

Subject B.S. produces the same accuracy ratio under either condition.

Ten day recall--accuracy. The ten day maintenance measures

for Subjects S.B. and E.H. under the HRWI condition produce more

accurate responding compared with the LRWI condition. For B.S.

there is no difference between the two conditions. This is also

true for Subject A.P.


Question Five Results for Experiment 1

Question five: (A comparison between the presumed best and

worst conditions) How does high rate of oral reading with inflection

(HRWI) influence comprehension and maintenance of comprehension compared

with low rate of oral reading without inflection (LRNI).

Subjects D.D. and J.J. generate data under the high rate with

inflection condition that have a greater accuracy ratio when compared

with the low rate without inflection condition for all measures of

accuracy.











Question Five Results for Experiment 2

The HRWI condition initially produces more accurate responding

for three subjects. For Subject B.S. there is no difference.

Appendix C shows the accuracy multipliers range from 1.0 to 3.77.

Three day recall--accuracy. Four subjects under the HRWI

condition produce more accurate responding at the three day main-

tenance measure compared with LRNI condition.

Ten day recall--accuracy. More accurate responding are also

produced by three subjects while under the HRWI condition at the 10

day measures of maintenance.


Question Six Results for Experiment 1

Question six: (A comparison between inflection and no inflection

at low rates) How does low rate of oral reading without inflection

(LRNI) influence comprehension and maintenance of comprehension

compared with low rate of oral reading with inflection (LRWI)?

Criteria recall--accuracy. For Subject D.D. the LRWI condition

initially produces an accuracy ratio that is 2.4 times greater than

the ratio produced by the LRNI condition. Subject J.J. generates

an accuracy ratio that is greater for the LRNI condition compared

with the LRWI condition.

Three day recall--accuracy. At three days both subjects are

producing a greater accuracy ratio under the LRWI condition.

Ten day recall--accuracy. Responses by D.D. under the LRNI

condition yields a greater proportion of number correct to incorrect

when compared to the LRWI condition at the 10 day measure. For

Subject J.J. there are no differences in the accuracy of responding.












Question Six Results for Experiment 2

Initial responding by Subjects E.H. and A.P. under the LRWI

condition produce a better accuracy ratio compared with the LRNI

condition. For Subject B.S. there are no differences.

Three day recall--accuracy. For three subjects at the three

day measures of maintenance the LRWI condition is producing a better

accuracy ratio compared with the LRHI condition.

Ten day recall--accuracy. At the 10 day measure, two subjects

under the LRWI condition produce more accurate responses compared to

the LRNI condition. For Subjects S.B. and B.S. the LRNI condition

produces more accurate performances.



Free Recall--Words Per Fact


Figure 6 depicts six subjects on their measures of the number

of words read for each fact retained.

For this measure, a lower value indicates that while reading

less words, subjects were recalling more facts. From Figure 6 it

can be seen that the high rate conditions initially have lower

values for all six subjects when compared to the low rate condi-

tions. Three subjects are maintaining this trend at the three day

measures of maintenance. At 10 days, four subjects maintain this

trend while one subject produces no discernable differences.

Responding under the high rate condition with inflection more

frequently produces a lower value (more facts per word) compared

to the other conditions. Except for the HRWI condition, responding









86



















________ -0ll _-j -
r 7




-77


-- ___

r~-, ( ...
-- -- -























y Co
--.i.~~-- ---- _

-Cj o0




Full Text
EFFECTS OF ORAL READING RATE AND INFLECTION
ON COMPREHENSION AND ITS MAINTENANCE
BY
HENRY A. TENENBAUM
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOP THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
I 983

Copyright ”1983
by
Henry A. Tenenbaum

ACKNOWLEDGEMENTS
A doctoral dissertation and all the steps along the way to a
doctorate require one to commit much of his energy and lifestyle
toward this goal. For me, this dissertation was greater than any
one individual. Several people have contributed directly to my
dissertation and others have given me their encouragement and per¬
spective that I needed to succeed.
My acknowledgements must begin with my parents, Jack and Bertha.
I will probably never know the extent of their suffering when their
lives were threatened and when they were separated from each other
during World War II. You would think they would be embittered people
who had given up their desire to be a constructive part of society.
Yet, when they escaped the torments that existed for them in Europe
during and after the War, they stood strong and proud to become
Americans. From their pride and eagerness to begin a new life, they
provided me with the devotion, love, and their constant nudging to
go cn and to become the best of what I wanted to be. One may never
be able to provide parents such as mine with all the acknowledgements
due them. But, I will continue to live my life in a manner that will
continually make then proud.

I must also acknowledge my sister, Sara, and my brother,
Theodore. Throughout my life they have always provided me with
another view and the necessary wisdom to make the best choices
concerning my career.
Probably one of the most difficult tasks throughout a doctoral
program is maintaining a fulfilling relationship with your spouse.
My wife, Sarah, never once strayed from her dedication and love for
me. The sacrifices that she made and her understanding of what was
needed to pursue a doctorate were an inspiration to maintain my
motivation and perseverence.
My dedication and interest in the behavioral sciences were not
shaped by accident. Dr. Robert E. Anderson gave me the opportunity
to work with children and encouraged me to continually increase my
skills on the graduate level. My friends, Kay Kaldor and John
Carrier, provided me with many of the fundamental principles that
were needed to work with children successfully. I will always be
grateful for the time and patience they have shown me.
The people who have contributed more directly to my doctorate
will be remembered. As life is, people come under different con¬
tingencies and leave their friends. But comfort can be had when one
knows that whenever we meet or need to be close again, the likeli¬
hood of its occurrence is great.
My closest friend and the one who influenced me the most was
my chairman, Bill Wolking. Bill is a remarkable human being who
seems to thrive on friendship and his dedication to students. Bill

has given me unselfish support and the guidance to know what an
empirical science of behavior should entail. Bill was able to not
only maintain my interest in applied behavior analysis and precision
teaching, but to widen my interests and to develop my curiosity of
the world outside of the university halls. He is a magnificent
person and I will continue to seek his leadership throughout my
career.
In any doctoral program one becomes influenced by many people.
But, Hank Pennypacker was one of those people along with B. F.
Skinner whose writings were continua'My used as a fundamental guide
to enhance my understanding of the principles involved in a science
of behavior. To him I will be grateful for the wisdom that he has
given me and to humankind.
Special thanks go to my other committee members Ed Turner
and Bert Sharp. As a committee member one is constantly being asked,
along with everything else, to review manuscripts, attend meetings,
give counseling, and to provide leadership as it is needed. Without
Ed and Bert my doctoral work could have been a much more difficult
task.
While working towards my doctorate others outside of my
committee went above and beyond to provide me with much needed
support. These people were typically part of the Special Education
Department. To these people I give my thanks. More specifically
my gratitude goes to Dr. Cecil Mercer who many times offered and gave
his guidance. Special thanks also go to Dr. Bob Algozzine who always
v

provided me with the humor, encouragement, and his attendance at
the meetings needed to complete my degree.
I also wish to express my thanks to many of my friends who I
have grown close to and who never failed to ask whether or not I
needed their help in some way. These people have already completed
their doctoral degrees or are pursuing doctorates themselves. My
appreciation goes to Bonnie Engel for the many years of encourage¬
ment she gave me. My appreciation goes to Leonard Weiss who kept
me interested in my studies by offering his professional respect and
by reaching out to maintain a long friendship with me. Special
thanks go to Charles Hughes who helped me to express many ideas
and, whenever he could, reminded me that I needed to graduate.
Another colleague who always made me feel in charge and helped to
maintain my motivation was Sue Peterson. To her I send very warm
thanks. To all the people who I have not mentioned, but who influenced
my life in many subtle ways, I give my appreciation.
Lastly, my love goes to all the children past and future who may
never have the opportunity to pursue a doctorate. Through the child¬
ren I have met, each one helped me to better understand human behavior.
They truly were my inspiration.
vi

TABLE OF CONTENTS
ACKNOWLEDGEMENTS iii
ABSTRACT xiii
CHAPTER I INTRODUCTION 1
Rationale 1
Significance 4
Statement of the Problem 5
Delimitations 5
Definition of Terms 6
Summary 7
CHAPTER II REVIEW OF RELATED LITERATURE 9
The Nature of Reading Comprehension 10
Reading Comprehension as a Multifaceted Skill 10
Listening and Reading Comprehension 11
Reading Comprehension as Facts 12
Oral vs. Silent Reading: Which Way Maximizes
Comprehension? 12
Theories, Strategies, and Tactics of Reading and
Comprehension 13
Traditional Teaching Methods 13
Recent Strategies and Tactics for Teaching Reading and
Comprehension 18
Summary 26

Dependent Measures of Reading Comprehension 28
Reading Comprehension and Memory 31
Summary 32
Oral Reading, Rate, Punctuation, and Comprehension 33
Summary 36
Verbal Behavior 37
Experimental Analysis of Behavior and B. F. Skinner ... 38
Stimulus Control 41
Comprehension Questions and Free Recall as Discriminative
Stimuli for Intraverbals 42
Summary ' 44
CHAPTER III METHOD 47
Experimental Questions 47
Experiment 1 48
Experiment 2 49
Independent Variables 49
Apparatus 49
Dependent Variables 50
Subjects 52
Design 53
Procedure 58
Pretraining Period 58
Experimental Conditions 59
Data Analysis—Dependent Variables
vi i i
52

CHAPTER IV RESULTS 66
Questions 1-6 Results for Experiments 1 and 2
Free Recall Total Performance 67
Experiment 1 Results for Question One 67
Question One Results for Experiment 2 69
Condition Two Results for Experiment 1 70
Question Two Results for Experiment 2 71
Question Three Results for Experiment 1 71
Question Three Results for Experiment 2 74
Question Four Results for Experiment 1 75
Question Four Results for Experiment 2 75
Question Five Results for Experiment 1 76
Question Five Results for Experiment 2 76
Question Six Results for Experiment 1 76
Question Six Results for Experiment 2 77
Accuracy Results 73
Question One Results for Experiment 1 78
Question One Results for Experiment 2 78
Question Two Results for Experiment 1 80
Question Two Results for Experiment 2 80
Question Three Results for Experiment 1 81
Question Three Results for Experiment 2 82
Question Four Results for Experiment 1 82
Question Four Results for Experiment 2 83
Question Five Results for Experiment 1 83
Question Five Results for Experiment 2 84
ix

Question Six Results for Experiment 1 84
Question Six Results for Experiment 2 85
Free Recall--Words Per Fact 85
Summary Free Recall Questions 1 6: Total Performance,
Accuracy, and Words Per Fact 87
Ten Comprehension Questions 88
Question One Results for Experiment One 88
Question One Results for Experiment Two 88
Question Two Results for Experiment 1 91
Question Two Results for Experiment 2 91
Question Three Results for Experiment 1 92
Question Three Results for Experiment 2 93
Question Four Results for Experiment 1 93
Question Four Results for Experiment 2 93
Question Five Results for Experiment 1 94
Question Five Results for Experiment 2 94
Question Six Results for Experiment 1 95
Question Six Results for Experiment 2 95
Latency Measure Results 96
Summary of Results: Dependent Measures of Comprehension . . 96
Results Cloze Procedure--Total Performance 98
Question One Results 98
Question Two Results 99
Question Three Results 101
Question Four Results 102
Question Five Results 103
Question Six Results 103
x

Cloze Procedure Results--Accuracy 104
Question One Results 104
Question Two Results 106
Question Three Results 106
Question Four Results 107
Question Five Results 107
Question Six Results 108
Results Cloze Procedure-Percent Correct 109
Summary of Results--Cloze Procedure 109
Summary of Results Experiments One and Two Ill
CHAPTER V DISCUSSION 114
Variability 118
Replications 120
Practical Implications 124
Implications for the School Psychologist 124
Generalizations 125
Suggestions for Future Research 126
REFERENCES 129
APPENDIX A FREQUENCY MULTIPLIERS, TOTAL PERFORMANCE, FREE
RECALL 138
B ACCURACY RATIO MULTIPLIERS, FREE RECALL .... 145
C FREQUENCY MULTIPLIERS, TOTAL PERFORMANCE
CLOZE PROCEDURE 152
D ACCURACY RATIO MULTIPLIERS, CLOZE PROCEDURE . . 157
E 10 SAMPLE COMPREHENSION QUESTIONS, CLOZE
PROCEDURE 161
xi

APPENDIX F POST-EXPERIMENTAL INTERVIEWS 165
G RAW DATA 174
BIOGRAPHICAL SKETCH 190
x i i

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
EFFECTS OF ORAL READING RATE AND INFLECTION ON
COMPREHENSION AND ITS MAINTENANCE
By
Henry A. Tenenbaum
August, 1983
Chairperson: William D. Wolking
Major Department: Counselor Education
Rigorous experimental analysis of the effect of oral reading
rate on comprehension has only recently been performed. There is
still controversy over which rate of oral reading maximizes compre¬
hension and retention. Inflection and how it interacts with rate
of oral reading have received very little attention.
This study sought to determine how oral reading rate, when
approximating functional conversational speech (150 to 200 words/
minute) and inflection, impacts comprehension and maintenance of
comprehension. A single subject design was developed to determine
how oral reading at 150 to 200 words/minute with inflection compared
with oral reading at 40 to 60 words/minute (instructional rate), with
inflection on measures of comprehension and maintenance of comprehen¬
sion. High oral reading rates without inflection were also compared
with low oral reading rates without inflection to determine their

effects on comprehension and maintenance of comprehension. The
dependent variables were a free recall task, answers to 10 compre¬
hension questions and written responses to a Cloze procedure.
These occurred immediately following reading criteria, and at three
and 10 days after criteria was reached.
Six subjects were used in this study; two subjects were of high
school age and reading below grade level and four subjects were in
the third grade reading on grade level. For this study an ABCD
design was used with four subjects and a CDAB design was used for
two subjects so that any effect that sequence may have had could be
determined. Also, the high rate conditions were yoked to the low
rate conditions to keep the number of trials equal.
The results confirmed that the combination of high oral reading
rate with inflection (when reading approximates conversational
speech), increased both the accuracy and speed of comprehension and
its maintenance more than any of the other combinations. The combina¬
tion of high oral reading without inflection was found to increase
comprehension and maintenance of comprehension when compared with
low oral reading rate with and without inflection. Also, inflection
training in both the high and low rate oral reading conditions
improved comprehension. The results of Experiment 2 systematically
replicated the results of Experiment 1 across reading levels, reading
passages, settings, and subjects.
xiv

CHAPTER I
INTRODUCTION
This is a study of how inflection and oral reading rate affect
comprehension and maintenance of comprehension. The impact that
inflection may have on comprehension and maintenance of comprehen¬
sion has not received extensive investigation (Weaver, Holmes,
Curtis, & Reynolds, 1970). Many reading experts agree that reading
with the correct inflection is an important skill (Durrell, 1949;
Heilman, 1967; Spache & Spache, 1977). The importance of reading
fluently has also been considered a major variable for developing
comprehension by only a few investigators (Lovitt 5 Hansen, 1976;
Smith, 1972). Yet, research reported has not addressed the relation¬
ship between both levels of fluency and levels of inflection on
comprehension and maintenance of comprehension.
Rationale
There is a concern that there has been a paucity of research on
how to develop and improve reading comprenensicn (Armstrong, 1981;
Turner S Fey, 1377; Van Etten, 1973). The idea behind this investigation
came forth as a result of discussing the concepts in verbal behavior
1

2
(Skinner, 1957) with others. In beginning an investigation in
reading comprehension, the concepts embedded in Skinner's (1957)
analysis of verbal behavior appeared to be a good guide for research
in this area.
Skinner (1957) wrote about reading comprehension as a particular
instance of verbal behavior. Within his analysis, understanding
written language is similar to listening comprehension. They are
similar in that they both need a preceding set of events for either
one of them to occur. Also, both kinds of comprehension must be
functional to maintain their occurrence (i.e., produce benefits
for the reader).
A difference between listening comprehension and reading compre¬
hension is that listening comes under the control of another person's
vocal behavior while reading comes under the control of words. People
with normal hearing will probably have a greater chance to receive
reinforcement from listening than from reading. In other words,
listening appears to have a longer history of reinforcement than
does reading. Also, it is easier to have access to another person
who will engage in conversation than it is to find a person with whom
one can discuss what one has read. Another difference appears to be
the rate at which conversational speech occurs as compared to reading
rate. People have been trained to listen and understand conversations
(which include inflection) of 150 to 200 words per minute (Hoiking,
1530). The average elementary aged school student is trained to read
only 40 to 90 words per minute (Lovitt, 19S2).

Earlier researchers demonstrated that receptive vocabulary
is prerequisite for comprehension of written words (Sidman, 1971;
3
Staats & Staats, 1963). Also, there is a body of research to
suggest that when the rate of reading approaches conversational
speed, there is an improvement in comprehension (Lovitt & Hansen,
1976) and retention (Berquam, 1981). Although inflection has been
implicated as an important variable for the development of compre¬
hension, the literature contains very little on how rate and inflec¬
tion affect comprehension and the maintenance of comprehension.
From the above concepts and research an empirical question
concerning rate and inflection developed. Taking Skinner's (1957)
analysis of reading and listening comprehension, perhaps reading
comprehension can be improved when the reading episode approximates
conversational behavior. In other words, can comprehension and
maintenance of comprehension improve when reading sounds similar to
the way people speak? If so, students can learn that reading has
functional value. What is said can be equivalent in function to what
is read. To answer the question two independent variables will need
to be controlled. These are oral reading rate and inflection.
To determine the impact that the independent variables have on
comprehension and maintenance of comprehension, a single subject
design was used. Single subject research has been proven to be a
successful strategy in observing and subsequently controlling for
individual variation in responding (Johnston & Pennypacker, 1980;
Sidman, 1960). Precision Teaching will be used as a way to teach

4
and measure inflection, reading, comprehension, and the maintenance
of comprehension (Armstrong, 1981; Berquam, 1981; Hansen, 1979;
Haring, 1978; Wolking, 1982).
Significance
Society views reading and reading comprehension as an important
measure of the effectiveness of public school instruction (Haring,
1978). The importance of reading with comprehension cannot be
underestimated (Turner & Fey, 1977). This study attempts to describe
one or more variables that may have a positive effect on the develop¬
ment of comprehension and the maintenance of comprehension. Hope¬
fully, this will increase the efficiency with which reading comprehension
comes under the control of teacher-environmental arrangements.
There appears to be a correlation between the illiteracy index
and public opinion. The 13th Annual Gallup Poll, funded by Phi
Delta Kappa, reported that 20 percent of the parents of children
attending public schools gave the schools a grade of D or F (Kappan,
1980). This was the highest percentage of failing grades given in
eight years. A teaching strategy that uses effective and accountable
techniques may help to improve the reading performances of students.
More than likely, when the reading performances improve significantly,
suoport for public education will increase. Familiarity with the
mechanisms that enhance reading and comprehension will give school
personnel the opportunity to take a leadership role in enhancing
the accountability of public school instruction and improving public
opinion.

5
Statement of the Problem
The present study investigates the effect of oral reading rate
with and without inflection has on the dependent variables, compre¬
hension and maintenance of comprehension.
This study also serves to extend Lovitt and Hansen's (1976)
findings. In their study, high oral reading rate was not compared
with other rates. Also, they did not report any control for inflec¬
tion. This study also systematically extends Berquam's (1981) findings
that high rate increases retention, by replication with a different
skill.
Delimitations
Comprehension has been defined by many investigators as having
several components. This study will only describe the relationship
of the independent variables with measures of comprehension and
maintenance of comprehension. The dependent variables measure
instances of recall. This limits the generality of the findings
across other components of comprehension. However, as already noted,
increased oral reading rates resulted in improved comprehension
across other components (Lovitt & Hansen, 1976).

6
Definition of Terns
Accuracy Ratio: This is a measure of accuracy derived from the
rat.¡0 Frequency Correct
Frequency Error '
Behavior: The behavior of an organism is that portion of the
organism's interaction with its environment that is characterized
by detectable displacement in space through time of some part of the
organism and that results in a measurable change in at least one
aspect of the environment (Johnston & Pennypacker, 1980).
Celeration: This is the basic unit of measurement to describe
behavior change; change in frequency per unit time.
Comprehension: Comprehension is an instance of verbal behavior
known as an intraverbal. An intraverbal is a verbal response in
which the prior controlling variable is a verbal stimulus with no
point-to-point correspondence between the stimulus and the response.
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 equivalent to rate in this
definition.
Inflection: A change in pitch or loudness of the voice. The
change of form that words undergo to mark case, gender, number,
tense, person, mood, or voice (Merriam-Webster, 1974). In addition
to the above definition, inflection in oral reading comes under the
control of punctuation marks.

7
Latency: The amount of time between the occurrence of a
signal and the beginning of a movement.
Learning: A change in performance per unit time; also called
celeration.
Level of difficulty of material: Level of difficulty of
material is defined by the frequency at which the student demon¬
strates performance of 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).
Readability: Readability is defined by Fry's readability formula.
This formula will provide an estimation of the grade level of reading
material.
Response Class: A response class is a set of responses that
have-at least one characteristic in common. These characteristics
may involve the 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 a hand, foot, knee, or body, but the
result is the same on the environment) (Whaley 4 Malott, 1971).
Standard Behavior Chart: A standard six-cycle semi-logarithmic
chart that displays frequency as movements/time and celeration as
movements/minute/week.
Summary
Experimental questions were developed to determine whether oral
reading, when it approximates conversational speech, can enhance

reading comprehension and maintenance of comprehension. More
specifically, the purpose of this investigation is to understand
the interaction of oral reading at two rates with and without
inflection as it impacts reading comprehension and maintenance of
comprehension. Precision Teaching, single subject technology, and
Skinner's Verbal Behavior paradigm, are the strategies that will be
employed to engineer the research design, data analysis, and conclu¬
sions.
Special educators who are interested in the area of reading
have only recently begun to determine the efficácy of oral reading
in enhancing comprehension. Oral reading at high rates has been
shown to have a positive effect on reading and reading comprehension
(Lovitt & Hansen, 1976). Studies that have investigated the effects
of varying reading rates on retention have suggested a positive
relationship between high rates of performance and retention of facts
(Berquam, 1981). This study is a systematic replication of Lovitt
and Hansen's study (1976) and also looks at the effects of inflection
on reading comprehension and its maintenance over time.

CHAPTER II
REVIEW OF RELATED LITERATURE
The following review is divided into six sections. The first
section focuses on the nature of reading comprehension. The second
section describes theories, strategies, and tactics that relate to
reading comprehension. The third section describes the most typical
methods of assessing the dependent measure, reading comprehension.
The fourth section refers to how oral reading, rate of oral reading,
and punctuation affect reading comprehension. Section five describes
reading comprehension as an instance of verbal behavior (Skinner,
1957). Although each main section has a summary, section six is a
summary of the previous main sections.
Only those studies and position papers that were relevant to
the experimental questions were reviewed. First, oral and/or silent
reading must be the major independent variable. Second, studies
which contained at least one of the following dependent measures of
reading comprehension were chosen: answers to comprehension questions,
Cloze procedures, and free recall. Third, studies that were performed
within a laboratory setting were sought. This was thought to be an
efficient way to provide the reader with studies that have undergone
experimental analysis. Moreover, by reviewing these studies, one
9

10
begins to set the occasion to systematically replicate known
laboratory principles of reading comprehension. So few laboratory
studies were found that research conducted in applied settings with
less than adequate control had to be included.
The Nature of Reading Comprehension
A definitive statement as to what constitutes reading comprehen¬
sion and how to teach it is hard to come by. It is no wonder then
that the literature contains very little experimental data on how to
improve reading comprehension (Armstrong, 1981; Van Etten, 1978;
Vogel, 1975).
Reading Comprehension as a Multifaceted Skill
Reading comprehension has usually been conceptualized as a complex
skill. There are many definitions of this skill. Smith (1969)
outlined seven categories of reading comprehension by using the
Taxonomy of Educational Objectives (Bloom, Engelhart, Furst, Hill,
and Krathwohl, 1956). These categories are memory, translation,
interpretation, application, analysis, synthesis, and evaluation.
The use of these categories has not gone without criticism.
Smith (1969) reported that the list was too large and proposed
a four level approach to reading comprehension. These were literal,
interpretive, critical, and creative reading comprehension. Miller
(1976) also believed that most definitions and components of reading
comprehension were too long. He described four types of reading

n
comprehension. These were (a) recalling information from long-term
memory, (b) elaborating and abstracting, (c) interpreting, and (d)
naming. In addition to these categories, listening comprehension
and word comprehension have also been included as a definition of
reading comprehension (Lerner, 1976; Vogel, 1975, 1977).
Listening and Reading Comprehension
Studies (Sidman, 1971; 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; Sidmarr, 1971).
Some differences between reading and listening are (a) listen¬
ing (language or speech) is typically acquired first, (b) the process
of speech acquisition is more gradual (Staats & Staats, 1962), (c)
speech acquisition is under the control of stronger reinforcers
(Staats & Staats, 1962), and (d) speech is typically taught in a
cne-to-one setting (Bijou S Baer, 1961). The use of vocal language
is usually learned before reading. Reading comprehension is strongly
influenced by spoken language (Staats & Staats, 1962). Along the
same lines matching-to-sample (visual and/or auditory equivalence
trainingjy has been used as a definition of reading comprehension
(Sidman, 1971; Wolking & Greenwood, 1979). This is in line with
Skinner's (1957) understanding of comprehension. In his classic
book Verbal 3ehavior, Skinner (1957) defines reading comprehension
as an intraverbal. For example, "A person comprehends a text when
he can describe it in different words" (Johnson & Chase, 1981, p.
110).

12
Reading Comprehension as Facts
The most widely used measure of comprehension in the public
schools is the ability to remember specific facts from what has been
read (Guszak, 1969; Johnson & Chase, 1981; Rystrom, 1970). The
ability to recall facts does appear to be a fundamental prerequisite
for all other measures of comprehension. Without the ability to
recall facts, concept formation and other abstractions will be
limited (Johnson & Chase, 1981). Thus, one might propose that train¬
ing a reader to recall facts should be the primary comprehension
task.
Oral vs. Silent Reading: Which Way Maximizes Comprehension?
During the early 1900's, E. L. Thorndike recommended that an
exclusive emphasis on precise pronounciation and oral reading was not
adequate for developing comprehension (Thorndike, 1917). His conclu¬
sions initiated a greater awareness and concern for reading comprehen¬
sion and silent reading; however, data to support his contention were
not provided. Yet, the emphasis shifted from oral reading to silent
reading within the public schools. Along with silent reading, an
emphasis on decoding words has been used with some research indicating
gains in comprehension (Rystrom, 1970). However, a new body of
research has been supporting the use of oral reading to increase
comprehension.
This new research suggests that when oral reading rate reaches
to a standard of performance, comprehension improves (Armstrong,
1981; Hansen, 1973; Lovitt & Hansen, 1976). Similarly, Samuels

13
(1979) demonstrated the effectiveness of using a standard within
his repeated readings paradigm was an effective way to increase
comprehension.
Summary
Reading comprehension has usually been viewed as a multifaceted
complex skill. Silent reading with an emphasis on teaching decoding
skills and word comprehension has been the accepted way to teach
comprehension. Other ways of teaching comprehension have been
related to listening comprehension. The most widely used measure
of comprehension has been recall of'facts. Since the early turn
of the century, silent reading along with teaching decoding skills
has been the predominant method of teaching reading and comprehension.
In more recent years a body of research has evolved to support the
use of oral reading at high rates to increase comprehension.
Theories, Strategies, and Tactics of
Reading and Comprehension
Many different ideas of how to teach reading and to maximize
comprehension exist. Many reading programs have been developed to
coincide with stages of child development. Theories of child develop¬
ment are myriad.
Traditional Teaching Methods
Teaching reading usually begins as a sequential process. More
often, beginning readers are taught to read from a basal series

14
(Heilman, Blair, a Rupley, 1981). This early reading curriculum
usually involves letter naming, saying letter sounds, word-picture
recognition and word attack or phonetic analysis, respectively.
Mere often than not, little emphasis is given to speed or fluency
building. Also, mastery of reading skills is usually assessed over
a long period of time. Teaching in the regular classes is also
performed in large groups. Students are asked to read in a round
robin format with each child reading approximately 40-50 words in
30 to 45 minutes.
Since the early 1900's many reading methods have been developed.
One such method that existed since the 19301s is the Language
Experience Approach (LEA).
Many educators have advocated the LEA rather than a skills
approach (Hasselriis, 1982). LEA requires teachers to use printed
material in its "natural" form; then whole sentences are combined to
form stories, articles, magazines, newspapers, and books. It requires
that if instruction is focused on words, parts of words, or other
subsets of language, those fragments of language should not be
removed from their original context (Hasselriis, 1982). This method
recognizes the relationship between reading, speaking, writing, and
listening as part of an individual's personal experiences (Spache &
Spache,1977). It further requires that the students be given
extensive opportunities to read, write, listen, and speak in natural
settings rather than through contrived means. Proponents see that
a child cannot be expected to deal with ideas of language in reading
that are much further advanced than those he can speak or write.

15
Linguistics and reading comprehension. Reading has been
related to language development. Concerned with language and its
components, the linguistic approach has been drawn upon to guide
curriculum development, instruction, and methods of teaching reading
and comprehension. Linguistics has been defined as the study of
language. Proponents of this system attempt to explain the phonemic
and grammatical structures of language (Karlin, 1975). Two of the
more salient domains in linguistics are structural linguistics and
generative or transformational linguistics. Structural linguists
study oral language to identify its sounds; its units of meaning,
carried by words and by parts of words; and its syntax. Transforma¬
tional linguists seek to find the knowledge one must have in order
to make such utterances or to understand them. They typically try
to produce grammars that describe the often unobservable knowledge
of language (Karlin, 1975).
Transformational grammarians believe that knowledge of the
grammar of a language enables one to produce it, and that meaning
is derived not from the words in a sentence, or the surface structure,
but from knowledge of the deep levels of the language. Grammar
serves as a bridge between the deeper structure of a language, in
which meaning resides, and the surface structure of the language
(Chomsky, 1965). Comprehension then is a process which occurs between
the surface and the deep structure by means of the vocabulary,
phonology, and syntax of a particular language. Therefore, compre¬
hension of what is read is assessed through techniques that rely on

16
inferences and/or prediction of a next word within the context of
a passage.
Similar to the area of linguistics, psycholinguistics has
recently had an impact in the arena of reading comprehension.
Psycholinguists believe that reading and comprehension are an
active process. For example, people can read words without decoding
and can understand without knowing how to read words (Smith, 1S71).
Psycholinguists believe that comprehension of words and word identi¬
fication may be a result of comprehending the story before reading
it. Also, psycholinguists believe that readers read for meaning
rather than word identification (Smith, 1971). A technique that has
been developed to test comprehension within this model is the "Cloze"
procedure. This procedure uses an automatic word deletion process
(e.g., every nth word), whereby words are removed from a printed
passage. The pupil's task is to say the exact word that was removed
and replace it (Rankin, 1965). Proponents of this method indicate
that pupils can predict the next word on a Cloze procedure when their
prior knowledge, general understanding of the material, context
clues, and a knowledge of word usage are well developed.
Reading in context with a focus on punctuation has been described
as a way to increase comprehension (Heilman, 1981). This is consistent
with the linguist point of view. The linguists and psycholinguists
have also emphasized comprehension to be partly influenced by intona¬
tion. In addition to the consonant vowel phonemes, they relate that
English uses a number of intonational phonemes. Speech consists of

17
a flow of words arranged in particular patterns which result in
distinctive rhythms or melodies which are unique to English
(Heilman, 1967). Reading in context along with using the correct
intonational patterns is viewed as having an impact on comprehension.
Reading as operant behavior. In 1938, B. F. Skinner published
The Behavior of Organisms. In that book he describes the differences
between operant behavior and reflexes. He described the use of fre¬
quency as a basic datum of behavioral research to describe how behavior
changes under operant control. It wasnotuntil Azrin and Lindsley
(1956) used operant procedures with children that the power and
generalizability of Skinner's fundamental principles could be
realized. In 1957, Skinner's book Verbal Behavior conceptualized
how reading and comprehension can be defined as and come under operant
control.
The inaugural study that used principles of operant behavior to
explore the relationship between reading behavior and reinforcement
was performed by Arthur Staats and associates (Staats & Staats, 1962).
This led to the systematic development of strategies and tactics for
the investigation of reading behavior. Staats and his colleagues
(Raygnor, Mark, & Warren, 1966; Staats, Minke, Filey, Wolf, & Brooks,
1964; Staats & Staats, 1962) demonstrated that reading, like any
other operant behavior, was influenced and controlled by its conse¬
quences .
Often, operant technology has focused on arranging contingencies
of reinforcement for academic behaviors. For example, the use of

18
"token economies" (Ayllon & Azrin, 1969) has been used to increase
academic performances across most subject areas (Ayllon & Roberts,
1974; O'Leary, Becker, Evans, & Saudergas, 1969; Wolf, Giles, &
Hall, 1968). The efficacy of using reinforcement contingently has
also been demonstrated on a national level with Project Follow
Through (Benjamin, 1981).
Reading comprehension is also a class of responses which a
science of behavior can and has investigated (Staats & Staats, 1962).
However, the study of reading as a natural science phenomenon has
only recently gained momentum (Indrisano, 1982).
Recent Strategies and Tactics for Teaching Reading and Comprehension
Recent approaches to teaching reading have focused on strategies
that have proven empirical value. Some of these are still undergoing
parametric evaluations. One particular approach which has gained in
momentum over the past decade is Direct Instruction.
Direct Instruction. Direct Instruction has emerged as an effec¬
tive method to develop reading ar.d comprehension. The beginning
sequence of this approach begins with first having the child say the
letter sound, word-picture identification, blending sounds into words,
and letter naming, respectively (Englemann & Bruner, 1974). Much of
the direct instruction approach requires the child to continually read
aloud. Comprehension at the beginning levels emphasizes visual dis¬
crimination tasks, matching to sample, and visual equivalence training
(Sidman, 1971). Direct Instruction emphasizes teaching children
letter sounds before naming them. This was developed empirically as

19
a method to teach children that reading is similar to speaking.
An important hallmark of Direct Instruction is the interaction
between the instructor and the students. Immediate feedback is
typically provided. Also, Direct Instruction is under the philosophy
that all children can learn. Therefore, instruction does not pro¬
ceed unless the pupil demonstrates that a skill or a lesson has been
learned. There are important studies published that have described
the efficacy of Direct Instruction with a program such as DISTAR
(Benjamin, 1981). These studies typically have shown how DISTAR,
compared to many other popular teaching strategies, managed to
raise the performances of low achieving students so that they were
on and above grade level (Benjamin, 1981; Englemann, 1975).
Precision Teaching and reading comprehension. Recent investiga¬
tors have used the technology of Precision Teaching (PT) and reported
studies in which reading comprehension was the primary dependent
variable (Armstrong, 1981; Fleisher, Jenkins, & Pany, 1 979 ; Hansen
& Lovitt, 1976; Jenkins, Barksdale, & Clinton, 1978; Lovitt & Hansen,
1976; Roberts & Smith, 1980; Weaver, Holmes, Curtis, & Reynolds,
1970). The findings of and the paradigms that were used in each of
these studies can be seen in Table 1. Included in Table 1 is the
only study that could be found comparing oral reading and inflection
with comprehension (Poulton & Brown, 1967).
Precision Teaching (PT) has five basic ingredients: (a) replic¬
able teaching procedures, (b) individual analysis, (c) experimental
control, (d) direct measurement, and (e) daily measurement (Haring,
1978). Within this framework, 0. R. Lindsley introduced Precision

20
Table 1
Review of Selected Studies on Oral Reading Rate
and Comprehension
Author(s) Subjects Measures Intervention
Poulton a 24 Housewives Answers to Oral and silent
Brown, 1967 questions. reading.
Results: Memory for the first 30 percent of the passage was less
after reading aloud. Remainder of the passage memory was equally
good for both conditions. The last 10 percent of the passage was
remembered reliably after reading aloud. Difficulty level of compre¬
hension questions was not reported. Also, questions restrict the
total number of facts that could have been recalled.
Limitations: Readability was not controlled. Rereading passages
was not controlled. Also, rates of oral and silent reading was not
controlled. The use of punctuation and inflection was not controlled.
Weaver, Holmes, 18 volunteer Cloze proc. Oral and silent
Curtis, & undergraduates reading with
Reynolds, 1970 punctuation. Oral
reading without
punctuation.
Results: Reading silently with punctuation resulted in higher scores
on the Cloze questions.
Limitations: Reading rates were not controlled. The number of trials
read silently and orally could not be managed successfully. Cloze
procedure restricts that total number of facts that could have been
recalled.
Levitt 5 7 LD boys
Hansen, 1976
Words read
per minute
correct and
error.
Percentage of
correct
answers on
Contingency of
skipping. Drill
on words and
questions read/
answered incor¬
rectly.
comp, questions.

21
Table 1--Continued
Author(s) Subjects Measures Intervention
Results: Subjects oral reading rate increased during the skip and
drill phase. There was also a simultaneous increase in number of
correct answers to comprehension questions.
Limitations: Minimal changes from the skip and drill phase to the
final baseline phase. The effects of punctuation training was not
reported in the study.
Jenkins, 3 LD students Correct oral Money contingent on
Barksdale, reading rate. correct answers to
& Clinton, Answers to comprehension
1978 -comp, questions, questions.
Results: Reading comprehension improved under the money contingency.
Then number of words read correctly increased also under the money
contingencies.
Limitations: Subjects began at a high level of reading correct words.
Comprehension questions restricts that total number that could have
been recalled.
Fleisher,
Jenkins, &
Pany, 1979
36 students.
14 good readers.
22 poor readers.
Words read per
minute. Per¬
cent correct
on Cloze.
Number correct
on factual and
inferential
questions.
Decoding training.
Oral reading rate
of 90 wpm. In¬
struction in
reading for
understand!' ng.
Results: Poor readers improved their number of words read correctly
with training. Good readers did not improve significantly.
Limitations: Low proficiency standard for oral reading. Did not
control punctuation and inflection. Difficulty level of comprehension
questions not reported. Comprehension questions restricts the total
number of facts that could have been recalled.

22
Table l--Continued
Author(s) Subjects Measures Intervention
Roberts & 8 LD boys
Smith, 1980
Oral reading rate
corr. Percent
correct of comp,
questions.
Contingencies to
increase oral
reading. Model¬
ing and instruc¬
tion.
Results: Oral reading rate increased, but answers to comprehension
questions improved less.
Limitations: Punctuation and inflection training was not controlled.
Many variables involved with the study. Difficult to separate out
the controlling variables. Level of difficulty of the comprehension
questions was not reported.
Contingent
pennies for
correct answers
to comprehension
questions for
each level of
material pre¬
sented at
staggered inter¬
val s.
Results: There was an inconclusive relationship between the contin¬
gencies and the percent of comprehension questions answered correctly.
In all six cases there was a positive relationship between the median
number of words read correctly and comprehension. There was a nega¬
tive relationship between words read correctly per minute and mean
response latency. In all six cases there was a positive relationship
between errors per minute and response latencies.
Limitations: Did not control for punctuation training. The rate of
oral reading was not controlled. Also used comprehension questions
that may have limited the total number of facts that could have been
recalled. The level of difficulty of the comprehension questions was
not reported.
Armstrong,
6 LD students Type and number
1981
of responses to
comp, questions.
Response latency.

23
Teaching at the University of Kansas in 1964 (Haring, 1978). A
hallmark of PT is the i nsistance on daily charting of targeted
skills on a six cycle semilogrhithmic chart (Pennypacker, Koenig,
Lindsley, 1972). PT also incorporates the use of frequency as the
basic datum to describe behavior. Many classrooms throughout the
country use PT techniques, although great variability exists between
classrooms (Lovitt, 1977). Lovitt (1977) describes five characteris¬
tics of PT that are the best practices within these classrooms (p.
175).
1. The teacher or pupil must pinpoint each behavior of a
child's program. If one goal is to increase ability to read orally
from a certain text, a situation should be arranged to deal directly
with that behavior.
2. An aim (e.g., goal) must be determined for each identified
behavior. In order to determine an aim, the teacher or pupil must
decide the rate at which the selected behavior should occur and the
date on which that criterion should be achieved. Lines of progress
are then drawn on a chart of that activity from the intersection of
the current date and rate to the point of the projected date and rate
intersect.
3. Third, the teacher or pupil must count the number of times
the behavior occurs.
4. The teacher or pupil should chart each day the frequency
of the pinpointed behavior.

24
5. The teacher and pupil should evaluate the performance of
each charted behavior every day. If the correct rate is above and
the incorrect rate below the corresponding progress lines, the
current instructional technique should be continued. If, however,
the progress is not satisfactory, an instructional change should be
considered.
Direct Instruction and PT are two methods that can be used
in a coordinated way within the classroom. Both advocate direct
interventions. PT and Direct Instruction focus on observable data
to guide interventions that are necessary to have a learner reach a
mastery criterion. PT also uses frequency as its basic datum. PT
differs from direct instruction methods by its insistence on frequency
of response as being the datum of interest. Typically, Direct
Instruction uses a percent correct criterion. In other words, PT
relies on accuracy and rate to measure performance while Direct
Instruction will emphasize the number of trials to criterion or
accuracy.
The ecological approach. This approach is similar to the Language
Experience Approach and to ABA. The approach deals with the inter¬
actions between the organism as a whole and its entire environment,
which includes the existence and behavior of other organisms in that
environment (Bijou & Baer, 1961). In the educational sense, ecology
is the relationship of the learners to their instructional environment
(Indrisano, 1982). The learner is viewed as a dynamic part of an
ongoing changing environment.

25
Automa ti city theory. The tactics advocated by the proponents
of automaticity are in some ways similar to PT. This method
advocates the use of repeated readings (Samuels, 1979). “Fluent
readers automatically process information at the visual and
phonological levels and are therefore able to focus attention on
the meaning codes in the text" (Box, 1982, p. 51). Advocates of
this theory believe that when each level of processing receives
attention and assisted repeated reading to a fluent rate, selective
attention will be improved. Samuels (1979) advocates oral reading
at 85 wpm as the criterion for fluent reading. Although not proponents
of automaticity theory, Smith (1969) and Beck (1982) advocate fluency
to be at 150 to 200 wpm.
PT and automaticity theory both include fluency training and
rereading as an important tactic to increase comprehension. Consistent
with PT is the notion that fluent reading is a more functional way to
increase reading comprehension for three reasons. First, fluent
reading can increase retention. Second, fluent reading increases the
probability that the reader and/or audience will attend to what has
been read. Consequently, this may provide the reader with a greater
variety and frequency of reinforcement because of the faster pace.
Third, as this study seeks to substantiate, when oral reading approxi¬
mates conversational speech, comprehension will improve. These
reasons are similar and coincide with those given by the proponents
of automaticity and PT.

26
Microcomputers and reading. Another approach which has recently
been revived because of advances in electronics is the use of
programmed instruction through the use of microcomputers. The use
of machines and programmed instruction are contrived methods that
can be made a valuable tool for instruction (Skinner, 1958). More¬
over, through programmed instruction and teaching machines the student
can rehearse newly learned skills, receive immediate reinforcement,
and develop independent work habits. Although much more technologically
advanced than teaching machines, microcomputers can be used as a very
sophisticated learning device as Skinner had envisioned. Furthermore,
microcomputers can reinforce small increments of behavior that other¬
wise may go unnoticed by the instructor. However, the use of micro¬
computers have been limited because of software design. The importance
and/or the expertise to incorporate fluency building into educational
software has typically not been made part of the recently developed
software for students. However, some educators have recognized this
deficiency and have begun expressing their concerns and have suggested
alternative strategies (Eshelman, 1983; Wolking & Buss, 1983).
Summary
The nature of reading comprehension is difficult to delineate.
Several theorists have believed that reading comprehension is a
multifaceted skill. Others have actively sought to decrease the
number of skills necessary to constitute what is commonly called
reading comprehension. There is agreement as to the importance of
retaining facts as a minimum skill that is the foundation for later

27
types of comprehension. During the early 1900's silent reading
was considered to be the best method to. teach reading and comprehen¬
sion. Then along with silent reading, decoding and word comprehension
were thought to be the best ways to optimize comprehension. Some
researchers in the 1960's provided data which support the notion
that when the rate of oral reading reaches a proficiency standard,
there is a concomitant increase in comprehension. Lovitt and Hansen
(1976) provided data that indicate that although improvements were
not as great during oral reading training, reading silently at high
rates improved comprehension of materials that were read silently.
There are many theories as to how comprehension should be
conceptualized and subsequently taught. Some theorists have con¬
jectured that language and comprehension are closely related. The
linguists, which include the psycholinguists, and the transformational
grammarians advocate developing comprehension through increased under¬
standing of the deep structures of language. The Language Experience
Approach advocates the use of already learned language as a place to
begin instruction. Also, reading materials should consist of func¬
tional value to the students. Similarly, proponents of Applied
Behavior Analysis see reading as operant behavior that comes under
the control of contingencies within the immediate environment. The
fundamentals of operant research were developed by B. F. Skinner
(1938). Skinner's (1957) conceptualization of reading and compre¬
hension was presented in his book Verbal Behavior. He defined
comprehension as an intraverbal. As an intraverbal (see definition

28
of comprehension), the aim of teaching reading comprehension will
be to functionalize it. More recent approaches to reading compre¬
hension are Direct Instruction, Precision Teaching, and Automaticity
theory. The research supporting these recent approaches is providing
evidence as to their effectiveness across many environments. More
specifically, instruction which incorporates a standard of per¬
formance, direct and frequent measurement of performance, high rate
of oral reading along with practice of the same or similar stimuli
has the greatest probability of increasing reading comprehension.
Dependent Measures of Reading Comprehension
Without a standard definition of reading comprehension, strategies
and methods to measure its occurrence are difficult. Without a standard
definition, it is difficult to decide which treatment effects are to
be considered significant. In spite of this difficulty investigators
have developed ways to measure comprehension.
In general, responses to comprehension questions are the most
common format. Typically multiple choice type questions are used.
However, true/false questions, fill-in-the-blanks, and vocabulary
definitions have also been used (Hansen, 1979). Methods of assessment
that employ questions have many drawbacks. Comprehension questions
reveal only a portion of the available information that may be obtained
from textual material. Also, students are assessed on what another
person considers to be the major points made in the text.

29
More recently other methods have been developed to correct
the problems inherent in the use of questions to measure compre¬
hension. Short-answer fill-ins have been suggested in an attempt
to facilitate recall of response rather than recognition. These
open-ended questions do not resolve the other concerns about questions
as a measure of comprehension. Cloze techniques have been suggested
as another alternative to comprehension questions. Difficulty arises
with the Cloze procedure because it does not provide evidence of the
ability to understand entire passages. Moreover, this technique
does not permit evaluation of a person's critical and inferential
comprehension skills (Hansen, 1979).
A more direct measure of comprehension was proposed by Sidman
(1971). He believed that reading comprehension occurs initially when
a beginning reader can match a word to a corresponding picture. In
this way the child indicates that written words are symbolic repre¬
sentations of objects found in the environment. He also proposed
that pictures and words became equivalent to each other, because
they both can serve as discriminative stimuli for the same operant
(same class of responses), of the same auditory word. Several
studies have reported that matching spoken words to pictures and
to printed words is a sufficient prerequisite for the emergence of
reading comprehension at the word level (Sidman, 1971).
Another direct measure of comprehension is whether the child can
locate the actual object or perform the prescribed action. This
method relies on recall as a prerequisite skill. The ability to

30
recall facts does appear to be a fundamental prerequisite for all
other measures of comprehension. Without the ability to recall
facts concept formation and other abstractions will be limited
(Johnson & Chase, 1981). Thus, one might propose that training a
reader to recall facts should be the first comprehension level.
A free recall task that uses frequency of facts told correctly
to criterion (e.g., 15 to 30 correct facts recalled per minute), can
significantly aid retention, type of information remembered and
provide a direct measure of its occurrence (Hansen, 1979). The
ability to recall facts to criterion might then be considered as a
tool skill for all other types of comprehension. Once proficiency
has been demonstrated, another type of comprehension should be taught.
Johnson and Chase (1981) proposed a hierarchical instructional
typology for adult learners based on Skinner's Verbal Behavior (1957).
Their suggested sequence begins with teaching cf free recall and then
moving on to more abstract methods of teaching and measuring compre¬
hension. Within the instructional typology all types of comprehension
measures that use answers to literal questions, and fill-in-the-
blanks, are considered as fundamental steps in developing abstract
comprehension. Other measures of comprehension such as Sidman's
(1971) equivalence training would likewise be regarded as a funda¬
mental step in developing comprehension. This-.study will measure
comprehension on a fundamental level. In other words, all responses
will be considered as free recall across different stimulus events.
Since free recall is the basic building block of all later

31
comprehension, steps should be taken to determine the accuracy
and rate that will improve more abstract types of comprehension.
This is regarded as a step towards standardizing a definition of
comprehension.
Reading Comprehension and Memory
Although recall of facts is the most sought after measure of
comprehension, many reading programs do not employ teaching
strategies that enhance memory. Reading curricula usually rely on
one trial learning per story. Students are expected to read a
story one time and then answer comprehension questions. This does
not appear to be the most efficient teaching strategy to enhance
memorization. Ebbinghaus (1885) learned that the more exposure or
practice he had to learn a list of words, the greater was his recall
for these words.
Although practice is highly correlated with an increase of
retained material, Berquam (1981) investigated the relationship
between rate of performance and extra practice of learning nonsense
syllables. He used a paired associate task in a trigram form.
His study confirmed the results of earlier investigations, in that
level of previous learning has a very strong relation to subsequent
retention of skills. The group that worked to increase performance
required less time to relearn the nonsense trigrams than did the
group which received extra practice without fluency training.
Since Berquam's study used frequency of response data, the results
indicate that the basic learning/retention relation is found with

32
frequency data as well as with the accuracy measures of previous
studies. Berquam (1981) concludes from these findings, "It seems
likely that teachers could increase their students' skill levels
and subsequent retention by using short, concentrated periods of
fluency training, similar to those used by precision teachers"
(p. 73).
This study builds on Berquam's findings. In Berquam's study,
he used nonsense trigrams. For this study, reading passages will
be used. As Berquam used rate in terms of frequency, this study
will also use frequency of oral reading as a basic datum. This
sets the occasion to systematically extend Berquam's (1981) findings.
Summary
Many different kinds of measures of reading comprehension exist.
Most of the ones used are question and answer type. Problems with
questions usually center around the subjectivity of the answers.
Also, the format in which the questions are asked usually does not
contain a time component. The question answer format usually solicits
the fundamental skill of recall. Since recall is a fundamental skill
for later types of comprehension, then a format that directly assesses
facts per some unit of time appears to have great merit. Memory has
been an important issue for experimental psychologists and educators.
Studies have been performed in which practice and, more recently,
practice with fluency increase retention on later measures.

33
Oral Reading, Rate, Punctuation, and Comprehension
Since retention increases with practice and fluency, the rela¬
tionship between oral reading rate and comprehension should be
investigated. The relationship between oral reading and comprehen¬
sion has received little attention. An early investigation of the
efficacy of oral and silent reading was conducted by Poulton and
Brown (1962). In their study 24 housewives read some passages
aloud and others silently and then answered questions. Time
allowed for reading silently was matched for time reading aloud.
The researchers concluded that memory for the first 30 percent of
the passage was less after reading aloud. For most of the remainder
of the passage memory was equally good. The last 10 percent of the
passage was remembered better after reading aloud.
To improve oral reading Lovitt and Hansen (1976) used a high
rate of oral reading along with word drill made up of words taken
from the pupil's reading passage that were read incorrectly. They
also reported a simultaneous increase in comprehension. Their pro¬
cedure also generalized to silently read material. Also, when allowed
to read the passage silently after they were trained to read it
correctly orally, their silent reading also improved. The results
from this study may also be generalized to silent reading as well.
Oral reading appears to improve comprehension, but the rate at which
oral reading maximizes comprehension has recently begun to receive
attention.

34
There is little empirical data at this point to determine
which rates of reading are needed to maximize comprehension
(Buss, 1982). A wide range of proficiency rates for academic
skills are reported in the literature (Mercer, Mercer, S Evans,
1982). Haughton (1982) reported that oral reading reaches competent
levels in the 250 to 400 words correct per minute range, but in the
first three grades most children receive instruction in the 50 to
70 words per minute range. Smith (1971) reported that "normal"
readers call words at 200 worn. After reviewing the reading per¬
formance of over 3,000 second and third graders, Kunzelman (1973)
reported that the students designated as top readers by their
teachers read orally between 150 and 210 correct words per minute
with two or less errors at a particular grade level is considered
instructional level (Starlin üStarlin, 1973).
There appears to be a great deal of discrepancy of what consti¬
tutes proficient reading. However, Hoiking (1973) reported that,
by sixth grade, the median child has attained approximately 95
percent of adult accuracy and 65 percent of adult speed in the
performance of basic academic skills, Wolking's (1973) data appear
to translate into an easy formula by which a proficiency can be
judged on a criterion referenced basis. However, this does not
settle the question of which rate maximizes comprehension. Reading
at or near conversational adult speech has been regarded as a
possible goal for oral reading rate that may maximize comprehension
(Hoiking, 1982). This also appears to be in line with Skinner's

35
understanding of verbal behavior (1957). Moreover, in the Wolking
findings, comparing children with adults to determine rates appears
to approximate the development of a reading rate that resembles
conversational speech. Conversational speech has been clocked at
150 to 200 wpm (Wolking, 1982). Although this review does not
report any studies that used a proficiency standard of 150 to 200
wpm, some researchers have used high rates of oral reading to
determine its effect on comprehension.
In their study, Fleisher, Jenkins, and Pany (1979) trained 36
fourth and fifth grade students to read at 90 wpm to see the effects
it would have on comprehension. They had two groups of children.
The first group were good readers while the second group were poor
readers. The investigators concluded that there was no significant
change in any of the three measures of reading comprehension between
the groups.
In a more recent study Roberts and Smith (1980) demonstrated the
efficacy of improving reading fluency by using high rates of oral
reading with eight learning disabled boys. However, comprehension
improved less than oral reading correct did. In support of Roberts
and Smith's (1980) study, Armstrong (1981) reported a positive
correlation to exist between high rate of oral reading and compre¬
hension.
The impact that punctuation has on oral reading should not be
overlooked (Weaver, Holmes, Curtis, & Reynolds, 1970; Wolking, 19821.
One study that was significant to this research was found which

36
directly addressed the question of punctuation and its relationship
with comprehension. In this study 18 volunteer undergraduate
students were assigned randomly to one of six conditions: oral
reading with punctuation, oral reading unpunctuated; reading final
word in sentence, punctuated; reading words in sentences unpunctuated
(Weaver et al., 1970). A Cloze procedure was used to assess compre¬
hension. The investigators concluded that reading silently with
punctuation resulted in higher scores on the Cloze procedure. They
concluded that the imposing of the verbal production has an inhibit¬
ing effect upon comprehension (p. 82).
Summary
In summary, investigators studying the relationships between
reading and reading comprehension have reported results which suggest
that oral reading at high rates can improve comprehension. Except
for Lovitt and Hansen (1976) and Armstrong's (1981) studies, most
results have not been replicated. The rates at which oral reading
will optimize comprehension and retention have not been reported in
the studies reviewed. Studies which sought to determine the effects
of oral reading rate did not use proficiency standards that were
advocated by precision teachers. Experts in the area of Precision
Teaching report a rate of 200 wpm to be the proficiency aim for
reading aloud. Many of the studies cited did not control for the
possible effects that punctuation training may have had on oral
reading and comprehension. A study that did investigate the
relationship between reading and use of punctuation had design

37
difficulties. In that study, they could not prohibit rereading
of the silent reading phase. The effects of overt and covert
reinforcement have been well documented. Yet, many of the studies
cited in this review do not report the use of contingencies that were
operating during the study. This investigation will use a minimum
of intermittent praise as the only source of reinforcement. Also,
some studies which used group designs did not adequately control
for receptive vocabulary differences in their subjects.
Verbal Behavior
Skinner's (1957) conceptualization of verbal behavior can help
to improve the strategies and methods that are presently being used
to research and to teach reading comprehension. Skinner (1957)
made the case in his book Verbal Behavior that all instances of
verbal behavior such as reading, writing, answering questions, are
defined by their function. As noted earlier, comprehension is an
intraverbal. This concept uses the idea that comprehension like
other kinds of verbal behavior is part of a dynamic system that is
functional in some regard. For example, if a teacher has established
herself as either possibly rewarding and/or aversive, she can elicit
responses from a student that are functional to both the student and
the teacher. The response by the student would be functional
because he/she would be either seeking a positive reinforcer or
possibly avoiding a punishment. The response by the student then
is a stimulus for the teacher to behave in some fashion. Thus, the

38
student's response sets the occasion for some teacher behavior.
Viewed in this light, intraverbals such as free recall can be
taught for some functional value. For instance, when a child learns
to provide answers at a rate of 20 per minute, he/she will probably
move on to a new story and/or understand similar stories at a faster
rate. Since the occurrence of a high rate behavior typically
implies that some reinforcement contingency is operating (Ferster
& Skinner, 1957), it can be said then that high oral reading rate
has some functional value. Since recalling facts at 20 per/minute
may increase understanding a similar story at a faster rate, then
recalling facts at a high rate will take on a positive functional
value.
Experimental Analysis of Behavior and B. F. Skinner
Since this study uses B. F. Skinner's conceptualization of
reading comprehension as an instance of verbal behavior, a discussion
of his contributions and related research are included.
Behavioral practices resulting from the pioneering work of
Skinner have become the system called Applied Behavior Analysis (ABA).
Analysis on the individual level is probably the most salient charac¬
teristic of ABA. This strategy is closely related to the concept of
experimental control (Johnston & Pennypacker, 1980). Other research
methods often report group data that compare the means of experi¬
mental and control groups (Johnston & Pennypacker, 1980). By using
single subject designs the variation in individual responding is
readily available for visual inspection and possible subsequent

39
control (Johnston & Pennypacker, 1980; Sidman, 1960). Since
control of variation is feasible, statistical procedures are
often unnecessary. Systematic removal of sources of variability
in the data leads to stable responding (Sidman, 1960). When stable
responding has been achieved, systematic replication of the effects
of an intervention can be reported with greater confidence
(Johnston & Pennypacker, 1980; Sidman, 1960).
Johnston and Pennypacker (1980) have developed a definition of
behavior that includes the properties of a physical science as they
relate to behavior (see Appendix A). Verbal behavior (Skinner, 1957)
is a class of behavior within this generic definition of behavior.
Peterson (1978), based on Skinner's book Verbal Behavior (1957),
defines the verbal episode as having the following features.
1. It is established and maintained by reinforcement.
2. The reinforcement is mediated by another person.
3. The other person's action that results in the rein¬
forcement must have been specifically trained in order
to reinforce speakers.
Peterson then goes on to list features which are irrelevant to a
definition of verbal behavior (p. 9). These are as follows;
1. The topography of the behavior: which muscles are used
in making the response.
2. Dynamic characteristics of the response: speed,
intensity, repetition.
3. Verbal or non-verbal stimulus.

40
4. Stimulus mode: auditory, visual, gustatory, etc.
5. Reinforcement features: conditioned, unconditioned,
type of schedule.
Many people involved with reading research have made the point
that reading, like speaking and listening, should be considered as a
language process (Goodman, 1976; Mcleod & Crump, 1978; Peaster, 1976;
Ryan & Semmel, 1969; Staats & Staats, 1962). Skinner (1957) defined
reading in terms of verbal behavior which is explicitly different from
language. He referred to reading as textual behavior. It is con¬
trolled by a prior stimulus that is the response product (end
result) of writing behavior (such as a book) and there is point-to-
point correspondence between the stimulus and the response. 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. For example,
copying a sentence word for word from a blackboard will have a point-
to-point correspondence between what was on the blackboard and what
was copied.
As with speaking, reading comes under the control of an audience.
The audience is a type of controlling variable (Peterson, 1978). The
audience is usually a listener in the presence of whom verbal behavior
is typically reinforced. Moreover, it controls a group of response
forms. The teacher and the school setting can serve as the audience
effecting control over the response form likely to be emitted. For'
example, in a high school Spanish class, the teacher may only wish to

41
respond to questions and answers given in Spanish by the students.
In this way, Spanish is strengthened and English, in that particular
class, is weakened. Thus, the teacher defines the occasion for only
certain types of stimuli to be reinforced. This also holds for
reading and comprehension. For example, when a teacher responds
only to words read correctly, the teacher sets the occasion to be
an audience that will reinforce only correct word calling. In other
words, the audience is developing or has developed stimulus control
over a behavior or a class of behaviors. Closely related to this
concept of audience is the notion of stimulus control.
Stimulus Control
Verbal behavior develops as the audience mediates reinforcement.
Stimulus control is a function of discrimination training (Whaley &
Malott, 1971). The discrimination training procedure consists of
reinforcing a response in the presence of one stimulus and extinguish¬
ing it in the presence of other classes of stimuli. When the response
is more likely to occur in the presence of the discriminative stimulus
than in the presence of the nonreinforced stimulu, stimulus control
has been established. Transfer of stimulus control has been demon¬
strated through fading and errorless discrimination training (Moore &
Goldiamond, 1964; Sidman & Stoddard, 1967; Terrace, 1963; Touchette,
1968).
Skinner (1957) regarded reading as a type of stimulus-response
relation (stimulus control) in which the controlling stimuli are
visual words--written or printed text. In a mediated transfer study,

42
Sidman (1971) sought to discover if saying a word in the presence
of the visual equivalent of the word would suffice to establish
reading comprehension. Also, would teaching auditory-visual word
matching suffice for oral reading to emerge? He reasoned that
children normally understand words they hear before they learn to
read with comprehension (p. 6). He used receptive auditory visual
training for reading comprehension. By bringing receptive language
under discriminative control of words and pictures (he used pennies
and a bell contingent on correct response), Sidman was able to
demonstrate that retarded children could match sound to word and then
word to picture without instruction. Simply, he taught the retarded
child the word symbol association by saying the word in the presence
of the written word. The child was then able to read the word and
point to the correct picture that was associated with the word. This
phenomenon has been replicated with normal preschool children (Hoiking
& Greenwood, 1978).
Comprehension Questions and Free Recall as
Discriminative Stimuli for Intraverbals
The episode that is usually known as comprehension can be inter¬
preted as an instance of verbal behavior. More specifically, compre¬
hension in Skinner's metaphor of Verbal Behavior is called an intraverbal.
An intraverbal is a verbal response in which the prior controlling
variable is a verbal stimulus (e.g,, a discriminative stimulus
situation such as reading) with no point-to-point correspondence
(i.e., a thematic response instead of an exact duplication of what

43
has been seen or heard), between the stimulus (the question) and
response (the answer). In other words, there is no point-to-point
correspondence between the question and the answer. In this study,
comprehension is defined as an intraverbal. The maintenance of
comprehension is simply an attempt to evoke elementary intraverbals
(Johnson & Chase, 1981) over longer time periods.
Comprehension questions can be designed to evoke intraverbals as
the mediated transfer studies (Sidman, 1971; Wolking & Greenwood,
1978) have demonstrated. Any questions or procedures that require
the subject to perform recall behavior engage that subject in per¬
forming intraverbals (Johnson & Chase, 1981). However, if compre¬
hension questions are not designed carefully, basic intraverbals
(the foundation for extended intraverbals and concept formation)
can be disrupted (Johnson & Chase, 1981).
Two investigators 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 compre¬
hension answers missed. This should help them to discriminate the
stimuli that evoke the proper intraverbals from those that lead to
incorrect ones. The reader is encouraged to read Peterson's (1973)
introductory text to Verbal Behavior for a more detailed account of
other types of verbal behavior.

44
The concepts of Verbal Behavior (Skinner, 1957) may be readily
applied to the professional working with school aged children. As
Johnson and Chase (1981) have pointed out, an entire instructional
sequence can be mapped out for each individual student. This will
include a functional approach to teaching skills from the funda¬
mental to the abstract level. Moreover, by engineering functional
assessment strategies, interventions can become more precise and
capable of evolving into a practical solution for many of the
difficulties that may burden school children (Grimes, 1981).
Summary
Several definitions of reading comprehension exist. This, in
part, has contributed to the difficulties involved with the develop¬
ment of a standard measure of reading comprehension. Major theorists
have promoted their own definitions and measurement schemes congruent
with their theory. This has made it difficult to replicate and
extend reported findings. Based on Verbal Behavior (Skinner, 1957),
comprehension is defined as an intraverbal. This is a functional
definition which can be used as a standard definition. Along with
frequency, the basic datum of behavioral research, this can set the
occasion for researchers and practitioners to make direct comparisons
of other research that finds its way into the professional literature.
This is thought possible because frequency uses a universal dimensional
property of behavior known as countability (Johnston & Pennypacker,
1980). Also, frequency uses an absolute and standard unit of

45
measurement to express countability. Along with frequency as a
basic datum, conceptualizing comprehension as a functional behavior
can help to pinpoint exactly the conditions, topography, and the
outcome of the episode known as comprehension. This is considered
desirable because each of the dynamic parts of comprehension can be
analyzed and understood in relationship with the entire episode.
One common type of intraverbal which appears to be a prerequisite
for all other types of intraverbals is recall. Free recall or facts
told per minute appears to have much promise as a measure of an
elementary intraverbal. Free recall and other types of recall type
situations rely heavily on memory. Several studies indicate that
relearning skills to criterion accelerate as a function of the number
of learning trials needed to reach criterion during initial learning.
More recently, Berquam (1981) demonstrated that daily practice with
an emphasis on high rate performance decreases the number of trials
needed to relearn a skill compared with practice without fluency
building.
During the turn of the century, E. L. Thorndike (1917) spear¬
headed the movement towards silent reading. His contentions prevailed
until the early 1960's and 1970's. Since then, use of oral reading has
been demonstrated as being an efficient method of increasing compre¬
hension (Lovitt & Hansen, 1976). More recently, Samuels (1979)
demonstrated that repeated readings along with a fluency criterion
increase comprehension. The minimal rate range of oral reading at
which comprehension will be maximized has not yet been established.

46
Most comprehension studies have not controlled for the effects
that inflection may have had on comprehension and the maintenance
of comprehension. Similarly, studies that manipulated punctuation
did not control reading rate and/or the number of trials (resulting
in a practice effect). Inflection provides the reader and listener
with auditory cues that may aid in the comprehension of written
materials for normal hearing persons. The strategies and tactics
designed to view individual variation as it is needed in this study
have been derived from applied behavior analysis and its related
area, precision teaching.

CHAPTER III
METHOD
Experimental Questions
This study addressed six questions as follows:
Questions 1-3: How does high rate of oral reading without inflection
(Condition HRNI) influence comprehension and mainten¬
ance of comprehension compared with the following
conditions:
Condition LRNI: Low rate of oral reading without
i nfl ection?
Condition HRWI: High rate of oral reading with
inflection?
Condition LRWI: Low rate of oral reading with
inflection?
Questions 4-5: How does high rate of oral reading with inflection
(Condition HRWI) influence comprehension and mainten¬
ance of comprehension compared with the following
conditions:
Condition LRWI: Low rata of oral reading with
inflection?
47

48
Condition LRNI: Low rate of oral reading without
inflection?
Question 6: How does low rate of oral reading without inflec¬
tion influence comprehension and maintenance of
comprehension compared with low rate of oral reading
with inflection?
The format of the experimental questions lend themselves to
making a summary statement about which conditions had the greatest
influence on comprehension and maintenance of comprehension. This
can be found in the results chapter.
Experiment 1
Experiment 1 began in November, 1981, and the last day of data
collection was in February of 1982. The study was done to probe the
procedures and the design to see if they were adequate for observing
the suspected relationships between rate, inflection, comprehension,
and maintenance. Also, it seemed particularly important to test
ways to sequence the occurrences of the dependent measures. The
experimental questions listed above were used. Since the first
experiment yielded data that were systematic and orderly, it was
decided to continue this line of research. Experiment 2 contained
the necessary modifications that were developed as a result of having
performed experiment one. The findings of experiment one are included
in the results section.

49
Experiment 2
This study was designed to systematically replicate the first
experiment by using subjects of different age, sex, and skill level.
Two other changes were made. A Cloze procedure was added to the
dependent measures and a design change was also made. The design
was changed from an ABCD design to one which inreased the probability
of finding sequence effects.
Independent Variables
There were two independent variables. These were (1) reading
rate and (2) vocal inflection. Two rates of oral reading were used.
First, a proficiency reading rate of 150 to 200 words per minute
with two or less errors (Starlin & Star!in, 1973) was used. Secondly,
instructional rate or reading at 30 to 60 wpm with two or less
errors (Starlin i Starlin, 1973) was used. Vocal inflection was
defined as those audible qualities of oral reading that come under
the control of punctuation marks and make oral reading approximate
conversational speech.
Apparatus
During Experiment 1 a General Electric tape recorder (model
number 3-5311) was used to record oral reading with and without
inflection. A Casio Melody Alarm wrist watch was used for timing.
It has a stop watch function that records time to a tenth of a second.

50
The watch provides an auditory signal at 60 second inter-als. This
made it possible to make reliable measurements of frequency,
latency, and duration.
Experiment 2 used the same apparatus to serve the same
functions as in Experiment 1. The tape recording permitted the
experimenter to check and verify the onset of the independent
variables. The cassette recording helped the experimenter to check
the accuracy and reliability of recording oral reading with or with¬
out inflection. The recordings were also used to record the verbal
responses to the different comprehension measures.
Dependent Variables
There were three dependent variables. They were (1) free recall
of facts, (2) answers to 10 comprehension questions, and (3) written
answers to a Cloze procedure (every fifth word was deleted). Compre¬
hension and maintenance of comprehension were measured in exactly
the same way. Comprehension was defined as the score received on
each of the dependent measures immediately following criteria for
oral reading for each condition. Maintenance of comprehension was
defined by the three scores on each of the dependent variables after
intervals of three and 10 calendar days. Each subject received
three different measures of comprehension in each of the four condi¬
tions, at three different times. This yielded a total of 36 measures
of the dependent variable that each of the four subjects received.

51
The dependent measures were taken in the following order:
free recall, ten comprehension questions, and then the Cloze
procedure. With this format the comprehension questions and the
Cloze procedure would not offer any cues or give answers to the
subject during the free recall episode. Since the Cloze procedure
used the entire reading story with every fifth word deleted, it
would have been possible for the subjects to recall facts that
otherwise would not have been recalled. To prevent this, the Cloze
procedure was always given last.
The ten comprehension questions consisted of the "what, when,
where, why, and how" variety. Whenever the textbooks listed compre¬
hension questions for a story, they were included as part of the ten
questions. Questions other than the ones taken from the textbook
were developed by the experimenter. The Cloze procedure was written
with a word processor on the TRS 80 Model 1 microcomputer. The Okidata
Microline 82A dot matrix printer was used to print the text.
Each of the dependent variables was measured by using universal
dimensional quantities of behavior (Johnston & Pennypacker, 1980)
These are countability, latency, and duration. When behavior occurred
often within a particular response class, frequency was employed
When changes in the frequency, latency, and duration of behavior
occurred over time relative to the independent variables, a ratio
multiplier was used to describe the changes. For other measures, a
ratio index of change was used.

52
Subjects
For Experiment 1, two subjects from Gainesville Job Corps
were selected: a black student and a white student, ages 18 and
16 years, respectively. The 18 year old was on a sixth grade
reading level. The 16 year old was found to be on a second grade
reading level. These reading grade levels were determined by a
modified informal reading inventory taken prior to the experiment.
All the research was conducted within the subjects' regular reading
class at eight o'clock in. the morning located at the Gainesville Job
Corps. The Open Court Basic Reader,'was used with the 16 year old
(Open Court Basic Readers, 1972) and the SRA Reading Kit entitled
"Black Like Me" (Science Research Associates, 1976) was used with
the 18 year old.
For Experiment 2, four white females in the third grade were
selected from three different classrooms located within a school in
Alachua County, Florida. Subjects were selected randomly from a list
of children who, in addition to the above, met the following two
criteria: they must have been between the ages of eight and nine
years of age. They also had to have been reading on the third grade
level, which was determined by their completed work in the Ginn
Reading Series (Ginn, 1976). Prior to Experiment 2, the Inventory
of Basic Skills (Brigance, 1980) vas used to assess each subject's
reading level. The purposes for using informal assessments have been
documented elsewhere (Alpar, Nowlin, Lemoine, Perine, & Bettencourt,
1974; Brigance, 1980; Lovitt & Hansen, 1976). The results indicated

53
that all subjects were on grade level. Reading materials for these
students were selected from a reading series which was grade appro¬
priate. The students read entire stories from the Open Court
Reading Series (Open Court, 1972). Table 2 lists all the subjects
and describes the materials they each read in each condition.
Fry's (1972) readability formula was applied to increase the
probability that the stories from the reading series were grade
appropriate for the subjects. All materials used for these experi¬
ments were on or near each subject's instructional reading level.
An attempt was made to insure that different reading passages would
appear at different conditions across subjects. For example, the
story in condition HRNI for subject three would also appear in
different conditions for subjects four through six. This was thought
to decrease the probability of experimental effects being attributable
to interest levels that subjects may have had for one story over
another.
The above differences stated between Experiment One and
Experiment Two subjects are in part the elements that comprised
a systematic replication of experiment one.
Design
A single subject design (Sidman, 1960) was used so that an
analysis at the individual level could be performed. This strategy
for conducting research has been well developed and documented
(Baer, Wolf, & Risley, 1968; Baily, 1977; Johnston & Pennypacker,

54
Table 2
Table of Subjects' Condition, Sequence, Reading
Passages, Humber of Trails to Criteria, and Total Time in
Experiment 1 Subjects
Subject 1: 18 year old black male; Gr. Level 6
Readings from: SRA School House Kit Black Like Me.
Phase Sequence ABCD
Story Trials
Phase A: The Skeleton's Song. 8
Phase B: The Burning Hand. - 8
Phase C: In Jai 1. 11
Phase D: The Beauty of Me. 11
Subject 2: 16 year old white male; Gr. Level 2
Readings from: Open Court Basic Readers Level 2.
Phase Sequence ABCD
Phase
A:
The Camel's Hose.
12
Phase
B:
The Sphinx.
12
Phase
C:
Hot and Cold From One Mouth.
18
Phase
D:
The Monkeys.
18
Experiment 2 Subjects
Subject 3: 8 year old white female; Gr. Level 3
Readings from: Open Court Basic Readers A Trip Around
Phase Sequence ABCD
Phase A: The Blind Men and The Elephant. 4
Phase B: The General and the Arrows. 4
Phase C: Antarctica. 7
Phase D: Cortez and Montezuma. 7
Mi ñutes
Time
78 min.
101' 29"
85' 48"
123' 33"
55' 3"
65' 48"
57' 16"
63’ 15"
the World.
12' 8"
21' 9"
26' 25"
66' 24"

55
Subject 4: 8 year old white female; Gr. Level 3
Readings from: Open Court Basic Readers A Trip Around the World.
Phase Sequence CDAB
Story Trials
Time
Phase C
Cortez and Montezuma.
8
48' 54"
Phase D
Antarctica.
8
64' 42"
Phase A
The General and the Arrows.
4
16r 54"
Phase B
The Blind Men and the Elephant.
4
33' 48"
Subject 5:
8 year old white female; Gr. Level 3
Readinqs from: Open Court Basic Readers A Tri_p
Around
the World.
Phase Sequence ABCD
Phase A
How the World Looks.
5
15' 18"
Phase B
Copernicus and Galileo.
5
34' 26"
Phase C
The Blind Men and the Elephant.
4
101 25"
Phase D
The General and the Arrows.
4
21' 2"
Subject 6:
8 year old white female; Gr. Level 3
Readinqs from: Open Court Basic Readers A Trip
Around
the World.
Phase Sequence CDAB
Phase C
The Meltinq Pot.
7
47' 42"
Phase D
Cortez and Montezuma.
7
61' 5"
Phase A
How the World Looks.
5
22' 1"
Phase B
Antarcti ca.
5
29' 36"

56
1981; Sidman, 1960). The strategy allows for an interactive approach
between the experimenter and the independent variable. This is made
possible because M=1 designs allow for the identification of experi¬
mental sources of intraindividual and interindividual variability
while the data are being generated by the subject (Pennypacker, 1982)
Also, this strategy permits the use of ratio comparisons that can be
made within experimental conditions and across experimental condi¬
tions .
Since this study required that the independent and dependent
variables occur in a predetermined sequence, a design was used so
that if sequence and practice effects occurred, they could be
observed (Johnston & Pennypacker, 1980). This study involved four
conditions with two different sequences. Condition HRNI required
that the subjects read at a high rate without inflection. Condition
LRNI required oral reading at low rate and without inflection. Condi
tion HRWI consisted of reading orally at a high rate and with correct
inflection. Reading at low rate with inflection comprised Condition
LRWI. The sequence of conditions for Subjects one, two, three, and
five was HRNI, LRNI, HRWI, LRWI. For Subjects four and six the
conditions were ordered HRWI, LRWI, HRNI, LRNI. This design also
made it possible to replicate the results across subjects and across
sequences. (See Figure 1.)
For the LRNI and LRWI conditions, the number of opportunities
to read a story was determined by the number of times stories were
read in HRNI and HRWI conditions, respectively. For example, the

u\
WI
Subject 1
Oral Reading
Classroom
Subject 2
Oral Rending
Classroom
Subject 3
HR
WI
Lit
NI |
MR
HI
I.J1
NI
HR
NI
LR
WT
hr
wi
LR
WI
IIR
WI
I R
NI
HR
nT
IR
ÑI
IIR
HÍ
LR
NI
MR
iii
I R
wi
UR
WI
IIR
ÑI
1 R
WI
rigure
Experimental
I
Design
Subject 4
High Rate Oral Reading
Without inflectínn
low Jt a te Oral Readi ng
Wi thout Inflection
J ligh Ra te Oral Rea ding
With Inflection
Low Rate Oral Reading
Hi Ill Inflection

58
number of times a subject read a story to meet reading rate criteria
in condition HRNI was the same number of times that the subject
could read a story in LRNI. In this way, practice effects were
equal so that the experimental effects of rate on the dependent
measures could be detected.
Procedure
The experimenter met with the two subjects in Experiment 1
every day for half hour each beginning at 8 a.m. The sessions took
place at a desk in the regular clas'sroom at Job Corps. The experi¬
menter met with the four subjects in Experiment two every instructional
day. Each subject was seen for one half hour between 8 and 10 a.m.
each day during the regular school week. The experimental sessions
took place behind the Media Center of the library at the school.
Pretraininq Period
High rate for oral reading was defined as 150 to 200 wpm with
two or less errors. Reading a story at 30 to 60 words per minute
with two or less errors was considered the low rate for each subject.
To teach the idea that reading can be as fluent as conversational
speech, fluency training with an easier prerequisite skill was con¬
ducted for all subjects. This is also known as tool skill training
(Haughton & Binder, 1982). Subjects were asked to say the letters of
the alphabet for one minute. Two to four of these timings were con¬
ducted. Also, subjects were asked to think/say "Star Wars is one of

59
my favorite movies" (Wolking, 1982). This was also done to shorten
the number of trials needed to reach proficiency criteria (Haughton
& Binder, 1982). Tool skill development was done prior to reading
a story and at times just before rereading a story. For subjects
who displayed little growth in oral reading, their reading rate
criteria for conditions A and C were then reduced to 100 correct
words per minute with two or less errors. Little growth was defined
to mean those subjects who were accelerating less than 10 percent per
session and their rates were lower than 100 words per minute on each
reread for two or three days. This occurred for three of the six
subjects.
During oral reading, when a subject misread a word, the experi¬
menter said the word aloud to the subject so that fluency would not
be disrupted. Furthermore, misread words were corrected and recorded
for later practice by the subject. These words were practiced every
day until they were read to a criterion of 40 to 60 per minute with
two or less errors.
Experimental Conditions
The high rate no inflection condition (HRNI) required the subjects
to reread a story orally and without inflection until 150 to 200 wpm
with two or less errors was reached. At times when the error rate
went above two per minute the experimenter modeled reading without
inflection. The modeling procedure was as follows:
(a) The experimenter asked the subject to read quickly and not
to pay attention to the punctuation marks.

60
(b) The experimenter then said that reading should sound like
the Federal Express commercial that they had seen on T.V. so that
their reading was fast and monotone and with equal emphasis on all
words.
(c) The experimenter then began reading a passage in a monotone
voice at a rate equal to 150 to 200 words per minute.
When criterion was achieved, measures of comprehension were taken
before starting the next condition. The number of trials to criteria
was recorded. This number was then used in a yoked manner in which
rereading in the LRNI condition only occurred for the same number of
trials that was required to reach criteria in the HRNI condition.
A new story for each subsequent condition was used. A measure of
the maintenance of comprehension of the story read in the HRNI
condition was then taken three and 10 calendar days from after the
end of HRNI condition. During the interval between the three and 10
days of the maintenance of comprehension measures for one condition
the next successive condition was in progress.
The LRNI condition required subjects to reread a story orally
without inflection at a rate of 30 to 60 wpm with two or less errors.
Modeling and verbal instructions were given similar to the instruc¬
tions for the HRNI condition, except the subjects were asked to
read slowly. The experimenter modeled this by reading in a monotone
voice with equal emphasis on each word, at a rate of one word per
second, so that the 60 wpm criteria with two or less errors would
not be exceeded. When the requirements in the LRNI condition had
been met, measures of comprehension were then taken. Maintenance

61
of comprehension was assessed for the LRUI condition three and 10
calendar days from the end of the condition.
Prior to the HRWI condition punctuation and inflection training
was provided. The punctuation marks that were taught and drilled
were the period, comma, quotation marks, exclamation mark, and the
question mark. First, naming punctuation marks for one minute was
a daily task prior to reading. Then the experimenter told the
subjects the function of each punctuation mark. Soon after, when
the experimenter pointed to each of the punctuation marks, the
subject told the examiner the function of each mark. This continued
until the subject was able to say the function of each punctuation
mark with 100 percent accuracy. Once this was completed, the subject
then pointed to each mark and then told the examiner the function
for each one until 100 percent accuracy was obtained. A probe sheet
with eight isolated phrases was used to practice oral reading at
high rates with correct inflection. The experimenter read these
phrases quickly and with the correct inflection. The subjects then
were asked to read these phrases as if they were speaking with some¬
one. At first, daily one minute timings were used. Later, the
subjects reread these phrases as many times as they could in two
one-minute timings. Reading a phrase with the correct inflection was
counted as one correct movement. If the subject inflected the
phrase correctly, but made an error in word calling the phrase was
still counted as correct.

62
In the HRWI and LRWI conditions an error was recorded when
incorrect inflection occurred during oral reading. The HRWI
condition involved rereading a story until criteria were reached
with correct inflection. When criteria were met, measures of com¬
prehension were then taken. The maintenance of comprehension measure
for the HRWI condition story occurred three and 10 calendar days
after criteria had been met.
The LRWI condition required subjects to reread a story for the
same number of trials it took to reach criteria in the HRWI condi¬
tion. When criteria were met, measures of comprehension followed.
Maintenance of comprehension was measured three and 10 calendar
days after the subject reached criterion.
Data Analysis--Dependent Variables
All performance data except for the answers to the 10 comprehen¬
sion questions and percent correct on the Cloze procedure were
charted on a Standard Celeration Chart—a six cycle semiTogarithmic
chart (Pennypacker, Koenig, & Lindsley, 1972). (See Figure 2.)
Oral reading was plotted as words read correct and incorrect per
minute. Free recall required two levels of analyses. First, fre¬
quency of correct and incorrect facts per minute recalled were plotted.
Second, words read per minute were divided by the total number of
facts recalled. This was done so that a measure of words read per
fact could be charted.

COUNT PER MINUTE
CALtNDAR WEEKS
SUCCESSIVE CALENDAR DAYS
SUPERVISOR ADVISER MANAGER
AGI l ADEL COUNTtD
DLPOSiTOR AGENCY
COUNTER CHARTER
CTi
CO
Figure 2
Standard Celeration Chart

64
For the ten comprehension questions, the latency to respond
to each of the ten questions and the count correct and incorrect
were plotted.
The Cloze procedure required two levels of analyses. First,
percent correct was recorded as the datum. Second, words written
correctly and incorrectly per minute were also plotted. To obtain
a rate correct per minute the total amount of seconds to complete
the Cloze test was divided into the number correct. This number
was then multiplied by 60 to get rate correct per minute. This was
also done for obtaining incorrects per minute.
The dependent measures of comprehension, maintenance after
three, and maintenance after 10 calendar days were compared. A
frequency multiplier was used to compare one condition with another
condition for measures of total performance (total performance =
rate correct + rate incorrect). An accuracy multiplier ratio was
used to describe accuracy ratio differences between conditions. A
total number correct for the 10 comprehension questions was used
to describe differences between conditions. A ratio index of change
was used to describe differences on the latency measures. The
measures for the Cloze procedure were the same for free recall.
In addition, a percent correct was used to describe differences
between conditions. (See Figure 3.)
A gain score for each of the independent variable conditions
was tabulated. This was done by counting the number of times the
dependent measures of a particular condition showed improvement over
the other conditions within the corresponding dependent measure.

65
A
f-
0*
3 4 5 6 7
Instructional Days
10
£ = Measures of comprehension immediately following criterion
performance for each phase.-
+ = Measures of maintenance occurring 3 and 10 instructional
days following the end of each phase.
Figure 3
Sequence of Dependent Measures
(Comprehension and Maintenance)

6fi
For example, if the accuracy ratio for facts recalled per minute in
the HRWI condition for the three days after criteria measure was
greater than the number of facts recalled in the LRWI, HRNI, and
LRNT conditions for the same corresponding dependent measure, then
the HRWI condition would have a gain score of three for that dependent
measure. Then all of the gain scores are added together and pre¬
sented as total number of gains for a particular condition on a
particular dependent measure.

CHAPTER IV
RESULTS
This chapter is designed to present data as they relate to each
of the experimental questions. The data are presented in the order
they were gathered while doing the research. Therefore, the data
generated from Experiment 1 are presented first. The results of
Experiment 1 consist of the free recall measures (total performance,
accuracy, and words per fact), count correct on the comprehension
measures and the median latencies, respectively. The Cloze procedure
was used only in Experiment 2. The results of the Cloze procedure
include total performance, accuracy and percent correct, respectively.
Since the Cloze procedure occurred after the free recall measures and
answers to the 10 comprehension questions, it is the last dependent
variable presented in this chapter. For Experiment 1 there was a
total of five measures that occurred three different times for two
subjects during each of the four conditions. Therefore, a total of
120 measures for Experiment 1 was taken. Experiment 2 had a total
of eight measures that occurred three different times for each of the
four subjects during each of the four conditions. Therefore, there
is a total of 354 measures for Experiment 2.
67

68
Six questions in a comparison format were asked in this study.
Since there were four conditions, a 3 by 2 by 1 matrix is presented
as an aid to the reader when following the results of the comparisons.
(See Table 3.)
Questions 1-6 Results for Experiments 1 and 2
Free Recall Total Performance
Question One: (A comparison between two reading rates) How
does high rate of oral reading without inflection (HRWI) influence
comprehension and maintenance of comprehension compared with low rate
of oral reading without inflection (LRNI)?
Experiment 1 Results for Question One
Figure 3 shows the total performance (speed) for facts recalled
per minute for all six subjects.
Criteria performance-free recall. Subjects in Experiment 1
were D.D. and J.J. For these subjects the HRNI condition produces
a faster rate for each measure of the dependent variable than the
LRNI condition for facts recalled per minute. For D.D., HRNI
condition is faster for each of the maintenance measures. For
J.J. the HRNI condition produces a faster performance compared with
the LRNI condition. The table provided in Appendix A displays how
many times faster or slower a condition is when compared to another
condition. This number is called a frequency multiplier index. A
number with an "X" indicates that the referent condition is faster.
A number with a division sign means that the referent condition is

69
Table 3
Graphic Display of Experimental Questions
Conditions
HRNI
LRNI
HRWI
LRWI
HRNI
Quest
1
Quest
2
Quest
3
HRWI
Quest
4
LRNI
Quest
5
Quest
6
HRNI = High rate without inflection
LRNI = Low rate without inflection
HRWI = High rate with inflection
LRWI = Low rate with inflection

70
slower. From the table provided in Appendix B, it can be seen that
initially for D.O. the HRNI condition is 1.25 times faster than the
LRNI condition. For J.J. the HRNI condition initially is 2.14
times faster than the LRNI condition.
Three day performance--free recall. Maintenance at three
days shows the HRNI condition to be 1.21 times faster than the LRNI
condition for subject D.D. For subject J.J. the HRNI condition is
2 1/2 times faster than the LRNI condition.
Ten day performance—free recall. Maintenance at ten days for
D.D. shows the HRNI condition to be 2.0 times faster than the LRNI
condition. For J.J. maintenance at 10 days the HRNI condition is
2 1/2 times faster than the LRNI condition.
Question One Results for Experiment 2
The subjects in this experiment were S.B., E.H., A.P., and
B.S.
Criteria performance--free recall. For these subjects the
HRNI condition initially produces a faster rate of recalling facts
than the LRNI condition. The frequency multipliers (see Appendix C)
for the HRNI condition are times XI.44, XI.27, X2.67, and XI.71
faster than the LRNI condition.
Three day performance—free recall. Three subjects under the
HRNI condition at the three day maintenance measures produce faster
rates compared with the LRNI condition. The HRNI condition influences
three subjects to produce slower rates, while the HRNI condition
influences E.H. to produce faster rates. The LRNI condition produces
a slower rate of responding for three subjects.

71
Ten day performance-free recall. It can be seen that for two
subjects, the HRNI condition at the ten day maintenance measure
influenced a faster response rate of recalling facts than the LRNI
condi ti on.
Condition Two Results for Experiment 1
Question Two: (A comparison between inflection conditions)
How does high rate of oral reading without inflection (HRNI) influence
comprehension and maintenance of comprehension when compared with high
rate of oral reading with inflection (HRVJI)?
Criteria performance--free recall. For both subjects the HRWI
condition initially produces a rate of recalling facts that is quicker
than facts recalled under the HRNI condition. From Appendix 3 it
can be seen that Subject D.D. under the HRWI condition is initially
1.4 times faster than his performance under the HRNI condition. And
the HRWI condition influences J.J. to be a 1.13 times faster than his
performance under the HRNI condition. (See Figure 4.)
Three day performance-free recall. For both subjects at
maintenance after three days the HRWI condition produces a faster
rate of recalling facts than the rates reported under the HRNI
condition.
Ten day performance-free recall. The HRWI condition produces
a rate for D.D. during the maintenance measure at 10 days that is
faster than those produced by the HRNI condition. However, J.J.
produces a faster rate under the HRNI condition than he does under
the HRWI condition.

COUNT pER MINUTE
CALENDAR WEEKS
SUCCESSIVE CALENDAR DAYS
SUi'l ftVISOK AOviSEH MAUAulH
BHIAVLR
ACl IAI!(1 COLIN TK)
OU'USITOH AGENCY
TIMER COUNILH
CHAKIIH
Figure 4
Free Recall Performance
IX>

73
Question Two Results for Experiment 2
For S.B. and E.H. initially, the HRWI condition produces rates
of recalling facts that are 1.07 and 1.21 times faster than facts
recalled under the HRNI condition, respectively. From Figure 3 it
can be seen that initially, there are no differences between the
HRWI and HRNI for Subject A.P. Under the HRNI condition Subject
B.S. initially produces a rate that is 1.71 times faster than her
rate generated under the HRWI condition.
Three day performance—free recall. Maintenance at three days
shows two subjects, A.P. and B.S., under the HRWI condition producing
a degree of performance that is faster than the HRNI condition. For
Subject S.B. the HRWI and HRNI conditions produce a similar rate of
performance. Subject E.H., when the HRNI condition is in effect,
yields a rate of recalling facts that is faster than facts recalled
under the HRWI condition.
Ten day performance—free recall. The speed of performance for
maintenance at 10 days shows that for two of the subjects, S.B. and
E.H., produced faster rates under the HRWI condition than they did
under the HRNI condition. For B.S. rate of performance under the
HRWI and HRNI conditions are the same. The HRNI condition for
Subject A.P. produces a rate at ten days that was faster than her
rate under the HRWI condition.
Question Three Results for Experiment 1
Question three: (A rate versus inflection comparison) How
does high rate of oral reading without inflection (HRNI) influence

74
comprehension and maintenance of comprehension when compared with
low rate of oral reading with inflection (LRWI)?
From Figure 4 it can be seen that under HRNI both subjects,
initially and after three days, are able to recall facts faster
than those under the LRWI condition for the same measures. At the
ten day measure, the HRNI condition for Subject J.J. produces a
faster rate. The HRWI and HRNI conditions yield similar rates for
Subject D.D.
Question Three Results for Experiment 2
Initially, the four subjects produced faster rates under the
HRNI condition compared with the LRWI condition for facts recalled.
From Appendix B it can be seen that performances under the HRNI
condition are 1.63, 1.56, 1.33, and 1.71 times faster, respectively,
than their performances under the LRWI condition.
Three day perforaiance--free recall. For maintenance after three
days three subjects under the HRNI condition produced rates that are
faster than their rates under the LRWI condition.
Ten day performance—free recall. At the ten day maintenance
measures, the HRNI condition produces rates for two subjects that
are about 1.5 to 2.0 times faster than their rates under the LRWI
condition. For E.H. and B.S. the LRWI condition produces a faster
rate of recalling facts.

75
Question Four Results for Experiment 1
Question four: (A high and low rate comparison, with inflection)
How does high rate of oral reading with inflection (HRWI) influence
comprehension and maintenance of comprehension compared with low
rate of oral reading with inflection (LRWI).
Criteria performance--free recall. By comparing the LRWI condi¬
tion with the HRWI condition Subjects D.D. and J.J. initially produce
faster recalls per minute under the HRWI condition. From the table
in Appendix B it can be seen that D.D.'s performance under the HRWI
condition initially is 1.62 times faster than his comparable measure
under the LRWI condition. For J.J., the HRWI condition produces a
rate that is b.7 times faster than his LRWI condition rate.
Three day performance-free recall. The HRWI condition for both
subjects produces a faster rate of recalling facts compared with the
LRWI condition at the three day measures of maintenance.
Ten day performance-free recall. For Subject D.D. the HRWI
produces a faster rate when compared with the LRWI condition at
the 10 day measures of maintenance. For Subject J.J. the HRWI
condition produces a frequency multiplier that is 1.25 times faster
than the LRWI condition.
Question Four Results for Experiment 2
For three subjects initially the HRWI condition produces a faster
rate of recalling facts per minute compared with the LRWI condition.
Subject B.S. under the HRWI condition produces the same rate as the
LRWI condition.

76
Three day performance--free recall. For three subjects,
maintenance after three days under the HRWI condition produces a
faster rate of recalling facts than under the LRWI condition.
Ten day performance--free recall. Three subjects produce a
faster rate of recalling facts under the HRWI condition compared
with the LRWI condition at the ten day measures of maintenance.
Question Five Results for Experiment 1
Question five: (A comparison between the presumed best and
worst conditions) How does high rate of oral reading with inflection
(HRWI) influence comprehension and maintenance of comprehension
compared with low rate of oral reading without inflection (LRNI)?
Subjects D.D. and J.J., under the HRWI condition, yield fre¬
quency multipliers that range from a times 1.25 to a times 2.42
faster than rates that are produced under the LRNI condition, across
all measures.
Question Five Results for Experiment 2
For three subjects, the HRWI condition produces rates that are
faster across all measures in comparison with the LRNI condition.
Initially Subjects S.B., E.H., and A.P., under the HRWI condition
produce frequency multipliers that are 1.56, 1.55, and 2.5 times
faster than the LRNI condition.
Question Six Results for Experiment 1
Question six: (A comparison between inflection and no inflection
at low rates) How does low rate of oral reading without inflection

77
(LRNI) influence comprehension and maintenance of comprehension
compared with low rate of oral reading with inflection (LRWI)?
Criteria performance—free recall. Initially, for Subjects
D.D. and J.J., the low rate with inflection condition increases
their rate for recalling facts per minute.
Three day performance--free recall. This trend maintains for
J.J. at the three day maintenance measure. Subject D.D. produces
identical rates under conditions B and D at the three day maintenance
measure.
Ten day performance--free recall. For Subject D.D. the LRWI
condition produces faster rates when compared with the LRNI condition.
For Subject J.J. there is no difference between the speed of recalling
facts produced by either condition.
Question Six Results for Experiment 2
Initially, S.B. and E.H. under the LRNI condition produce faster
rates when compared with the LRWI condition. Subject A.P. produces
a faster rate of responding under the LRWI condition. Subject B.S.
produces the same speed of performance regardless of condition.
Three day performance--free recall. For three subjects, the
LRWI condition produces faster rates of performance.
Ten day performance—free recall. The trends are very similar
to those seen at the three day measure. That is, three subjects at
the ten day measure of maintenance produce faster rates under the
LRWI condition compared with the LRNI condition.

78
Accuracy Results
The following results are presented in the same format as the
performance data.
Question One Results for Experiment 1
Question one: (A comparison between two reading rates) How
does high rate of oral reading without inflection (HRNI) influence
comprehension and maintenance of comprehension compared with low rate
of oral reading without inflection (LRNI)? Figure 5 depicts the
accuracy ratios for all subjects.
Criteria recaí 1--accuracy. From the table in Appendix C it
can be seen that, for Subjects D.D. and J.J. under the high rate
without inflection condition (HRNI), they initially produce accuracy
ratios that are 1.3 and 2.5 times more accurate compared with the
low rate without inflection condition (LRNI).
Three day recall—accuracy. For Subject J.J. the HRNI condi¬
tion continues to produce an accuracy multiplier that indicates a
more accurate performance compared with the LRNI condition. This
trend does not maintain for Subject D.D.
Ten dav recal 1—accuracy. Subject D.D. produces more accurate
responding under the HRNI condition at the 10 day measures of mainten¬
ance compared with the LRNI condition.
Question One Results for Experiment 2
Initially, all subjects have a greater accuracy multiplier index
under the high rate without inflection condition (HRNI) compared with
the low rate without inflection condition (LRNI). From Appendix C,

COUNT PER MINUTE
CALENDAR WEEKS
SUCCESSIVE CALENDAR DAYS
SlJI't hVHOR AUVIStH MAN AGÍ K III HAVI H Aí.f I A Hi I CIIIJN1III
GII'O'-IIDK AGfNCf I IMt A COUNILK CHARI £ti
Figure 5
Free Recall Accuracy

80
these subjects produce accuracy multiplier indices under the HRNI
condition that range from a times 1.44 to 3.2.
Three day recall--accuracy. For the three day maintenance
measure, three subjects produce more accurate responding under the
HRNI condition compared to the LRMI condition.
Ten day recaí 1--accuracy. Two subjects under the HRNI condi¬
tion produce more accurate responding compared with the LRNI condition.
For one subject there is no difference.
Question Two Results for Experiment 1
Question two: (A comparison between inflection conditions)
How does high rate of oral reading without inflection (HRNI) influence
comprehension and maintenance of comprehension compared with high
rate of oral reading with inflection (HRWI condition)?
From Figure 5 it can be seen that responding for Subjects D.D.
and J.J., under the HRWI condition, yields a greater accuracy ratio
compared with the HRNI condition. This trend is seen across all
three temporal measures of accuracy.
Question Two Results for Experiment 2
For Subjects S.B. and E.H. initially the HRWI condition produces
a greater accuracy ratio compared with the HRNI condition. For A.P.
there is no difference. Subject B.S. responds under the HRWI condi¬
tion by producing a better accuracy ratio compared with the HRNI
condition.

31
Three day recall--accuracy. For maintenance at three days,
Subjects A.P. and B.S. under the HRWI condition produce a greater
accuracy ratio compared with the HRNI condition. There is no
difference observed for Subject S.B. Subject E.H. under the HRWI
condition produces a greater accuracy ratio compared with the HRNI
condition.
Ten day recall—accuracy. For two subjects, the HRWI condition
produces more accurate responding at the ten day measures of main¬
tenance compared with the HRNI condition. The same measure for Subject
S.B. reveals a no difference effect. Figure 5 shows that for E.H.,
a steeper decelerating slope is present under the HRWI condition when
compared to the HRNI condition.
Question Three Results for Experiment 1
Question three: (A rate versus inflection comparison) How does
high rate of oral reading without inflection (HRNI) influence compre¬
hension and maintenance of comprehension compared with low rate of
oral reading with inflection (LRWI)?
Criteria reca11--accuracy. Subject D.D. initially responds
under the LRWI condition by producing an accuracy multiplier that
is 2.0 times greater compared with the HRNI condition. For Subject
J.J., initially the HRNI condition produces an accuracy ratio that
is 3.75 times greater than the one produced by the LRWI condition.
Three day recaíl--accuracy. Maintenance after three days for
D.D. is similar to the trends found under the initial measurement.
At three days, Subject J.J. under the HRNI condition produces an

82
accuracy multiplier that indicates a more accurate performance
compared with the LRNI condition.
Ten day recall--accuracy. At the ten day maintenance measure,
Subject D.D. under the HRNI condition produces an accuracy multiplier
of 3.25 compared with the LRWI condition. For J.J. a reverse trend
occurs. That is, the HRNI condition produces less accurate per¬
formance compared with LRWI condition.
Question Three Results for Experiment 2
From Figure 5 it can be seen that initially more accurate per¬
formance is produced by all subjects under the HRNI condition. The
accuracy multipliers range from a times 1.33 to a times 1.71.
Three day recall--accuracy. For three subjects, the HRNI
condition produces more accurate performances compared with the
LRWI condition at the three day measures of maintenance.
Ten day recall--accuracy. The ten day measure describes two
subjects under the HRNI condition as yielding more accurate performances.
For B.S. there is no difference, while for A.P. the LRWI condition
results in more accurate responding compared with the HRNI condition.
Question Four Results for Experiment 1
Question four: (A high and low rate comparison, with inflection)
How does high rate of oral reading with inflection (HRWI) influence
comprehension compared with low rate of oral reading with inflection
(LRWI)?

For both subjects, Figure 5 depicts high rate of oral reading
with inflection (HRWI condition) to produce more accurate respond¬
ing across all temporal measures compared with the LRWI condition.
Question Four Results for Experiment 2
As seen in Figure 5, three subjects under the HRWI condition
initially produce more accurate performances compared to the LRWI
condition. For Subject B.S. there is no difference.
Three day recall--accuracy. Accuracy at three days under the
HRWI condition produces more accurate performances for three subjects.
Subject E.S. produces the same accuracy ratio under either condition.
Ten day recal 1--accuracy. The ten day maintenance measures
for Subjects S.B. and E.H. under the HRWI condition produce more
accurate responding compared with the LRWI condition. For B.S.
there is no difference between the two conditions. This is also
true for Subject A.P.
Question Five Results for Experiment 1
Question five: (A comparison between the presumed best and
worst conditions) How does high rate of oral reading with inflection
(HRWI) influence comprehension and maintenance of comprehension compared
with low rate of oral reading without inflection (LRNI).
Subjects D.D. and J.J. generate data under the high rate with
inflection condition that have a greater accuracy ratio when compared
with the low rate without inflection condition for all measures of
accuracy.

84
Question Five Results for Experiment 2
The HRWI condition initially produces more accurate responding
for three subjects. For Subject B.S. there is no difference.
Appendix C shows the accuracy multipliers range from 1.0 to 3.77.
Three day reca!l--accuracy. Four subjects under the HRWI
condition produce more accurate responding at the three day main¬
tenance measure compared with LRNI condition.
Ten day recall--accuracy. More accurate respondí"ngs are also
produced by three subjects while under the HRWI condition at the 10
day measures of maintenance.
Question Six Results for Experiment 1
Question six: (A comparison between inflection and no inflection
at low rates) How does low rate of oral reading without inflection
(LRNI) influence comprehension and maintenance of comprehension
compared with low rate of oral reading with inflection (LRWI)?
Criteria recall--accuracy. For Subject D.D. the LRWI condition
initially produces an accuracy ratio that is 2.4 times greater than
the ratio produced by the LRNI condition. Subject J.J. generates
an accuracy ratio that is greater for the LRNI condition compared
with the LRWI condition.
Three day recall--accuracy. At three days both subjects are
producing a greater accuracy ratio under the LRWI condition.
Ten day recall--accuracy. Responses by D.D. under the LRNI
condition yields a greater proportion of number correct to incorrect
when compared to the LRWI condition at the 10 day measure. For
Subject J.J. there are no differences in the accuracy of responding.

85
Question Six Results for Experiment 2
Initial responding by Subjects E.H. and A.P. under the LRWI
condition produce a better accuracy ratio compared with the LRNI
condition. For Subject B.S. there are no differences.
Three day recall--accuracy. For three subjects at the three
day measures of maintenance the LRWI condition is producing a better
accuracy ratio compared with the LRNI condition.
Ten day recall--accuracy. At the 10 day measure, two subjects
under the LRWI condition produce more accurate responses compared to
the LRNI condition. For Subjects S.B. and B.S. the LRNI condition
produces more accurate performances.
Free Recall—Words Per Fact
Figure 6 depicts six subjects on their measures of the number
of words read for each fact retained.
For this measure, a lower value indicates that while reading
less words, subjects were recalling more facts. From Figure 6 it
can be seen that the high rate conditions initially have lower
values for all six subjects when compared to the low rate condi¬
tions. Three subjects are maintaining this trend at the three day
measures of maintenance. At 10 days, four subjects maintain this
trend while one subject produces no discernable differences.
Responding under the high rate condition with inflection more
frequently produces a lower value (more facts per word) compared
to the other conditions. Except for the HRWI condition, responding

COUNT PER MINJTE
CALENDAR WEEKS
succrsswi: caundar days
$Ul l hVI'.IlH AUVIMR MAMAU II
fl) M A VI << Al.i 1 AIM I
1*1 tom i oh
11 w III COUNT I II CIIARII II
Figure 6
Free Recall Words/Fact
20
r;i
' i:
, i
I
i
jji
MO
r.oiiMii o

87
under the high rate condition without inflection most frequently
produces the lowest values compared to other conditions. Respond¬
ing under the low rate v;ith inflection condition produces the
lowest values compared with the low rate without inflection condition.
Summary Free Recall Questions 1-6: Total
Performance, Accuracy, and Words Per Fact
From Figures 4 and 5 it can be seen that in general the high
rate condition with inflection (HRWI) produces faster and accurate
performances more often than the other conditions. The high rate
without inflection condition (HRNI) produces faster and more accurate
performances for facts recalled compared to all other conditions
except for the HRWI condition. The LRWI condition more often pro¬
duces more accurate and faster performances compared with the LRNI
condition. From Figure 5 it can be seen that the above trends
seem to hold for the words per fact measure. By simply comparing
the frequency multipliers of the yoked conditions (HRNI to LRNI,
and HRWI to LRWI) it becomes evident that inflection training had
a greater influence on the speed of recalling facts under the low
rate condition for most subjects. This comparison is only done for
the initial measure of performance. The initial measure of accuracy
for free recall describes inflection training to have a greater
influence on increasing the accuracy of recalling facts under the
high rate condition for most subjects.

88
Ten Comprehension Questions
The 10 comprehension questions are presented similar to the
above format.
Question One Results for Experiment One
Question one: (A comparison between two reading rates) This
question compares high rate of oral reading without inflection (HRNI
condition) with low rate of oral reading without inflections (LRNI
condition).
Immediate answers. From the bar graph in Figure 6 it can be
seen that for Subjects D.D. and J.J., the HRNI condition initially
produces a higher percent correct compared with the LRNI condition.
Three day measure. For both subjects, the HRNI condition pro¬
duces a higher percent correct on answers to comprehension questions
compared with the LRNI condition.
Ten day measure. At the 10 day measure of maintenance the HRNI
condition for D.D. has a greater number of correct answers compared
to the LRNI condition. For the same measure of maintenance Subject
J.J. under the HRNI and LRNI conditions produce the same number of
correct responses.
Question One Results for Experiment Two
Subjects S.B. and B.S. under the HRNI condition initially produce
more correct answers to comprehension questions compared with their
performances under the LRNI condition. For Subject A.P. the HRNI and
LRNI conditions initially show no differences.

Experiment i
JXD.
Figure 7
Number Correct 10 Comprehension Questions
CO

Experiment 2
SJB.
JO
10
JO
10
1
Vi
7
1
7
t
i
t
t
7
u
H
7
7
7
Í
6
<
%
W
Í
5
Í
5
0
5
H
$
7
i
*
3
'-Z
i
i
}
1
f
l
l
l
!
l
1
A A
c
x>
A A
c
®
A
B
C
T>
Comp >íairtí~ JJ)atp Alai rtf "JOJUip
ABC?
Corttp
5.5.
7
7
i
Í
Í
C
i
5
Vh
t
7
*
a
3
3
Z
2
/
ASCO
3 DajjS
ncp
Jiunt'/OAup
JO
CAP.
<1
7
7
¿4
i
i
7
•a
7
1
w
$
le
Í
$
-4
s
6
7
H
*
3
i
3
Z
f
l
t
z
t
A £ C D
Comp
ABC®
Aunt' 3 Buy
ABC®
AUnH0I\iiJs
A B
Co>
£>/.
c ®
>P
to
to
/O
7
1
7
i
7
7
7
7
7
6
rf
—
t>
(j
t>
Vs
â– *:
w
5
5
H
£
Vi
7
3
-J.
3
3
t
\
2
J
Z
1
A 6 C D
Aliíní—BIVui»
Figure 7
Number Correct 10 Comprehension Questions
A B c d
Aiaiiil~/OPays

91
Three day measure. All subjects under the HRNI condition for
the maintenance at three day measures produce more accurate
responding compared with the LRNI condition.
Ten day measure. Two subjects at the 10 day maintenance measure
yield a greater percent correct under the HRNI condition compared
with the LRNI condition.
Question Two Results for Experiment 1
Question two: (A comparison between inflection conditions)
This question compares high rate of oral reading without inflection
(HRNI) with high rate of oral reading with inflection (HRWI).
Immediate answers. Initially, responding under both conditions
produce no differences for either of the two subjects.
Three day measure. Subject D.D. under the HRWI condition
produces more accurate responding at the three day maintenance
measure compared with the HRNI condition. For J.J. the HR’II condi¬
tion produces more accurate responding to comprehension questions
during the three day maintenance measure.
Ten day measure. Subjects under the HRWI condition produce more
correct responses at the 10 day maintenance measure compared with the
HRNI condition.
Question Two Results for Experiment 2
Initially, four subjects yield more correct responding under the
high rate with inflection condition compared to the high rate with
inflection condition.

92
Three day measure. This trend maintains at the three day
measure for Subjects S.B. and E.H. For Subject A.P. there are no
differences between the two conditions. Subject B.S. under the
HRNI condition produces more accurate responding at the three day
measure.
Ten day measure. Two subjects yield a greater percent correct
under the HRWI condition at the 10 day measure compared with the HRNI
condition. For one subject there are no differences. Subject A.P.,
while under the HRNI condition, produces more accurate responses
compared with the HRWI condition.
Question Three Results for Experiment 1
Question three: (A rate versus inflection comparison) This
question is used to compare high rate of oral reading without inflec¬
tion (HRNI) with low rate of oral reading with inflection (LRWI).
Immediate answers. Initially, both subjects during the HRNI
condition, produce higher percents correct compared to their per¬
formances while under the LRWI condition.
Three day measure. At the three day measure the HRNI condition
increases Subject J.J.'s performance by 43 percent compared with the
LRWI condition. For D.D. there are no differences between the two
conditions.
Ten day measure. The 10 day maintenance measures describes
more accurate responding for J.J. while under the HRNI condition.
The inverse is true for Subject D.D.

93
Question Three Results for Experiment 2
Initially, two subjects produce a greater percent correct under
the HRNI condition compared with the LRWI condition. For S.B. there
is no difference in percent correct between the two conditions. For
E.H. responding under the LRWI condition produces a higher percent
correct compared with the HRNI condition.
Three day measure. Three subjects at the three day measure
have greater percents correct under the HRNI condition compared
with the LRWI condition. This trend maintains at the 10 day measures
of maintenance.
Question Four Results for Experiment 1
Question four: (A high and low rate comparison, with inflection)
This questions compares how rate of oral reading with inflection
(HRWI) compares with low rate of oral reading with inflection (LRWI).
For Subjects D.D. and J.J. the HRWI condition produces a higher
percent correct across all measures compared with the LRWI condition.
Question Four Results for Experiment 2
All subjects under the HRWI condition initially produce more
correct responding when compared to the LRWI condition.
Three day measure. Three subjects yield higher percents correct
under the HRWI condition compared with the LRWI condition at the
three day maintenance measures.
Ten day measure. Two subjects produce a higher percent correct
under the HRWI condition at the 10 day maintenance measures when

94
compared with the LRWI condition. Subjects A.P. and E.H. produce
no differences under either condition at the 10 day measures.
Question Five Results for Experiment 1
Question five: (A comparison between the presumed best and
worst conditions) This question compares high oral reading rate
with inflection (HRWI) with low oral reading rate without inflec¬
tion (LRNI).
Immediate answers. Both subjects under the HRWI condition
initially produce a higher percent correct compared with the LRNI
condition.
Three day measure. Subject D.D. yields a 50 percent increase
of corrects under the HRWI condition at the three day maintenance
measure. There is no difference in percent correct for Subject J.J.
Ten day measure. The HRWI condition produces a higher percent
correct for both subjects compared with the LRNI condition.
Question Five Results for Experiment 2
Initial responding under the HRWI condition yields a higher
percent correct for all subjects compared with the LRNI condition.
Three day measure. At the three day maintenance measure, three
subjects yield a higher percent correct under the HRWI condition.
Ten day measure. Three subjects at the 10 day measure produce
a higher percent correct under the HRWI condition compared with the
LRNI condition.

95
Question Six Results for Experiment 1
Question six: (A comparison between inflection and no inflec¬
tion at low rates) This question compares low rate of oral reading
without inflection (LRNI) with low rate of oral reading with inflec¬
tion (LRWI).
Immediate answers. For Subject D.D., there is no difference
in percent correct at the initial measure. For J.J., the LRWI condi¬
tion produces a higher percent correct.
Three day measure. At the three day maintenance measure, one
subject under the LRWI condition produces a higher percent correct
while the inverse is true for the other subject. This trend main¬
tains at the 10 day measure.
Question Six Results for Experiment 2
Initial responding under the LRWI condition for two subjects
yield a higher percent correct compared with the LRNI condition.
The LRNI condition produces a higher percent correct for the other
subjects.
Three day measure. At the three day maintenance measure the
LRWI condition increases the number correct for two subjects. For
the other two subjects there is no difference in the percent correct.
Ten day measure. At the 10 day maintenance measure two subjects
yield a higher percent correct under the LRWI condition. The
inverse occurs for the other two subjects.

96
Latency Measure Results
Figure 0 depicts the median latencies for subjects in
Experiments One and Two. A condition that shows the lowest
value is regarded as the fastest condition for answering compre¬
hension questions. (See Figure 8.)
Immediate latencies. Initially, subjects in both experiments
produce decreased latencies (answered questions faster) when under
the control of a high rate condition. High rate with inflection
(HRWI) initially influences responding to produce the fastest
latencies for five subjects.
Three day measure. At the three day maintenance measure,
five subjects under the HRWI condition produce the fastest
latencies compared to the other conditions.
Ten day measure. At the 10 day maintenance measure four
subjects are yielding decreased latencies under the HRWI condition
compared to the other conditions.
Summary of Results: Dependent Measures
of Comprehension
Responses under the high rate conditions for subjects in
Experiment One more often produces a higher percentage of correct
answers compared to other conditions. Of the high rate conditions,
the HRWI condition most often produces a higher percent correct
compared with the other conditions. The low rate with inflection
condition more often produces a higher percent correct compared with

COUNT PER MINU1
CALENDAR WEEKS
SUCCESSIVE CALENDAR DAYS
Sill mviSOfl AUVI&IR MANACt£ft IHIlAvEH A (it I AllII CDUNIIO
oiposirdft
All INC Y
Figure 8 ^
Median Latencies

98
the low rate without inflection condition. These trends are also
seen for the measures of the median latencies. Experiment Two
subjects produce results similar to Experiment One subjects. By
comparing the absolute number of correct responses of the yoked
conditions (high rate of oral reading without inflection to low
rate of oral reading without inflection and high rate of oral
reading with inflection to low rate of oral reading with inflection),
it becomes evident that inflection training had a greater influence
on increasing the absolute number of correct responses under the
high rate condition for most subjects. This trend also maintains
for the measures of median latency. In other words, inflection
training has a greater impact on increasing the speed of answering
comprehension questions under the high rate condition compared with
all other conditions.
Results Cloze Procedure--Total Performance
Since Experiment One subjects did not receive the Cloze procedure
the results are for subjects in Experiment Two.
Question One Results
Question one: (A comparison between two reading rates) How does
high rate of oral reading without inflection (HRNI) influence compre¬
hension and maintenance of comprehension compared with low rate of
oral reading without inflection (LRNI)?

99
Experiment Two subjects' speeds of performance for writing
answers to the Cloze questions are depicted in Figure 9.
Immediate performance. Initially, Subject S.B. performs faster
under the HRNI condition compared to the LRNI condition. From
Appendix E, it can be seen that this subject yields a frequency
multiplier of times 1.9. For E.H. there is no initial difference
between the two conditions. For Subjects A.P. and B.S. the LRNI
condition produces a faster response rate compared with the HRNI
condi ti on.
Three day measure. At the three day maintenance measures the
HRNI condition increases these subjects production of writing
facts compared with the LRNI condition. Subject B.S. under the
LRNI condition produces the same rate of responding as did her
performance under the HRNI condition.
Ten day measure. Maintenance at 10 days describes Subject A.P.
as producing a faster rate under the HRNI condition compared with
the LRNI condition. There is no difference for Subject B.S. Both
S.B. and E.ll. under the HRNI condition produce slower rates of
responding compared with the LRNI condition.
Question Two Results
Question two: (A comparison between inflection conditions)
How does high rate of oral reading without inflection (HRNI) influence
comprehension and maintenance of comprehension compared with high rate
of oral reading with inflection (HRWI).

0UN1 ÍJEK MINlJTF
CALENDAR WfEKS
Í.Ut'tWVIiOh All VIS i H MANACTK liHIAvlR ACC l Ahí l COUNJIÜ
DlEOSlTtlfl AGENCY flMtK COUNTER CIIARICR
Figure 9
Cloze Procedure Performance
o
O
100

101
Immediate performance. Two subjects initially produce faster
performances under the HRWI condition compared with the HRNI condi¬
tion. From the table in Appendix E, the HRWI condition for these
subjects yield frequency multipliers of 1.30 and 1.33. For one
subject the two conditions produces a no difference in rate of
responding. For Subject S.B. the HRNI condition yields a frequency
multiplier of 1.35 compared with the HRNI condition.
Three day measure. For the three day maintenance measures, the
HRWI condition outpaces the HRNI condition for two subjects. One
subject produces no difference between the two conditions. Tiie HRNI
condition produces, for Subject E.H., a faster rate of writing facts
compared with the HRWI condition.
Ten day measure. At the 10 day measure, three subjects are pro¬
ducing a faster rate under the HRWI condition compared with the HRNI
condition. Subject A.P., under the HRNI condition, produces a
faster rate of responding.
Question Three Results
Question three: (A rate versus inflection comparison) How
does high rate of oral reading without inflection (HRNI) influence
comprehension and maintenance of comprehension compared with low
rate of oral reading with inflection (LRWI)?
Immediate performance. Initially Subjects S.B., E.H., and
B.S. produce a faster rate of writing answers under the HRNI condi¬
tion compared with the LRUI condition. The frequency multipliers
to describe the differences are 1.28, 1.06, and 1.17, respectively.

102
Three day measure. The maintenance measure at three days
displays a faster rate under the HRNI condition for E.H. compared
with the LRHI condition. For two subjects there are no differences
between the two conditions.
Ten day measure. The 10 day measures for two subjects under
the HRNI condition produces increasing rates compared with the LRWI
condition. For one subject there are no differences between the two
conditions. For S.B. the LRWI condition is faster compared with
the HRNI condition.
Question Four Results
Question four: (A high and low rate comparison, with inflection)
How does high rate of oral reading with inflection (HRWI) influence
comprehension and maintenance of comprehension compared with low
rate of oral reading with inflection (LRWI).
Immediate performance. The HRWI condition initially produces
a faster rate of writing facts for Subject S.B. compared with the
LRWI condition. For the other subjects there are no initial differ¬
ences in speed of writing facts on the Cloze procedure under the HRWI
and LRWI conditions.
Three day measure. For the three day maintenance measures, one
subject under the HRWI condition produces a faster rate compared to
the LRWI condition. For the other three subjects there are no
differences.
Ten day measure. Maintenance measures at 10 days for three
subjects describes a faster performance under the HRWI condition com¬
pared with the LRWI condition.

103
Question Five Results
Question five: (A comparison between the presumed best and
worst conditions) How does high rate of oral reading with inflection
(HRWI) influence comprehension and maintenance of comprehension com¬
pared with low rate of oral reading without inflection (LRNI)?
Immediate performance. Two subjects under the HRWI condition
initially yield faster performances compared with the LRNI condition.
The LRNI condition initially produces a faster rate of performance
for Subjects A.P. and B.S. compared with the HRWI condition.
Three day measure. Two subjects under the HRWI condition at the
three day measure of maintenance produce a faster rate compared with
the LRNI condition. One subject does not produce a change in respond¬
ing. Subject B.S. produces faster rates under the LRNI condition.
Ten day measure. The 10 day maintenance measures describe
faster performances for three subjects under the HRWI condition
compared with the LRNI condition.
Question Six Results
Question six: (A comparison between inflection and no inflection
at low rates) How does low rate of oral reading without inflection
(LRNI) influence comprehension and maintenance of comprehension com¬
pared with low rate of oral reading with inflection (LRWI).
Immediate performance. Subject S.B. initially produces faster
rates under the LRWI condition when compared with the LRNI condition.
For E.H. there are no differences between the two conditions. Two
subjects initially produce a faster rate of writing responses under
the LRNI condition compared with the LRWI condition.

104
Three day measure. These trends maintain at the three day
measure, except that the two conditions made no difference on rate
for Subject B.S.
Ten day measure. For Subject E.H. measures of maintenance at
10 days describes the LRWI condition as producing a faster rate.
For the other subjects there are no differences.
Cloze Procedure Results--Accuracy
Question one: (A comparison between two reading rates) How
does high rate of oral reading without inflection (HRNI) influence
comprehension and maintenance of comprehension compared with low
rate of oral reading without inflection (LRNI)?
Figure 10 depicts the four subjects in Experiment Two with their
accuracy measures for every condition initially, maintenance at three
and ten days.
Immediate measure of accuracy. From Figure 10 it can be seen
that Subjects E.H. and A.P. under the HRNI condition initially produce
a greater accuracy ratio compared with the LRNI condition. From the
table in Appendix E, the accuracy multiplier for these subjects are
1.95 and 3.6, respectively.
Three day measure. At the three day maintenance measures, the
HRNI condition produces more accurate responding for two subjects
compared with the LRNI condition.

COUNT PER MINUTE
CALENDAR WEEKS
SUCCESSIVE CALENDAR DAYS
SIM E «VISOR AOVIStH M AM Atif R
HI IIAVEH
AGÍ LABEL COUNIfD
depositor
Ar.tMCv
Figure 10
Cloze Procedure Accuracy
o
Cn
SOL

106
Ten day measure. At the 10 day measure of maintenance the
HRNI condition produces the same accuracy ratio as the LRNI condi¬
tion for Subject B.S. For three subjects under the LRNI condition
produce a better accuracy ratio compared with the HRNI condition.
Question Two Results
Question two: (A comparison between inflection conditions)
How does high rate of oral reading without inflection (HRNI)
influence comprehension and maintenance of comprehension compared
with high rate of oral reading with inflection (HRWI)?
Immediate measure of accuracy. Four subjects initially produce
more accurate responding under the HRWI condition compared with the
HRNI condition for writing correct responses on the Cloze procedure.
Three day measure. At the three day maintenance measures,
two subjects produce a better accuracy ratio under the HRWI condition.
Ten day measure. The 10 day measure for Subjects S.B., E.H.,
and B.S. show that the HRWI condition increases accurate responding
compared with the HRNI condition.
Question Three Results
Question three: (A rate versus inflection comparison) How does
high rate of oral reading without inflection (HRNI) influence compre¬
hension and maintenance of comprehension compared with low rate of
oral reading with inflection (LRWI)?
Immediate measure of accuracy. Subjects S.B., E.H., and A.P.
initially produce a greater accuracy ratio' under the HRNI condition
compared with the LRWI condition.

107
Three day measure. At the three day maintenance measures
Subjects S.B., E.H., and A.P. produce a higher accuracy multiplier
under the HRMI condition compared with the LRWI condition.
Ten day measure. Subject S.B. produces an increase in accurate
performance at the 10 day maintenance measure under the HRNI condi¬
tion compared with the LRWI condition. For one subject the same
accuracy ratio occurs.
Question Four Results
Question four: (A high and low rate comparison with inflection)
How does high rate of oral reading with inflection (HRWI) influence
comprehension and maintenance of comprehension compared with low
rate of oral reading with inflection (LRWI)?
Immediate measure of accuracy. Four subjects initially produce
increasing accurate performances under the HRWI condition compared
with the LRWI condition.
Three day measure. This trend maintains at the three day measure
of maintenance except for Subject A.P. where the same accuracy ratio
occurs.
Ten day measure. The HRWI condition produces a greater accuracy
multiplier for three subjects at the 10 day maintenance measures.
Question Five Results
Question five: (A comparison between the presumed best and
worst conditions) How does high rate of oral reading with inflection
(HRWI) influence comprehension and maintenance of comprehension com¬
pared with low rate of oral reading without inflection (LRNI)?

108
Immediate measure of accuracy. Four subjects initially produce
more accurate performance under the HRWI condition compared with the
LRNI condition.
Three day measure. Three subjects at the three day measures
produce a greater accuracy multiplier under the HRWI condition
compared with the LRNI condition.
Ten day measure. At the 10 day measures of maintenance four
subjects are yielding more accurate responding under the HRWI condi¬
tion.
Question Six Results
Question six: (A comparison between inflection and no inflection
at low rates) How does low rate of oral reading without inflection
(LRNI) influence comprehension and maintenance of comprehension
compared with low rate of oral reading with inflection (LRWI)?
Immediate measure of accuracy. Three subjects initially produce
more accurate performances under the LRWI condition compared with the
LRNI condition.
Three day measure. At the three day measures Subject A.P.
produces more accurate performances under the LRWI condition. The
same accuracy ratio occurs for Subject E.H.
Ten day measure. Maintenance at 10 days describes greater
accuracy ratios for Subjects S.B. and A.P. under the LRNI condition
compared with the LRWI condition. Subject B.S. produces the same
accuracy ratios under both conditions.

109
Results Cloze Procedure-Percent Correct
Figure 11 is a bar graph describing subjects' percent of
correct responding under each condition.
Compared with all other conditions and across subjects, the
high rate with inflection condition more often produces a higher
percent of correct responding. Compared with the LRNI and LRWI
conditions, the HRMI condition more often shows a higher percent
correct. When comparing the low rate condition, the LRNI condition
has more gains than the LRWI condition. Subject S.B. is not typical
in responding under the LRNI condition when compared to the other
subjects. The other subjects more often produce a higher percent
correct under the LRWI condition compared to the LRNI condition.
Summary of Results--Cloze Procedure
From Figures 9, 10, and 11 it can be seen that the high rate
conditions most often produce a faster and more accurate rate of
responding when compared with the other conditions. The high rate
condition with inflection most often produces the fastest and most
accurate responding compared to the other conditions. The low rate
with inflection condition produces faster rates of performance only
slightly more often than the low rate without inflection condition.
The LRWI condition for most subjects more often produces accelerating
accuracy ratios compared with the LRNI condition. 3y comparing the
frequency multipliers of the yoked conditions inflection training had

S.J}
£>/.
<00
*
*
r¡ 7a
Jü
1*
^ lo
A B C P
Como.
cAT.
A B C D
^a»nI-3P.
10
1o
(pO
5o
Ho
io
%o
u>
AÍtW~/PPibj*
35.
loo
9o
<¡o
io
te
—
10
íí
lo
7a
Lo
3
3
*
to
k>
‘Jo
i
5
vi
¡>0
ix>
W
lo
30
3o
30
XO
XO
IO
lo
IO
c
p
A
B
C
p
A
B
c
P
A
P
c
p
I co
C-Otnp
Mtint~3 Ilkp Muint ~iO
Figure ll
Cloze Percent Correct

no influence on producing faster rates of writing responses on the
Cloze test. This is only performed for the initial measures.
The initial measure of accuracy for free recall describes inflec¬
tion training to have a greater influence on increasing the
accuracy of recalling facts under the high rate condition for
most subjects when compared with the low rate conditions.
Summary of Results Experiments One and Two
For these experiments the high rate with inflection condition
(HRWI) more often than the other conditions influences subjects to
comprehend more text, more quickly and more accurately. The HRWI
condition also increases their response time when answering compre¬
hension questions and to read fewer words and retain more facts.
The high rate without inflection is the other condition that pro¬
duces trends similar to those found under HRWI condition; however,
HRNI condition did not produce these trends as often as HRWI condi¬
tion. Subjects, when influenced by the low rate with inflection
condition (LRWI condition), more often produce better performances
compared with the low rate without inflection condition (LRNI condi¬
tion). In general, inflection has a bigger impact on producing
desirable performances for the high rate condition than it did for
the low rate condition. However, inflection increases subjects'
performances under both the high and low rate conditions.
As an aid to interpretation a table depicting the total number
of gains is presented (see Table 4).

112
Table 4
Summary Counts of Better Performance in Comparison to Each Condition*
Cond
DD
JJ
SB
AP
Cond
EH
BS
Cond
Total**
Possible
Free Recal 1--Accuracy
HRNI
3
5
6
5
HRWI
7
3
HRNI
28
54
Multi pi iers
LRNI
2
2
3
1
LRWI
3
3
LRNI
12
54
HRVJI
9
8
9
7
HRNI
6
3
HRWI
43
54
LRWI
4
2
0
5
LRNI
0
4
LRNI
17
54
Free Recall —
HRNI
3
5
7
8
HRWI
8
2
HRNI
26
54
Performance: Freq.
LRN I
2
2
0
0
LRWI
2
3
HRNI
10
54
Mult.
HRWI
9
8
8
6
HRNI
5
3
HRWI
41
54
LRWI
4
2
2
3
LRNI
2
4
LRWI
16
54
Free Recal 1--Words/
HRNI
2
6
7
8
HRWI
3
3
HRNI
36
54
Fact
LRNI
1
0
7
8
LRWI
3
5
LRNI
20
54
HRWI
8
9
8
6
HRNI
8
5
HRWI
37
54
LRNI
6
3
1
4
LRNI
0
4
LRWI
22
54
10 Comprehension
HRNI
4
6
5
5
HRWI
8
8
HRNI
25
54
Questions
LRNI
0
2
1
2
LRWI
7
2
LRNI
9
54
HRWI
9
6
9
7
HRNI
1
4
HRWI
47
54
LRWI
3
1
3
3
LRNI
3
1
LRWI
19
54
Median Latencies
HRNI
9
5
4
4
HRWI
9
8
HRNI
27
54
LRNI
2
2
5
1
LRWI
3
3
LRNI
18
54
HRWI
5
9
9
9
HRNI
2
3
HRWI
49
54
LRWI
2
1
0
3
LRNI
4
4
LRWI
12
54
Cloze Performance:
HRNI
4
3
HRWI
7
3
HRNI
17
36
Freq. Mult.
LRNI
2
7
LRWI
0
5
LRNI
21
36
HRWI
8
5
HRNI
5
5
HRWI
23
36
LRWI
4
4
LRNI
6
6
LRWI
13
36
Cloze Accuracy
HRNI
3
5
HRWI
8
8
HRNI
15
36
Mul tipi iers
LRNI
5
2
LRWI
6
4
LRNI
n
36
HRWI
8
7
HRNI
2
5
HRWI
31
36
LRWI
0
3
LRNI
2
2
LRWI
13
36
Cloze Percent Correct
HRNI
2
5
HRWI
8
8
HRNI
14
36
LRNI
7
2
LRWI
4
3
LRNI
15
36
HRWI
8
7
HRNI
2
5
HRWI
31
36
LRWI
1
4
LRNI
4
2
LRWI
12
36
* Better means any functional adgantage such as more speed, more accuracy, and
shorter latency.
**Total = The number of times a particular condition within a dependent
measure has a better performance in comparison with all other conditions.

113
Table 4 depicts how often each condition improved performance
when compared with the other conditions for the same dependent
measure. This was done for all subjects under each condition.
Condition HRWI always produces the best results across all dependent
measures.

CHAPTER V
DISCUSSION
The findings of this study are discussed relative to the
experimental questions. The questions are listed and a brief
discussion follows.
1. How does high rate of oral reading without inflection (HRNI)
influence comprehension and maintenance of comprehension compared with
low rate of oral reading without inflection (LRNI)?
Question one sought to determine the influence that rate has on
comprehension and retention when inflection was held constant. The
data presented indicate that generally high rate of oral reading
without inflection increased comprehension and maintenance of compre¬
hension. This is noted on dependent measures that yielded measures
of accuracy, fluency, response time to questions, and the number of
words read to retain a fact. The data in Table 4 provide a summary
of how often a particular condition had a greater impact on the
dependent variables compared with all other conditions. In total
there were eight measures. Subjects under the HRNI condition improved
their performance when compared with the LRNI condition on six
of the eight measures. Of a possible 378 gains, the HRNI condi¬
tion had 188 total gains or 50 percent while the LRNI condition
produced 116 gains or 23 percent. When compared with each other,
the HRNI condition was 62 percent better than the LRNI condition.
114

115
In light of these findings, it is reasonable to suggest that when
oral reading at least approaches the speed of conversational speech,
comprehension and retention improve.
2. How does high rate of oral reading without inflection (HRNI)
influence comprehension and maintenance of comprehension compared with
high rate of oral reading with inflection (HRWI)?
Question two sought to determine if oral reading with inflection
at a high rate can increase comprehension and maintenance of compre¬
hension when compared with the same high rate of oral reading, but
without the inflection. The HRWI condition on all eight measures
outperformed the HRNI condition. High rate with inflection con-
si stantly improved performance on the dependent measures when
compared with high rate without inflection. Of a possible 378 gains,
the HRWI condition produced 302 or 80 percent of the gains while the
HRNI condition produced 188 gains. In other words, the HRWI condi¬
tion had 61 percent of the gains when compared with the HRNI condition.
3. How does high rate of oral reading without inflection (HRNI)
influence comprehension and maintenance of comprehension compared
with low rate of oral reading with inflection (LR1JI)?
High rate of oral reading without inflection more often increased
comprehension and maintenance of comprehension when compared with the
LRWI condition. The HRNI condition produced 52 percent more gains when
compared with the LRWI condition. The LRWI condition had a total of
124 gains or 33 percent of the total number of possible gains. Thus,
inflection with low rate of oral reading seemed to improve comprehension

H6
and maintenance, but never as often as did high rate with or without
infiection.
4. How does high rate of oral reading with inflection (HRWI)
influence comprehension and maintenance of comprehension compared
with low rate of oral reading with inflection {LRWI)?
This question sought to determine if inflection at high and low
rates could produce differences on measures of comprehension and
maintenance of comprehension. The data indicated that high rate of
oral reading with inflection more often produced improved comprehen¬
sion and maintenance. On all eight measures the HRWI condition
outperformed the LRWI condition. When compared with each other, the
HRWI condition produced 1 1/2 times more gains than did the LRWI
condition.
5. How does high rate of oral reading with inflection (HRWI)
influence comprehension and maintenance of comprehension compared
with low rate of oral reading without inflection (LRNI)?
The answer to this question provides information as to how
rate differences with and without inflection could impact the dependent
measures. On all eight measures the HRWI condition outperformed the
LRNI condition. The HRWI condition was 2.6 times better than the LRNI
condition. The data strongly supports that low rate without inflec¬
tion does not yield higher gains in comprehension and retention when
compared with reading at conversational rate along with inflection.
6. How does low rate of oral reading without inflection (LRNI)
influence comprehension and maintenance of comprehension compared with
low rate of oral reading with inflection (LRWI)?

117
The answer to this question provides information on how inflec¬
tion while reading slow can either improve or inhibit comprehension
and its maintenance over time. The LRWI condition on five of the
eight measures showed a better performance compared with the LRNI
condition. The results indicate that inflection at low rates does
improve comprehension and maintenance. The LRWI condition was 7
percent better than the LRNI condition. However, the difference
is not as great as the HRWI condition compared with the HRNI.
Inflection effects on the dependent measure. An analysis was
performed to determine under which rate did inflection training have
the greater impact on comprehension and maintenance of comprehension.
The HRWI condition on all eight measures increased comprehension and
maintenance of comprehension compared with the HRNI condition. The
LRWI condition on five of the eight measures increased comprehension
and maintenance compared with the LRNI condition. Thus, it appears
that inflection has a greater effect on comprehension and maintenance
at high rates than it does at low rates.
Qualitative findings. At the conclusion of this study all
subjects upon interview reported that reading fast.was easier. Also,
most subjects indicated that they understood the story better when
asked to inflect at high rates. Appendix F contains transcripts of
the interviews. Many times after rereading a fast rate story the
subjects upon request recited lines and passages in the exact order
and using the same vocabulary. This rarely happened when they per¬
formed the same task after reading stories slowly. Many times the
facts provided from a slow condition story were out of sequence and

118
told at a slower pace. The effects on concurrent off task operants
were especially interesting during the reading slow conditions.
Many of the subjects while reading slowly exhibited an increased
frequency of competing responses such as head scratches, loosing
their place, erasing, slower response time to reread a story, and
asking for the time.
Variability
By visually inspecting the tables and figures in this paper
one can observe the variability of responses both within and across
subjects. One subject in particular produced data that was
atypical of the others. Once the study was under way, it was
discovered that Subject B.S. had an articulation problem. She was
being seen by a speech pathologist. At first this was thought not
to be a significant variable; however, as the measurements were
being taken it soon became evident that her rate and accuracy of
vocally producing facts under the high rate conditions were differ¬
ent from the other subjects. Inspection of the other dependent
measures revealed a reverse trend; that is, she was performing much
like the other subjects on the dependent measures other than free
recall.
To offer an explanation of these phenomena required additional
investigation by the experimenter. By contacting speech pathologists
it became evident that children with articulation problems generally
have difficulties recalling facts and experiences (Klein, 1983) and
speaking at rates that approximate "normal" conversational speech

119
(Thomsem, 1983). Perhaps this happens because oeople ask these
children to slow down or to repeat what was said. When this
happens children may be receiving reinforcement for slowed speech
or punishment for speaking, especially when they are only asked to
repeat what they have previously said. In this study, the subjects
did not practice saying facts at high rates nor were they reinforced
for it. Moreover, they were not told what a fact was. They were
only asked to tell the experimenter everything they could about a
story.
Variability was also noticed between Experiments One and Two.
Experiment 1 subjects typically provided data that were more orderly
and systematic than those provided by subjects in Experiment 2.
Also, the effects were larger with the Experiment 1 subjects. This
seems to suggest that the high rate conditions especially with
inflection were more beneficial to the older subjects who were
reading significantly below their age levels.
There also was a difference between the number of trials it took
to reach criteria for each condition between the subjects in the two
experiments. Subjects in Experiment 1 typically required twice as
many trials to reach criteria at the high rate conditions than it
did for the younger subjects except for Subject B.S.
The results also indicate that the subjects in this study
typically performed better than others of similar training, age,
and reading levels. For example, data on the Cloze procedure indicate
that initially no subject had less than 70 percent correct. In fact,

120
no subject had less than 50 percent correct during any of the
measures. This compares favorably with the findings reported by
other researchers. For example Howell (1983) indicated that for
subjects who did not have any experiences with the Cloze procedure,
40 percent correct on the Cloze was considered passing on his
criterion referenced test. The performances of subjects in this
study may be due to the number of rereadings to which they were
exposed. For example, every subject reread a story until they
reached criteria. The least number of times a subject reread a
story was four.
Another source of variability may be due to the nature of the
Cloze procedure. Since, when taking the Cloze test, only every
fifth word is deleted, subject in effect have an opportunity to
reread the story. This may also account for the seemingly elevated
performances. Also, the Cloze may have also contributed to the
reminiscence effect that was observed at the three and 10 day
period, especially for the low rate conditions.
Replications
Replications within experiments. For all subjects high rate
with inflection most often produced increased performances. High
rate in itself was potent enough to produce improved performances
compared with the low rate conditions. Inflection for most subjects
helped to improve performances within the respective rate condition.
In other words, although low rate with inflection typically did not
improve performance compared with the high rate conditions, it did

121
improve performance when compared with the low rate without inflec¬
tion. This trend was also true for both subjects under the high
rate conditions.
Replications between Experiments One and Two. The trends are
similar within and across Experiments One and Two. The HRWI condi¬
tion for Experiments 1 and 2 was the best condition. The next best
condition for these experiments was the HRNI condition. The HRWI
condition is associated with 93 percent of the gains in Experiment 1.
Seventy-six percent of the gains were for the HRWI condition in
Experiment 2. The next best condition in Experiment 1 was the HRNI
condition. This condition produced 60 percent of the gains reported
in Experiment 1. In Experiment 2 the HRNI condition captured 50
percent of the gains reported. Thus, these results were replicated
across subjects, age, sex, reading levels, reading materials, race,
and experimental designs.
Replications across skills. In Experiment 1 the dependent
measures were speed of free recall (performance), accuracy of free
recall, words per fact, correct answers to the 10 comprehension
questions, and the median latencies of responding to the 10 comprehen¬
sion questions. Each of the dependent measures should be understood
as being a separate skill. All of these measures occurred at three
different times for each subject under each condition. The performance,
accuracy, and the words/fact measures of the free recall were derived
from measuring facts recalled per minute. The median latencies were
derived from the episode of answering comprehension questions. In

122
addition to the above measures, Experiment 2 subjects were given a
Cloze procedure. The results from these experiments indicated that
high rate of oral reading with inflection more often produced
improved responses for these skills compared with the other
conditions. For Experiment 2 subjects, the rate and accuracy of
writing responses were also improved by the high rate conditions
and more so under the high rate with inflection condition. Thus,
a within and across subject generalization of skills was produced.
Replications across studies. This study sought to replicate,
in part, and extend the findings of Lovitt and Hansen (1976). Also,
it sought to partially replicate and extend the findings of Berquam
(1981). Lovitt and Hansen discovered that as the rate of oral
reading increased, there was a simultaneous improvement in compre¬
hension. Berquam found that practice alone on a paired associates
task (nonsense trigrams) did not improve performance as well as
daily practice that sought to increase the rate of accurate respond¬
ing. This study sought to determine which rate of oral reading would
produce the best comprehension and maintenance of comprehension if
practice were kept equal. Also, when the influence of rate could
be determined, the role of inflection could then be described.
Since in Lovitt and Hansen's study (1976), they did not describe
the influence that inflection may have had, the present work then
extends their work. Since Berquam did not use reading materials
and did not yoke the amount of practice, this study then extends his
researcn.

123
The findings of this study did support Lovitt and Hansen's
(1976) findings. Increasing oral reading rates improves compre¬
hension compared with reading at slow rates. Also, inflection
with high rate of oral reading increases comprehension and main¬
tenance of comprehension more so than does high rate of oral reading
without inflection. The findings of this study also support
Berquam's findings and permit his results to be generalized to
reading. When subjects are given the same number of opportunities
to read at high rate as well as low rate, they typically improve
comprehension and retention of facts more under high rate than low
rate of oral reading.
Several specialists in the area of reading have regarded read¬
ing fluently as an important skill for children (Samuels, 1979;
Smith, 1971). Samuels (1979) and his idea of automaticity has
demonstrated the need for repeated readings and also the need to
increase fluency. This study lends support to the concept that
when reading increases in fluency or when it approximates conversa¬
tional speech, comprehension and what is retained will be improved.
Although this study used repeated readings, its effect on comprehen¬
sion was not presented as an experimental question. However, some
generalizations can be made. Howell (1983) reported that criteria'^,
for passing a Cloze procedure is 40 percent correct. The subjects in
the present study always had at least 55 percent correct on the Cloze
procedure regardless of the last time they were exposed to the story.
Since the results of the Cloze procedure compares favorably
with Howell's findings it should be noted that the Cloze in this

124
study was done silently. Thus, as mentioned in Chapter Two, the
results of this study appear to generalize to silent reading.
Practical Implications
The results from this study support the notion that when
reading sounds like conversational speech it will increase compre-
henion and maintenance of comprehension. Put another way, when
reading is as fast and uses inflection similar to conversational
speech it may take on similar functional properties. The data from
this study indicate that a goal in reading instruction with students
who have normal speech and hearing should be having the reader learn
to read as if she/he is speaking. It is usually reasonable to assume
that teaching should begin with the complex skill first (Lindsley,
1983). If the student has trouble reading at high rate with inflec¬
tion, then the initial training focus should be on developing accuracy
and speed. The data in this study supports the idea that high rate
of oral reading, even without inflection, will increase comprehension
and maintenance of comprehension more often when compared with
reading slow with or without inflection. Once the student has
developed and knows what reading fast is like, inflection training
along with high rate should increase performance.
Implications for the School Psychologist
Bijou (1969) wrote that the school psychologist should work to
prevent school retardation (p. 71). He advocated the use of ABA as
a functional strategy to meet this end. Verbal Behavior (Skinner,

125
1957) was an application of the strategies of the experimental
analysis of behavior to explain language. By using the concepts
that were made explicit in Verbal Behavior (Skinner, 1957) school
psychologists can increase their capacity to prevent school
retardation.
As school psychologists become more involved with the reading
process there may be a concomittant decrease in the number of
assessments that may be required of them. P.L. 94-142 requires
that interventions be performed before a child is considered for
an evaluation for possible placement into special education. This
law can now serve as an important stimulus to which psychologists
can begin using their expertise in an area such as reading. Bijou
(1968) wrote about the prevention of school retardation. Moreover,
he recently indicated that only about 15 percent of the total popula¬
tion of mentally retarded individuals are retarded because of some
physical problem (Bijou, 1982). The rest simply require more effec¬
tive methods of training new skills. By becoming knowledgeable about
effective reading interventions school psychologists may be able to
decrease the number of referrals made to special education and to be
better able to serve the needs of their schools.
Generalizations
This study sought to develop and systematically replicate four
functional relationships. These replications occurred across
subjects, age levels, reading levels, reading materials, sex,
several related skills collectively known as comprehension, and

126
design. This study should generalize at least to other students
similar to the ones used in this study. This study is limited
only in its application to other variables not studied in this
research. It is important to note that in a previous study,
Lovitt and Hansen (1976) reported similar results for the effects
of high rate with learning disabled students. Thus, the effects
of high rate of oral reading appears to be gaining a strong empirical
base for its efficacy across a widening population. With this study,
the efficacy of using inflection along with high rate of oral read¬
ing has begun to develop an empirical base.
Suggestions for Future Research
The foregoing research stimulates a variety of questions related
to oral reading research and comprehension. Much research is needed
to determine the parameters of the variables that increase general¬
ization of skills that are taught through oral reading.
The first requirement for any science is that of replication
for it is the soundest empirical test of reliability (Johnston &
Pennypacker, 1980; Sidman, 1960). Single subject designs lend
themselves well to replication. Because of the reliance on experi¬
mental control, the great number of observations on relatively few
subjects, the strategy of holding many variables constant while
manipulating only one, and the technologically exact descriptions
of the procedure, replication is possible.

127
Systematic replication (Sidman, 1960) of the present study
should provide more detailed information of how high rates of
oral reading with and without inflection relate to comprehension
and maintenance of comprehension. First, replications with children
of different reading and/or grade levels should be performed. This
is thought to be important because the more a relationship is
replicated with a divergent population, the more widely and
confidently it may be applied.
Second, the effects of fast paced vocal stimuli have been shown
to increase a listening audience's comprehension (ABC Nightly News,
1983). A study could be done to determine how listening to fast
paced oral reading with and without inflection could be used to improve
oral reading comprehension of handicapped and/or nonhandicapped school
children.
Third, this study did not train the subjects on skills such as
free recall and Cloze procedure. It would be interesting to see if the
absolute magnitude of change would have been different under each
condition relative to other conditions. This would also help to
confirm the explanation that children with speech defects, such as
one subject in this study had, would do appreciably better under the
control of high oral reading with inflection when specifically train¬
ed to provide fluent answers to questions that are open ended.
Fourth, an investigation which can modify the procedures in
this study to be used in a larger group setting will help to deter¬
mine the generalizability of the findings.

128
Fifth, since this study used rereading as part of the pro¬
cedure, an investigation can be developed to see how much of the
results were due to rereading. The results can then be used to
determine more precisely the magnitude in which high rate of oral
reading contributes to comprehension and maintenance of comprehen¬
sion.
Finally, this study does not have a referent to which the
results could be compared. A study which uses the same measurement
strategies and tactics as those used in this study would help deter¬
mine more precisely how the results of this study compare with the
typical methods that are used to teach reading.
The relationship between high rate of oral reading and compre¬
hension and retention has a good empirical base to support its use.
Teachers are encouraged to use strategies that can increase the rate
in which students read text. The data presented in this study strongly
suggest that students be trained to read orally at high rates with
inflection.

REFERENCES
Alpa, T., Nowlin, L., Lemonine, K., Perine, M., & Bettencourt, B.
The rated assessment of academic skills Academic Therapy,
1974, 9(3), 151-164.
Armstrong, S. W. The effects of oral reading rate and reinforcement
on reading comprehension (Doctoral dissertation, University of
Florida, 1981).
Ayllon, T., & Azrin, N. The token economy: A motivational system
for therapy and rehabilitation. New York: Appleton-Century-
Crofts, 1969.
Ayllon, T., & Roberts, M. D. Eliminating discipline problems by
strengthening academic performance. Journal of Applied
Behavioral Analysis, 1974, _7(1), 71-76.
Azrin, N. H., & Lindsley, 0. R. Reinforcement of cooperation between
children. Journal of Abnormal and Social Psvcholoay, 1956, 52
(1), 100-102.
Baer, D. M., Wolf, M. M,, & Risley, T. R. Some current dimensions of
applied behavior analysis. Journal of Applied Behavior Analysis,
1968, 1(1), 91-97.
Bailey, J. S. A handbook of research methods in applied behavior
analysis. Tallahassee: Florida State University, 1977.
Beck, R. Personal communication. Precision Teaching Conference,
Orlando, Florida, March 1982.
Benjamin, R. Making schools work. New York: The Continuum, 1981.
Berquam, E. M. The relation between frequency of response and retention
on a paired associates task (Doctoral dissertation, University of
Florida, 1981).
Bijou, S. W. What psychology has to offer education--now. Journal
of Applied Behavior Analysis, 1970, 3(1), 65-75.

130
Bijou, S. W. Personal communication. Florida Association of
Behavior Analysts, 1982.
Bijou, S. W., & Baer, D. M. Child development: A systematic and
empirical theory. Mew York: Appleton-Century-Crofts, 1961.
Black like me--series. Chicago: Science Research Associates, 1976.
Bloom, B. S., Engelhart, M. D., Furst, E. J., Hill, W. H., &
Krathwohl, D. R. Taxonomy of educational objectives. New York:
David McKay, 1956.
Bos, C. S. Getting past decoding: Assisted and repeated readings
as remedial methods for learning disabled students. Topics in
Learning and Learning Disabilities, 1982, 1_(4), 51-57.
Brigance, A. H. Diagnostic inventory of basic skills. North
Bellerica, MA: Curriculum Associates, 1980.
Buss, M. Does reading rate correlate with reading comprehension?
Unpublished manuscript, University of Florida, 1982.
Chomsky, N. Aspects of the theory of syntax. Cambridge, MA: MIT
Press, 1965.
Durrel1, D. D. Development of comprehension and interpretation.
Reading in elementary school. Chicago: National Society of
the Study of Education, 1949.
Ebbinghaus, H. Memory: A contribution to experimental psychology
(H. A. Ruger a C. E. Bussenius translation) Mew York: Teachers
College, Columbia University, 1913. (originally published, 1885)
Engelmann, S. Preventing failure in the primary grades. Chicago:
Science Research Associates, 1975.
Engelmann, S., & Bruner, E. C. DISTAR reading 1: An instructional
system (2nd ed.). Chicago: Science Research Associates, 1974.
Eshelman, J. Self-corrective educational software: On line instruc¬
tional decisions by microcomputers. Paper presented at the
Precision Teaching Conference, Orlando, Florida, March 11, 1983.
Ferster, C. B., & Skinner, B. F. Schedules of reinforcement. New
York: Appleton-Century-Crofts, 1957.
Fleisher, L., Jenkins, R., & Pany, D. Reader's comprehension of train¬
ing in rapid decoding. Reading Research Quarterly, 1979, 15(1),
30-48.
Fry, E. B. Reading instruction for classroom and clinic. New York:
McGraw-Hill, 1972.

131
Ginn reading 720 series. Lexington, MA: Ginn and Co., 1976.
Goodman, K. Reading: A psycholinguistic guessing game. In H.
Singer & R. B. Ruddell (Eds.), Theoretical models and processing
of reading (2nd ed.). Newark, DE: International Reading
Association, 1976.
Grimes, J. Shaping the future of school psychology. The School
Psychology Review, 1981, J_0(2), 205-231.
Guszak, F. J. Questioning strategies of elementary teachers in
relation to comprehension. In L. A. Harris & C. B. Smith (Eds.),
Individualizing reading instruction. New York: Holt, Rinehart
and Uinston, 1369.
Hansen, C. L., & Lovitt, L. C. The relationship between question type
and mode of reading on the ability to comprehend. Journal of
Special Education, 1976, 10(1), 53-60.
Hansen, C. L. Chicken soup and other forms of comprehension. In
J. E. Bueton & 0. T. Rowland (Eds.), Communications research in
learning disabilities and mental retardation. Baltimore:
University Park Press, 1979.
Haring, N. G. Research in the classroom: Problems and procedures.
In N. G. Haring, T. C. Lovitt, M. D. Eaton, & C. L. Hansen
(Eds.), The 4th R: Research in the classroom. Columbus, OH:
Charles E. Merrill, 1978.
Hasselriis, P. I.E.P.s and a whole-language model of language arts.
Topics in Learning and Learning Disabilities, 1982, 1(4), 17-21.
Hauck, B., Metcalfe, J., & Bennet, P. Teaching early comprehension
skills: A case study. In N. G. Haring, T. C. Lovitt, M. D.
Eaton, & C. L. Hansen (Eds.), The 4th R: Research in the
classroom. Columbus, OH: Charles E. Merrill, 1978.
Haughton, E. Personal communication. Precision Teaching Conference,
Orlando, Florida, March 1982.
Haughton, E., & Binder, C. Personal communication. Precision Teaching
Conference, Orlando, Florida, March 1983.
Heilman, A. W. Principles and practices of teachinq readinq. Columbus,
OH: Charles E. Merrill, 1967.
Heilman, A. U., Blair, T. R., & Rupley, W. H. Principles and practices
of teachinq readinq (5th ed.). Columbus, OH: Charles E.
Merrill, 1981.
Howell, K. Personal communication. Precision Teaching Conference,
Orlando, Florida, March 12, 1983.

132
Indrisano, R. An ecological approach to learning. Topics in
Learning and Learning Disabilities, 1982, 1_(4), 11-15.
Jenkins, J. R., Barksdale, A., & Clinton, L. Improving reading
comprehension and oral reading: Generalization across behaviors,
settings and times. Journal of Learning Disabilities, 1978, 11
(10), 607-617. ~
Johnson, K. R., & Chase, P. N. Behavior analysis in instructional
design: A functional typology of verbal tasks. The Behavior
Analyst, 1981, 4(2), 103-121.
Johnston, J. M., & Pennypacker, H. S. Strategies and tactics of
human behavioral research. Princeton, NJ: Lawrence Erlbaum
Associates, 1980.
Kaplan, M. S., Chrin, M., & Clancy, B. Priority roles for school
psychologists as seen by superintendents. Journal of School
Psychology, 1977, 150 ), 75-80.
Karlin, R. Teaching elementary reading: Principles and strategies.
New York: Harcourt Brace Jovanovich, 1975.
Keogh, B., Kukic, S. J., Becker, L. D., Mcloughin, J. J., & Kukic,
M. B. School psychologists' services in special education
programs. Journal of School Psychology, 1975, 23.(2), 142-148.
Klein, M. Personal communication. University of Florida, April 4,
1983.
Knapczyk, D. R., & Livingston, G. The effects of prompting question
asking upon on-task behavior and reading comprehension.
Journal of Applied Behavior Analysis, 1974, 7.(1), 115-121.
Lahey, B. B., Mcnees, M. P., & Brown, C. C. Modification of deficits
in reading comprehension. Journal of Applied Behavior Analysis,
1973, 6(2), 475-480.
Lerner, J. W. Children with learning disabilities. Boston: Houqhton
Mifflin, 1976.
Lindsley, 0. R. Personal communication. Precision Teaching Conference,
Orlando, Florida, March 11 , 1983.
Lovitt, T. C. Personal communication. Precision Teaching Conference,
Orlando, Florida, March 1982.
Lovitt, T. C. In spite of my resistance . . . I've learned from
children. Columbus, OH: Charles E. Merrill, 1977.

133
Lovitt, T. C., & Hansen, C. L. Round one--placing the child in
the right reader. Journal of Learninq Disabilities, 1976,
9(6), 347-353.
Lovitt, T. C., & Hansen, C. L. The use of contingent skipping and
drilling to improve oral reading and comprehension. Journal
of Learninq Disabilities, 1976, £(8), 481-487.
McLeod, T., & Crump, D. The relationship of visuospatial skills and
verbal ability to learning disabilities in mathematics. Journal
of Learninq Disabilities, 1978, TJ_( 4), 237-241.
Mercer, C. D., Mercer, A., & Evans, S. The uses of frequency in
establishing instructional aims. Journal of Precision Teachinq,
1982, 3(3), 57-63.
Miller, G. A. Text comprehension skills and process model of test
comprehension. In H. Sager & R. B. Ruddel (Eds.), Theoretical
processes of reading. Neward, DE: International Reading
Association, 1976, 709-729.
Moore, R., & Goldiamond, I. Errorless establishment of visual dis¬
crimination using facing procedures. Journal of Experimental
Analysis of Behavior, 1964, 7_(3), 269-272.
National Association of School Psychologists. Special issue: Reading
assessment and intervention. School Psychology Review, 1982, 11_
O'Leary, K. D., Becker, W. C., Evans, M. B., & Saudergas, R. A.
A token reinforcement program in a public school: A replication
and systematic analysis. Journal of Applied Behavior Analysis,
1969, 2(1), 3-13.
Open Court Basic Reader: A trip around the world. Lasalle, IL: Open
Court Publishing, 1972.
Peaster, M. Y. Language and the reader in merritt. In E. John (Ed.),
New horizons in reading. Wilmington, DE: International Reading
Association, 1976.
Pennypacker, H. S. Personal communication. University of Florida,
October, 1982.
Pennypacker, H. S., Koenig, C. H., & Lindsley, 0. R. Handbook of
the standard behavior chart. Kansas City, KS: Precision Media,
797?:
Peterson, N. An introduction to verbal behavior. Minneapolis:
Behavior Associates, 1978.

134
Poulton, E. C., & Brown, C. H. Memory after reading aloud and
silently. British Journal of Psychology, 1967, 58(3-4), 219-222.
Rankin, E. F., Jr. The Cloze prccedure--A survey of research. In
E. L. Thurston & L. E. Hafner (Eds.), The philosophical and
sociological bases of reading. The 14th Yearbook of the National
Readinq Conference. Milwaukee: National Readinq Conference, 1965,
133-150.
Raygnor, A. L., Wark, D. M., & Warren, A. D. Operant conditioning of
readinq rate: The effect of a secondary reinforcer. Journal of
Readinq, 1966, 9(3), 147-T57.
Roberts, M., & Smith, D. D. The relationship among correct and error
oral reading rates and comprehension. Learning Disability
Quarterly, 1980, ¿(1), 54-64.
Ryan, E. B., & Semmel, M. L. Reading as a constructive language process.
Readinq Research Quarterly, 1969, 50), 59-84.
Rystrom, R. Toward defininq comprehension. The Readinq Teacher, 1970,
23(2), 144-158.
Samuels, S. J, The method of repeated readings. The Readinq Teacher,
1979, 32(4), 403-408.
Sidman, M. Tactics of scientific research. New York: Basic Books,
1960.
Sidman, M. Reading and auditory equivalences. Journal of Speech and
Hearing Research, 1971, J4(l), 5-7.
Sidman, M., & Stoddard, L. T. The effectiveness of fading in program-
the Experimental Analysis of
Behavior,
1967, 10(1), 3-15.
Skinner, B. F.
The behavior of organisms:
An experimental analysis.
New York:
Appleton-Century-
Crofts, 1938.
Skinner, B. F.
1957.
Verbal behavior.
Englewood
Cliffs, NJ: Prentice-Hall,
Skinner, B. F.
Science and human
behavior.
New York: Macmillan, 1953
Skinner, B. F.
Teaching machines
. Science.
, 1958, 128, 969-977.
Smith, F. Understanding readinq.
New York;
: Holt, Rinehart and
Winston, 1971.
Smith, N. B. The many faces of readinq comprehension. The Readinq
Teacher, 1969, 23(3), 249-259.

135
Smith, iN. B. The many faces of reading comprehension. In L. Harris
a C. Smith (Eds.), Individualized reading instruction: A
reader. New York: Holt, Rinehart and Winston, 1977.
Spache, G. D., & Spache, E. B. Reading in the elementary school (4th
ed.). Boston, MA: Allyn and Bacon, 1977.
Staats, A. W., & Staats, C. K. A comparison of the development of
speech and reading behavior with implications for research.
Child Development, 1962, 33_(4), 831-846.
Staats, A. W., Minke, K. A., Finley, J. R., Wolf, M. M., & Brooks,
L. 0. A reinforcer system and experimental procedure for the
laboratory study of readinq acquisition. Child Development,
1964, 35(1), 209-231.
Starlin, C. M-, & Starlin, A. Guides to decision making in oral
reading. Bemidji, MN: Unique Curriculums Unlimited, 1973.
Terrace, H. S. Discrimination learning with and without errors.
Journal of Experimental Analysis of Behavior, 1963, 60 ), 1-27.
Thirteenth annual gallop poll of the public's attitudes toward the
public schools. Phi Delta Kappan, 1980, 62(1 ), 33-48.
Thomsem, S. Precision teaching with speech and language. Paper
presented at the Precision Teaching Conference, Orlando, Florida,
March 11, 1983.
Thorndike, E. L. Reading as reasoning: A study of mistakes in
paragraph readinq. Journal of Educational Psycholoqy, 1917,
8(6), 323-332.
Touchette, P. E. The effects of gradual stimulus change on the
acquisition of simple discrimination in severely retarded boys.
Journal of the Experimental Analysis of Behavior, 1968, 11(1),
39-48.-
Turner, E. C., & Fey, F. T. Improving comprehension skills.
Gainesville, FL: Florida Educational Research Council, 1977.
Van Etten, G. V. A look at reading comprehension. Journal of Learning
Disabil ities, 1978, JJ_(1), 30-34.
Vogel, S. A. Syntactic abilities in normal and dyslexic children.
BaltimoreT University Park Press, 1975.
Vogel, S. A. Morphological ability in normal and dyslexic children.
Journal of Learning Disabilities, 1977, 10(1 ), 35-63.

136
Waters, L. G. School psychologists as perceived by school personnel:
Support for a consultant model. Journal of School Psychology,
1973, Jl(l), 40-46.
Weaver, W. W., Holmes, C. C., Curtis, J., & Reynolds, R. J. The
effects of reading variation and punctuation condition upon
reading comprehension. Journal of Readinq Behavior, 1970, 2(1),
75-84.
Whaley, D. L., & Malcott, R. W. Elementary principles of behavior.
New York: Appleton-Century-Crofts, 1971.
Wolf, M. M., Giles, D. K., & Hall, R. V. Experiments with token
reinforcements in a remedial classroom. Behavior Research and
Therapy, 1968, 60), 51-64.
Wolking, W. D. Rate of growth toward adult proficiency. Paper
presented at the International Symposium of Learning Disabilities,
Miami Beach, Florida, October, 1973. In C. D. Mercer (Ed.),
Children and adolescents with learning disabilities. Columbus,
OH: Charles E. Merrill, 1979.
Wolking, W. D. Personal communication. University of Florida,
Gainesville, Florida, September 1980.
Wolking, W. D. Personal communication. University of Florida,
Gainesville, Florida, July 1982.
Wolking, W. D., & Buss, M. A microcomputer program to teach precision
teaching technical terms and store learning histories. Paper
presented at the Precision Teaching Conference, Orlando, Florida,
March 11, 1983.
Wolking, W. D., & Greenwood, B. Teaching sight words: Using cross
model transfer. Unpublished manuscript, University of Florida,
Gainesville, Florida, 1979.
World Book Dictionary. Chicago, IL: Doubleday and Co., 1981.

APPENDIX A
FREQUENCY MULTIPLIERS,
TOTAL PERFORMANCE,
FREE RECALL

Subject B.S.
LRU I
URN I
LRNI
HRWI
ND
XI.71
ND
LRWI
—
XI. 71
ND
IIRNI
—
—
*1.71
Comprehension
LRWI
HRNI
LRNI
H'RWI
XI. 11
*1.29
*1.13
LRWI
—
*1.43
*1.25
HRNI
—
—
XI.14
Maintenance - 3 Days
LRWI
HRNI
LRNI
HRWI
ND
ND
X5.0
LRWI
—
ND
X5.0
HRNI
— •
—
X5.0
Maintenance - 10 Days

Subject S.B.
LRNI
HRWI
LRWI
HRWI
*1.44
X
o
00
*1.63
LRNI
—
XI. 56
*1.13
HRWI
—
—
*1.75
Comprehension
LRNI
HRWI
LRWI
LRNI
HRWI
LRWI
HRNI
*2
ND
*1.2
HRNI
*1.6
XI.88
*1.33
LRNI
-
X2
Í
LO
I ~
1
LRNI
—
X3.0
XI .2
HRWI
—
—
*1 .2
HRWI
—
—
*2.5
Maintenance - 3 Days
Maintenance - 10 Days

Subject E.H.
LRWI
HRNI
LRNI
HRWI
si.89
si .21
•si.55
LRNI
—
XI .56
XI .22
HRNI
—
-
sl.27
Comprehension
LRWI
HRNI
LRNI
HRUI
si.27
XI. 29
si.27
LRWI
—
XI. 64
ND
HRNI
—
-
si.64
Maintenance - 3 Days
LRWI
HRNI
LRNI
HRWI
si.07
si.6
il .14
LRWI
—
si.5
si .07
HRNI
—
—
XI.4
Maintenance - 10 Days

Subject A.P.
LRNI
HRWI
LRWI
URN I
*2.7
*1.07
*1.14
LRNI
—
X2.5
X2.0
HRWI
—
—
*1.25
Comprehension
LRNI
HRWI
LRWI
URN I
*1.4
X2.86
*1.4
LRNI
—
X4.0
ND
HRWI
—
—
*4.0
Maintenance - 3 Days
LRNI
HRWI
LRWI
HRNI
*1.88
*1.67
*1 .5
LRNI
—
XI. 13
XI. 25
HRWI
—
—
XI .11
Maintenance - 10 Days

Subject J.J
LRNI
HRWI
LRU I
URN I
*2.14
XI. 13
*1.5
LRNI
—
X2.43
XI.43
MRU I
—
—
*1.7
Comprehension
LRNI
HRWI
LRU I
LRNI
HRWI
LRWI
HRNI
*2.0
XI. 07
*1.2
MRNI
*1.5
*1.2
*1.5
LRNI
—
X2.14
XI.14
LRNI
—
XI.25
ND
HRUI
—
—
*1.88
HRWI
—
—
*1.25
Maintenance - 3 Days
Maintenance - 10 Days

Subject D.D.
LRNI
HRWI
LRWI
URN I
â– si. 25
XI.4
41.15
LRNI
—
XI .75
XI.08
HRWI
—
—
41.62
Comprehension
LRNI
HRWI
LRWI
LRNI
HRWI
LRWI
URN I
â– si .21
XI. 24
41.30
HRNI
42.0
XI. 12
XI. 08
LRNI
—
XI. 5
41.08
LRNI
—
X2.23
X2.08
HRWI
—
—
41.62
HRWI
—
—
41.07
Maintenance - 3 Days Maintenance - 10 Days

APPENDIX B
ACCURACY RATIO MULTIPLIERS
FREE RECALL

Subject D.D
LRNI
HRWI
LRWI
HRNI
â– il. 3
X3.08
X2.0
LRNI
—
X2.0
X2.6
HRWI
—
—
il .54
Comprehension
LRNI
HRWI
LRWI
LRNI
HRWI
LRWI
HRNI
XI .13
X6.46
X3.69
HRNI
*2.17
XI .12
-=3.25
LRNI
—
X5.74
X3.28
LRNI
—
X2.42
il .5
HRWI
—
—
â– si .75
HRWI
—
—
i 3.63
Maintenance - 3 Days Maintenance - 10 Days

Subject J.J.
LRNI
HRWI
LRWI
HRNI
*2.5
XI.13
*3.75
LRNI
—
X2.83
-1.5
HRWI
—
—
■¡4.25
Comprehension
LRNI
HRWI
LRWI
LRNI
HRWI
LRWI
HRNI
*2.17
T1
o
CO j
*1.86
HRNI
XI.4
X2.0
XI .4
LRNI
—
X2.33
XI.17
LRNI
—
XI .43
ND
HRWI
—
—
*2.0
HRWI
—
—
*1.43
Maintenance - 3 Days
Maintenance - 10 Days

Subject S.B.
HRNI
LRNI
HRWI
LRWI
HRNI
—
*1.44
XI. 08
*1 .86
LRNI
—•
—
XI.56
*1.29
HRWI
—
—
—
*2.0
LRWI
—
—
—
—
Comprehension
HRHI LRNI HRWI LRWI HRNI LRNI HRWI LRWI
HRNI
—
*2.4
ND
*3.0
HRNI
—
*1.6
XI.88
*8.0
LRNI
—
—
X2.4
*1.25
LRNI
—
—
X3.0
*5.0
HRWI
—
—
—
*3.0
HRWI
—
.
—
*15.0
LRWI
—
—
—
—
LRWI
—
—
—
Maintenance - 3 Days
Maintenance - 10 Days

Subject E.H.
HRWI
LRWI
HRNI
LRNI
HRWI
—
•si .89
*1.21
*3.78
LRWI
—
—
XI.56
*2.0
HRNI
—
—
—
*3.11
LRNI
—
—
—
—
Comprehension
HIM LRWI HRNI LRNI HRWI LRWI HRNI LRNI
HRWI
—
*1.4
XI.29
*8.0
HRWI
—
*1.08
XI .43
*1.17
LRWI
—
—
XI.8
*5.71
LRWI
—
—
XI. 67
*1.08
HRNI
—
—
—
*10.3
HRNI
—
—
—
*1.67
LRNI
—
—
—
—
LRNI
—
—
—
—
Maintenance - 3 Days
Maintenance - 10 Days
•t»
03

Subject A.P.
URN I
LRNI
HRWI
LRWI
URN I
—
+ 3.2
+ 1 .07
+ 1.33
LRNI
—
—
X3.0
X2.4
HRWI
—
—
—
+ 1 .25
LRWI
—
—
—
—
Comprehension
HRNI
LRNI
HRWI
LRWI
HRNI
—
+ 1.75
X2.71
+ 1 .4
LRNI
—
—
X4.75
XI.25
HRWI
■—
—
—
-=3.8
LRWI
—
—
—
—
Maintenance - 3 Days
HRNI
LRNI
HRWI
LRWI
HRNI
—
XI.08
+ 1.38
+ 1.54
LRNI
—
—
XI .29
XI .43
HRWI
—
—
—
XI .11
LRWI
—
—
—
—
Maintenance - 10 Days
-F»

Subject B.S.
HRWI
LRWI
HRMI
LRNI
HRWI
_
ND
XI .71
ND
LRWI
*
XI.71
ND
HRMI
—
—
—
il .71
LRNI
—
—
—
Comprehension
HRWI
LRWI
HRNI
LRNI
HRWI
—
XI .11
-1.29
-il. 13
LRWI
—
—
•il .43
-il .25
HRNI
—
—
—
XI .14
LRNI
—
—
—
—
Maintenance - 3 Days
HRWI
LRWI
HRNI
LRNI
HRWI
—
ND
ND
X4.5
LRWI
—
—
ND
X4.5
HRNI
—
—
—
X4.5
LRNI
—
—
—
—
Maintenance - 10 Days

APPENDIX C
FREQUENCY MULTIPLIERS,
TOTAL PERFORMANCE,
CLOZE PROCEDURE

Subject B.S.
LRWI
HRNI
LRNI
HRWI
XI.3
XI .53
XI.5
LRWI
—
XI.17
XI. 15
HRNI
—
—
â– ;! .02
Comprehension
LRWI
HRNI
LRNI
LRWI
HRNI
LRNI
HRWI
XI.41
XI .33
XI.44
HRWI
â– si.07
â– sl .30
*1 .2
LRWI
—
*1 .06
XI. 02
LRWI
—
CXI
M*
■íl. 12
I1RNI
—
—
XI. 08
HRNI
-
—
XI.09
Maintenance - 3 Days
Maintenance - 10 Days

Subject A.P.
LRNI
HRHI
LRU I
HRHI
X2.25
XI. 38
XI.49
LRNI
—
-il .63
■íl.51
II Rill
—
—
XI. 08
Comprehension
LRNI
HRHI
LRWI
LRNI
HRWI
LRWI
HRNI
XI.79
XI. 05
XI.49
HRNI
il.76
il .63
il .67
LRNI
—
â– il .70
â– il. 98
LRNI
—
XI .08
XI.06
HRHI
—
—
â– il. 17
HRWI
—
il .03
Maintenance - 3 Days
Maintenance - 10 Days

Subject E.H.
LRWI
HRNI
LRNI
MRU I
*1.37
*1.30
*1.27
LRWI
—
XI. 06
XI. 08
URN I
—
—
XI. 02
Comprehension
LRWI
HRNI
LRNI
HRWI
*1.17
CO
i n
*
X
XI. 21
LRWI
—
XI. 61
XI.42
HRNI
—
—
*1.14
Maintenance - 3 Days
LRWI
HRNI
LRNI
HRWI
*1.81
*1.13
*1.02
LRWI
—
XI .5
XI .67
HRNI
—
—
XI .11
Maintenance - 10 Days

Subject S.B
LRNI
I1RW1
LRWI
HRNI
«1.48
«1.04
«1.18
LRNI
—
XI.43
XI. 25
HRUI
—
—
«1.14
Comprehension
LRNI
HRWI
LRWI
LRNI
HRWI
LRWI
HRNI
«1.28
XI.72
XI.12
HRNI
XI. 37
XI. 57
XI. 34
LRNI
—
X2.21
XI.44
LRNI
XI. 15
«1 .02
HRWI
—
—
«1.54
HRWI
—
—
«1.17
Maintenance - 3 Days
Maintenance - 10 Days

APPENDIX D
ACCURACY RATIO MULTIPLIERS
CLOZE PROCEDURE

Subject B.S.
HRWI
LRWI
HRMI
LRNI
HRWI
—
■¡3.3
*1.97
*2.53
LRU I
—
—
■¡1.38
*1.3
1IRNI
—
—
—
XI. 06
LRNI
—
—
—
—
Comprehension
HRWI
LRWI
HRNI
LRNI
HRWI
LRWI
HRNI
LRNI
HRWI
—
-¡3.04
■¡4.78
-¡2.49
HRWI
—
*1.11
XI. 19
*1.26
LRWI
—
—
X2.13
XI .22
LRWI
—
—
XI .30
*1.15
MRNI
—
—
—
*1.75
IIRNI
—
—
—
*1 .50
LRNI
—
—
—
—
LRNI
—
—
—
—
Maintenance - 3 Days
Maintenance - 10 Days

Subject E.H.
HRWI
LRWI
HRNI
LRNI
HRWI
—
+ 1.46
+1 .04
+ 2.05
LRWI
—
—
XI.4
-fl .4
HRN I
—
—
—
-il .96
LRNI
—
—
—
—
Comprehension
HRWI LRWI HRNI LRNI HRWI LRWI URN I LRNI
HRWI
—
-¡4
+9.74
+4.29
HRWI
—
XI.76
+ 1.89
+ 1.65
LRWI
—
—
*2.44
+ 1.07
LRWI
—
—
+ 3.37
+ 2.91
HRNI
—
—
—
11.11
HRNI
—
—
—
XI .16
LRNI
--
—
—
—
LRNI
—
—
—
—
Maintenance - 3 Days
Maintenance - 10 Days
CJl
Co

Subject A.P.
HRNI
LRNI
HRWI
LRWI
HRNI
—
-¡3.57
CO
><
■¡2.36
LRNI
—
—
X4.22
XI .51
HRWI
—
—
—
â– si.13
LRWI
—
—
—
—
Comprehension
HRNI LRNI tIRWI LRWI URN I LRNI HRWI LRWI
HRNI
—
■¡6.40
■¡1.4
■¡1.4
HRNI
—
X2.63
X3.42
XI. 66
LRNI
—
—
X4.57
X4.57
LRNI
—
—
XI. 30
•¡1.59
HRWI
—
—
—
NC
HRWI
—
—
—
-¡2.06
LRWI
—
—
—
—
LRWI
—
—
—
—
Maintenance - 3 Days Maintenance - 10 Days

Subject S.B.
URN I
LRNI
HRWI
LRWI
URN I
—
XI. 05
X19.6
â– i 1.34
LRNI
—
—
X18.67
-il .43
HRWI
—
—
—
â– f26.29
LRWI
—
—
—
—
Comprehension
IIRNI
LRNI
HRWI
LRWI
HRNI
—
X9.5
X3.33
il .33
LRNI
—
—
â– f 2.86
fl 2.67
HRWI
—
—
—
i4.43
LRWI
—
—
—
—
Maintenance - 3 Days
HRNI
LRNI
HRWI
LRWI
HRNI
—
XI .79
X3.6
il.66
LRNI
—
—
X2.01
i2.97
HRWI
—
—
—
i5.97
LRWI
—
—
—
—
Maintenance - 10 Days

APPENDIX E
ID SAMPLE COMPREHENSION QUESTIONS
CLOZE PROCEDURE
IQ Comprehension Questions from Antarctica
Subject S.B.
1. How was Admiral Byrd able to keep his place in the air long enough
to fly over the South Pole?
Latency:
2. After the food was dropped, what did the pilot shout again?
Latency:
3. How was Admiral Byrd's plane different from the ones we have now?
Latency:
4. Admiral Byrd became the first man to do what?
Latency:
5. Why don't people like to live in Antarctica?
Latency:
6. What kind of place is Antarctica?
Latency:
7. Why did Admiral Byrd and his men stay in a place called Little
Ameri ca?
Latency:
8. How have other brave men explored the South Pole?
Latency:
9. How are penguins different from other birds?
Latency:
10.What other animals live with or near the penguins to keep them
company?
Latency:
161

162
Cloze Procedure
Phase C
Subject S.B.
Antartica
"We must either drop the gasoline or go
!" the pilot shouted. But Byrd did
not want drop the gasoline from
airplane because then they not reach the South
. Instead he shouted back, " a bag of
food !"
The food was dropped,
his men to
soon the pilot shouted
, "Quick! Dump more!" Admiral ordered
the
trap door. More
another bag of food
a month's supply of
food had been thrown
On their way to
his men had to
South Pole, Admiral Byrd
long distances and
over
mountains. Admiral Byrd's plane
slow, and it could fly as far or
fast
as airplanes can . Even though the weather
bitter cold and he few supplies,
Admiral Byrd
the first man to
over
the South Pole.

163
land around the South
is called
Antarctica. For
mean stayed in a
whole winter Byrd and
out more about the
called Little America to
of the world.
Antarctica
It is much
the coldest and windiest
loneliest continent in the world.
than the United States,
not a single person
there because it is
cold. Most of Antarctica
covered with ice which
melts. In some places
thick.
time and
all.
Like
the summer the sun
the winter it does
is a mile
al 1 the
shine at
Byrd, other brave men
al so
explored this land
ice and snow. With
sleds or snowmobiles they their way to the
Pole by riding across snow. But
everyone who
it is time to
Antarctica is always glad
home.
It is not
right to say that
one
lives in Antarctica penguins live there. The
make their home in ice and snow and
the shore. There
, and there
shore,
are
are a
the fish that swim
_ whales in the nearby
but otherwise the
insects and plants near _
do not have much

164
Penguins are the funniest
you have ever
seen.
all they can
ice. But in _____
and divers, and so
cannot fly, and they
is waddle and slide
water they are expert
can find plenty of
run;
the
to eat.

APPENDIX F
POST-EXPERIMENTAL INTERVIEWS
Experiment 1
Subject D.D.
E: How did you feel about doing this research with me?
S: It's been alright. I liked it.
E: What did you like about it?
S: Well, I don’t really know what I like about it, but it helped
me a little bit.
E: Did it? In what way?
S: Well, I got faith. It gave me faith, so that's where it helped
me. You need faith.
E: Yeah? What did it give you faith in?
S: In reading.
E: Did it do that?
S: Yeah.
E: Did you like the way we were doing the reading?
S: Yeah, I liked the way we did it.
E: Anything in particular that you liked about it?
S: Well, I don't know. I like being here every day and all that.
When I was off work detail.
E: What about the times I got you to read fast?
165

166
S: Oh, it was alright. Sometimes it got hard. But it wasn't
really hard. It was challening.
E: What did you like doing better-reading fast or reading slow?
S: Fast. I didn't like reading that slow, 'cause I couldn't do it.
You kept telling me to slow down, slow down.
E: Alright, if I didn't tell you to slow down which one did you like
better?
S: Reading fast.
E: Reading fast. How come?
S: I just like reading fast. Makes more sense when you read fast.
E: Good point. OK, Thank you, David.
Subject J.J.
E: How did you like doing all this?
S: It was alright, man. You know, at first I had a little fit about
it, but that was before you even came. When you asked me about it,
I didn't want to do it, but I said "what the hell." That's the way
I feel. I said that I'm going to give it a try.
E: So you gave it a try and what happened?
S: So, I gave it a try. I came in the first day and it was down.*
You asked me to start reading first. And I did begin and it came
alright. It was alright.
E: So did you like coming?
S: Yeah it was alright. I wouldn't mind keeping it up.
E: What did you like about it?
S: Everything, man. The whole thing’, I liked the whole thing.
E: Would you do it again?
S: Yeah.
E: Yeah, alright.
*Slang for it was good.

167
S: I'd keep on doing it now, you know. It wouldn't hurt nothing.
E: Yeah, you'd like that. When we were doing the readings, the fast
and the slow readings, which did you like doing better?
S: Well, I like the fast reading, but, you know, I like the fast
reading because it could come clearer to me more quickly and I
could remember what the story was about.
E: So, you said when you read fast it comes to you. Well, what comes
to you?
S: What I'm reading about. It's like when I'm reading fast I can read
so much and then I can kind of put the rest of it before I even
read it. I can do it when I'm reading slow, but I can do it
quicker when I'm reading fast.
E: So, you like reading fast?
S: Yeah.
E: Well, that's good. Now, there was also a time when I was telling
you to read without paying attention to the marks and then I asked
you to read with paying attention to the marks. Did that make any
difference to you?
S: It made a little difference. Well, it makes a big difference, really,
because when I was reading without the punctuation marks, I was just
shooting through. It was like you be running or something and just
go ahead on around. If you run a full mile, every time you run a
mile, you take a stop and think about what you're doing.
E: So, you're saying you liked it better with the punctuation marks?
S: Yeah.
E: Because you can think about what you're reading?
S: Yeah, you get an idea what he said and what someone else, what
everybody said.
E: That made it easier?
S: Yeah, it didn't make it easier but it makes you understand what
it's all about.
E: So reading fast with punctuation was the best for you.
S: Yeah.
E: 0.K., Thank you.

168
Experiment 2
Subject S.B.
E: Did you like coming here?
S: Yes.
E: You did? What did you like about it?
S: I liked to read.
E: Oh. Well, of the stories we read here, which one did you like most?
S: The General and the Arrows.*
E: You did. You liked that one? O.K. Do you remember when we were
reading and sometimes I'd asked you to read slow and sometimes I
asked you to read fast? Which did you like better?
S: Fast.
E: Fast? Why?
S: Because it was easier.
E: O.K. Now, remember I asked you to read fast and slow and sometimes
I asked you to not pay attention to the marks, then sometimes I
asked you to pay attention to them. What did you like better that
time?
S: Not to pay attention to them.
E: Not to pay attention to them? How come?
S: Because it was easier.
E: It was easier? You could just read without them? O.K. Now
which stories did you understand the most--reading fast stories
or reading slow stories?
S: (No response.)
E: Do you know what I mean by understand? Well, in which stories did
you know what was going on the best?
S: Urn. The General and the Arrows and Antarctica.*
E: Antarctica. O.K. Well, thank you very much. I'm glad that I got
to meet you.
*Both stories were high rate stories. The General and the Arrows was
a high rate with inflection story.

159
Subject E.H.
E: How did you like coming here?
S: I 1 iked it.
E: You did? What did you like about it?
S: Reading fast.
E: Well, what made you say that? Why reading fast, why do you
like reading fast?
S: Because whenever you read slow, whenever you tell me just to read
slow and then say don't stop at the periods or anything, I can't
help it, 'cause whenever you read slow it sounds like you have a
period every time.
E: O.K., but do you remember one time when I asked you to read slow
and pay attention to the periods? Which time did you like better¬
reading fast and paying attention or reading slow?
S: Fast.
E: Reading fast. Why?
S: I don't know.
E: Well, what did you like about reading fast?
S: I got through a story faster.
E: O.K., when you read faster, did you know what was going on in the
story?
S: (Nods.)
E: You did? That's good. Which was easier for you to understand--
reading fast or reading slow?
S: Fast.
E: Why do you suppose that is? You don't know? O.K. So, you like
reading that's good. Would you like to do this again?
S: (Nods.)
E: You would? O.K. I liked working with you. You're a very hard
worker. Thank you, Erica.

170
Subject B.S.
E: How did you like coming here?
S: I liked it.
E: You did like it?
S: I like comprehension.
E: Comprehension? What do you mean? What part of comprehension?
S: Just writing.*
E: Just reading and writing? You like writing those things. You did?
Oh, O.K. Do you remember what story you like the most?
S: The General and the Arrows.**
E: You did? O.K., now, when you were reading, a couple times I asked
you to read fast and then I asked you to read slow. Remember that?
What did you like better-reading fast or slow?
S: Fast.
E: You did. Why?
S: Because I can be done quicker.
E: O.K., that's one reason. When you read fast or slow, which way do
you think helped you understand the story better?
S: Slow.
E: O.K. And remember a few times we were reading and I asked you to
sometimes pay attention to the punctuation marks and sometimes not.
Which did you like better when I asked you to do that?
S: Pay attention to them.
E: That helped you understand things better? O.K., so you like reading
fast except reading slow, you thought, helped you understand better?
S: Yeah.
E: But you like reading fast. Did you understand the things you read
when you were reading fast?
* Subject with mild articulation difficulties.
**High rate story.

171
S: (Nods.)
E: You did? Well, which do you think you understood better-fast
or slow?
S: Slow.
E: Well, good. You know, I liked when you came here. It was fun.
Thank you very much.
Subject A.P.
E: I was curious about how you felt about everything that went on.
How did you like doing all of this?
S: Iliked it.
E: You did like it? What did you like about it?
S: Well, I liked the reading stories the best.
E: You did? Well, O.K., let me ask you about the stories. What story
did you 1ike the best?
S: The first story.
£: What story was that?
S: The Trip Around the World.*
E: Oh, you liked that, huh? You know, during the readings sometimes
I told you to read fast and sometimes I asked you to read slow.
Which did you like better?
S: Fast.
E: And why did you like reading fast better?
S: Because it's easier.
E: Remember, when I asked you to sometimes pay attention to the marks?
Remember the marks? And sometimes I said don't pay attention to
them. Which did you like better doing that way?
S: Not paying attention.
E: You didn't like paying attention to them?
High rate story, no inflection.

172
S: Well, I sort of like both but . . .
E: Well, when you were reading fast what did you like to do--
paying attention to the marks or not paving attention?
S: I don't know. I liked it either way, but reading slow I liked
not paying attention.
E: You did? O.K., well that's a good answer. Well, thank you, Amy.
I've really enjoyed working with you. You're a very nice person
to get to know. You're a good student.

APPENDIX G
RAW DATA

Subject B.S
Cloze Procedure--^ Correct
HRNI-1
LRNI-1
HRWI-1
LRWI-1
7T
72
87
77
HRNI-3
L RN1-3
H RW1-3
L RW1-3
83
73
89
69
HRNI-10
LRNI-10
HRWI-10
LRWI-10
83
77
80
78
Cloze
Procedure—Words/Min.
and Accuracy Ratio
HRNI-1
AR
LRNI-1
AR
HRWI-1
AR
LRWI-1
M
4.'3/1.8
2.39
4.3/1.7
2.53
3.3/ .7
4.71
4 /1.2
3.33
HRNI-3
LRNI-3
HRWI-3
L RW1-3
4.3/ .9
4.73
4.1/1.5
2.73
3.4/ .5
6.8
3.8/1.7
2.24
HRNI-10
LRNI-10
HRWI-10
LRWI-10
3.8/ .8
4.75
3.8/1.2
3.17
4.8/1.2
4
4.4/1.2
3.67
1 = Immediate measure
3=3 days after immediate measure
10 = 10 days after immediate measure
174

175
Free
Recal 1--Facts/Min.
and Accuracy
Ratio
HRNI-1
AR
L RN I -1
AR
HRWI-1
AR
LRWI-1
AR
12/0
12
7/0
7
7/0
7
7
7
H RN1-3
L RN1-3
H RW1-3
L RW 1-3
7/0
8
8/0
8
9
9
10
10
H RN 1-10
LRNI-10
HRWI-10
LRWI-10
2/0
2
9/1
9
2
2
8/4
2
Free Recall-
-Word/Fact
HRNI-1
LRNI-1
HRWI-1
LRWI-1
32
53
73
77
HRNI-3
L RN1-3
HRWI-3
L RW1-3
55
47
57
39
HRNI-10
LRNI-10
HRWI-10
LRWI-1 0
193
41
256
67

176
10 Comprehension Questions-
-Correct Answers
HRNI-1
LRNI-1
HRWI-1
LRWI-1
7/3
5/5
8/2
6/4
H RN1-3
L RN1-3
HRWI-3
L RW1-3
8/2
6/4
7/3
6/4
HRNI-10
LRNI-10
HRWI-10
LRWI-10
7/3
6/4
7/3
5/5
10 Comprehension Questions-
-Median Latency
HRNI-1
LRNI-1
HRWI-1
LRWI-1
2.05
1.65
T.3
1.5
H RN1-3
LRNI-3
HRWI-3
LRWI-3
1.05
1.15
1
1.3
HRNI-10
LRNI 10
HRWI-10
LRWI-10
1.5
.75
1.05
1.8

177
Subject J.J.
Free Recaí 1--Facts/Min.
HRNI-1
LRNI-1
HRWI-1
LRWI-1
15/0
6/1
17/0
8/2
HRNI-3
LRNI-3
H RW1-3
LRWI-3
13/1
6/1
14/1
7/1
HRNI-1a
LRNI-10
HRWI-10
LRWI-10
10/2
7/1
10/0
7/1
HRNI-1
Words/Fact
LRNI-1
HRWI-1
LRWI-1
63.4
137
53.1
101 .8
HRNI-3
LRNI-3
H RW1-3
LRWI-3
73
137
64.5
116.3
HRNI-10
LRNI-10
HRWI-10
LRWI-10
95.1
117
90.4
116.3

178
10 Comprehension Questions--Correct Answers
HRNI-1
LRNI-1
HRWI-1
LRWI-1
10/0
7/3
10/0
8/2
HRNI-3
L RN1-3
H RW1-3
LRWI-3
10/0
8/2
8/2
7/3
HRNI-10
LRNI-10
HRWI-10
LRWI-10
9/1
9/1
10/0
7/3
Median Latencies--
10 Comprehension Questions
HRNI-1
LRNI-1
HRWI-1
LRWI-1
.85
1.55
.7
1.1
HRNI-3
LRNI-3
HRWI-3
LRWI-3
.85
1.44
.55
1.55
HRNI-10
LRNI-10
HRWI-10
LRWI-10
.65
.65
.6
2.3
*HRNI-1, 3, 10 = immediately following criteria, 3 days after criteria,
10 days after criteria.

179
Subject D.D.
10 Comprehension Questions-
-Correct Answers
HRNI-1
LRNI-1
HRWI-1
LRWI-1
9/1
8/2
9/1
8/2
H RN1-3
LRNI-3
H RW1-3
L RW1-3
8/2
5/5
10/0
8/2
HRNI-10
LRNI-10
HRWI-10
LRWI-10
8/2
7/3
10/0
9/1
10 Comprehension Questions-
-Median Latency
HRNI-1
LRNI-1
HRWI-1
LRWI-1
.5
T i 25
.65
1.1
H RN1-3
LRNI-3
HRWI-3
L RW1-3
.5
1.25
.6
1 .2
HRNI-10
LRNI-10
HRWI-10
LRWI-10
.45
.5
.65
.3

180
Free Recall-
-Facts/Minute
HRNI-1
LRNI-1
HRWI-1
LRWI-1
13/2
10/2
20/1
13/0
HRNI-3
LRNI-3
H RW1-3
L RW 1-3
13/4
11/3
21/0
12/1
HRNI-10
LRNI-10
HRWI-10
LRWI-10
26/0
12/1
29/0
24/3
Free Recall-
-Words/Fact
HRNI-1
LRNI-1
HRWI-1
LRWI-1
25/3
26.0
12.7
13.5
HRNI-3
LRNI-3
HRWI-3
L RW 1-3
25.'3
23.6
10.2
14.6
HRNI-10
LRNI-10
HRWI-10
LRWI-10
12.6
22.0
7.45
7.3

181
Subject S.B.
Free
Recall--Facts/Min.
and Accuracy
Ratio
HRNI-1
AR
LRNI-1
AR
HRWI-1
AR
LRWI-1
M
13/0
13
9/0
9
14/0
14
7/1
7
H RN1-3
LRNI-3
HRWI-3
LRWI-3
12/0
12
5/1
5
12/0
12
8/2
4
HRNI-10
LRNI-10
HRWI-10
LRWI-10
8/0
8
5/0
5
15/0
15
3/3
1
Free Recall-
-Word/Fact
HRNI-1
LRNI-1
HRWI-1
LRWI-1
31
41
29
77
HRNI-3
LRNI-3
HRWI-3
LRWI-3
33
79
34
67
HRNI-10
LRNI-10
HRWI-10
LRWI-10
50
75
27
179

182
10 Comprehension Questions-
-Correct Answers
HRNI-1
LRNI-1
HRWI-1
LRWI-1
8/2
6/4
10/0
8/2
HRNI-3
LRNI-3
HRWI-3
L RW1-3
8/2
7/3
10/0
7/3
HRNI-10
LRNI-10
HRWI-10
LRWI-10
7/3
9/1
9/1
4/6
10 Comprehension Questions-
-Median Latency
HRNI-1
LRNI-1
HRWI-1
LRWI-1
1.2
1.9
.8
2.05
HRNI-3
LRNI-3
H RW1-3
LRWI-3
1.15
1.15
.8
1.45
HRNI-10
LRNI-10
HRWI-10
LRWI-1C
.75 .75 1.15
.75

183
Cloze Procedure—%
Correct
HRNI-1
LRNI-1
HRWI-1
LRWI-1
79
79
98
68
HRNI-3
LRNI-3
HRWI-3
LRWI-3
81
97
93
75
HRNI-10
LRNI-10
HRWI-10
LRWI-10
83
90
95
74
Cloze Procedure--Words/Min. and Accuracy Ratio
HRNI-1
AR
LRNI-1
AR
HRWI-1
AR
LRWI-1
AR
5.7/2
2.85
3 /I
3
5.6/.1
56
3.4/1.6
2.13
HRNI-3
LRNI-3
HRWI-3
LRWI-3
4 /I
4
3.8/.1
38
8 /. 6
13.3
4.2/1.4
3
HRNI-10
LRNI-10
HRWI-10
LRWI-10
2.9/.6
4.83
4.3/.5
8.6
5.2/.3
17.3
3.5/1.2
2.92

184
Subject E.H.
Free
Recaí1--Facts/Min.
and Accuracy Ratio
HRNI-1
M
LRNI-1
AR
HRWI-1
AR
LRWI-1
AR
14
14
9/2
4.5
17
17
9
9
H RN1-3
L RN 1-3
HRWI-3
L RW1-3
18
18
7/4
1 .75
14
14
10/1
10
HRNI-10
LRNI-10
HRWI-10
LRUI-10
10
10
12/2
6
14/2
7
13/2
6.5
Free Recall
--Word/Fact
HRNI-1
LRNI-1
HRWI-1
LRWI-1
27
44
32
46
H RN1-3
LRNI-3
HRWI-3
LRWI-3
21
57
38
41
HRNI-10
LRNI-10
HRWI-10
LRWI-10
37
33
38
32

185
10 Comprehension
Questions—Correct Answers
H RM1-1
LRNI-1
HRWI-1
LRWI-1
5/5
6/4
8/2
6/4
HRN1-3
LRNI-3
HRWI-3
LRWI-3
7/3
5/5
8/2
8/2
H RNI -10
LRNI-10
HRWI-10
LRWI-10
6/4
7/3
8/2
8/2
10 Comprehension Questions-
-Median Latency
HRNI-1
LRNI-1
HRWI-1
LRWI-1
3.9
1.65
1.15
1.5
H RN 1-3
LRNI-3
HRWI-3
LRWI-3
1.3
1.35
1.25
1 .45
HRNI-10
LRNI-10
HRWI-10
LRWI-10
1.85
1.5
1.25
1.35

186
Cloze Procedure--" Correct
HRNI-1
LRNI-1
HRWI-1
LRWI-1
85
75
86
82
H RN1-3
L RN1-3
H RW1-3
LRWI-3
66
82
95
80
HRNI-10
LRNI-10
HRWI-10
L RW1-10
70
74
82
83
Cloze
Procedure--Words/Min.
and Accuracy
Ratio
HRNI-1
AR
LRNI-1
AR
HRWI-1
AR
LRWI-1
AR
4.'6/.8
5.75
4.1/1.4
2.93
6/1
6
4.1/1
4.1
HRNI 3
L RN1-3
H RW1-3
LRWI-3
3.9/1.9
2.05
4.2/.9
4.67
4/.2
20
3/.6
5
HRNI-10
LRNI-10
HRWI-10
LRWI-10
3.8/1.6
2.38
4.4/1.6
2.75
5/1.1
4.55
3.2/.4
8

187
Subject A.P.
Cloze Procedure--"
; Correct
HRNI-1
LRNI-1
HRWI-1
LRWI-1
81
55
77
65
HRNI-3
LRNI-3
H RW1-3
L RW1-3
86
49
82
82
HRNI-10
LRNI-10
HRWI-10
LRWI-10
55
76
81
67
Cloze Procedure--Words/Min. and Accuracy Ratio
HRNI-1
AR
LRNI-1 AR
HRWI-1
AR
LRWI-1
AR
4.3/1
4.3
6.5/5.4 1.20
6.1/1.2
5.08
5.1/2.8
1 .82
HRNI-3
LRNI-3
HRWI-3
LRWI-3
6.3/1
6.3
6.5/6.6 10.98
6.3/1.4
4.5
5.4/1.2
4.5
HRNI-10
LRNI-10
HRWI-10
LRWI-10
7/5.7
1.28
5.5/1.7 3.24
6.3/1.5
4.2
5.1/2.5
2.04

188
Free Recal1--Facts/Min. and Accuracy Ratio
HRNI-1
AR
LRNI-1
AR
HRWI-1
ar
LRWI-1
AR
16/0
16
5/1
5
15/0
15
12/0
12
HRNI-3
L RN1-3
HRWI-3
LRWI-3
7/10
7
4/1
4
19/1
19
5/0
5
HRNI-10
LRNI-10
HRWI-10
LRWI-10
13/2
6.5
7/1
7
9/10
9
10/0
10
Free Recall-
-Word/Fact
HRNI-1
LRNI-1
HRWI-1
LRWI-1
24
77
26
34
HRNI-3
L RN1-3
HRWI-3
LRWI-3
55
97
21
82
HRNI-10
LRNI-10
HRWI-10
LRWI-10
30
55
44
41

139
10 Comprehension
Questions--Correct Answers
H RN1-1
LRNI-1
HRWI-1
LRWI-1
8/2
8/2
9/1
5/5
HRNI-3
LRNI-3
HRWI-3
LRWI-3
8/2
5/5
8/2
6/4
HRNI-10
LRNI-10
HRWI-10
LRWI-10
8/2
6/4
7/3
7/3
10 Comprehension Questions-
-Median Latency
HRNI-1
LRNI-1
HRWI-1
LRWI-1
2.85
1.95
.6
1.85
HRNI-3
LRNI-3
HRWI-3
LRWI-3
.75
2.35
.65
1.2
HRNI-10
LRNI-10
HRWI-10
LRWI-10
.5
1.35
.35
.85

BIOGRAPHICAL SKETCH
Henry Abraham Tenenbaum was born on February 28, 1951, in the
Brownsville section of Brooklyn, New York. His parents immigrated
from Poland after the Nazi occupation in 1949.
In 1965, he moved to Oceanside, Long Island, from Sheepshead Bay,
Brooklyn. In 1969, he graduated from Oceanside High School with an
academic diploma. After high school Henry went to Israel and worked
on a Kibbutz. Upon his return he entered Nassau Community College.
While studying for his Associates of Arts degree, he worked with
adolescents as a drug rehabilitation counselor. After studying for
1 1/2 years, he was awarded an Associates of Arts degree in liberal
arts in August of 1974. Henry then began his junior year at the
S.U.N.Y. at Stonybrook. His declared major was experimental psychology.
After a year of undergraduate study at Stonybrook, Henry moved
to North Miami Beach, Florida. He then completed his Bachelor of
Arts degree with honors in experimental psychology in 1976 from
Florida Atlantic University in Boca Raton, Florida.
In the fall of 1976, Henry began his graduate work in school
psychology at Florida International University, Miami, Florida.
During the time he was studying for his master's degree he was employed
at the Bertha Abess Children's Center. He worked full time as a
190

191
diagnostic teacher. The children at the center taught him how
behavioral principles can be used to teach children to gain the
most out of their educational experiences.
In the spring of 1978, Henry graduated with his Master of
Science degree in school psychology. After another year at the
Sertha Abess Children's Center, fie decided to obtain his doctorate.
He entered the University of Florida in the fall of 1979. In
August of 1982, Hentry was married to Sarah Margaret Sawyer. He and
his wife both are pursuing their careers in Gainesville, Florida.

I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
William D. Wolking, Chairperson
Professor of Counselor ^Education
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly ^presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
ry st Penny ^
Professor of Psychology
lyfi^cker
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
Bert L. Sharp «
Professor/of Counselor Education
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
c
Edward C. Turner
Assistant Professor of
Instructional Leadership
and Support

This dissertation was submitted to the Graduate Faculty of the
Department of Counselor Education in the College of Education and
to the Graduate Council, and was accepted as partial fulfillment
of the requirements for the degree of Doctor of Philosophy.
August, 1983
Dean for Graduate Studies and
Research

UNIVERSITY OF FLORIDA
3 1262 08552 9351