The effect of enhanced vs. non-enhanced directed positive imagery upon the reinforcement of effortful academic behavior


Material Information

The effect of enhanced vs. non-enhanced directed positive imagery upon the reinforcement of effortful academic behavior
Effect of enhanced vs. non-enhanced directed positive imagery ..
Physical Description:
xiii, 160 leaves : ill. ; 28 cm.
Edenfield, William Horace, 1946-
Publication Date:


Subjects / Keywords:
Imagery (Psychology)   ( lcsh )
Behavior therapy   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis--University of Florida.
Includes bibliographical references (leaves 148-158).
Statement of Responsibility:
by William Horace Edenfield.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 06505613
System ID:

Full Text

This is an authorized facsimile, made from the microfilm
master copy of the original dissertation or masters thesis
published by UMI.

The bibliographic information for this thesis is contained in
UMI's Dissertation Abstracts database, the only central
source for accessing almost every doctoral dissertation
accepted in North America since 1861.

TU -TI Dissertation
SIV I Information Service
University Microfilms International
A Bell & Howell Information Company
300 N. Zeeb Road, Ann Arbor, Michigan 48106
800-521-0600 OR 313/761-4700

Printed in 1991 by xerographic process
on acid-free paper





The University of Florida PH.D. 1979

International 300N.Zeeb Road,AnnArb M148106 18 Bedford Row. Loadon WC1R 4E. England


This was produced from a copy of a document sent to us for microfilming. While the
most advanced technological means to photograph and reproduce this document
have been used, the quality is heavily dependent upon the quality of the material

The following explanation of techniques is provided to help you understand
markings or notations which may appear on this reproduction.

1. The sign or "target" for pagesapparently lacking from the document
photographed is "Missing Page(s)". If it was possible to obtain the missing
page(s) or section, they are spliced into the film along with adjacent pages.
This may have necessitated cutting through an image and duplicating
adjacent pages to assure you of complete continuity.

2. When an image on the film is obliterated with a round black mark it is an
indication that the film inspector noticed either blurred copy because of
movement during exposure, or duplicate copy. Unless we meant to delete
copyrighted materials that should not have been filmed, you will find a
good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., is part of the material being photo-
graphed the photographer has followed a definite method in sectioningg"
the material. It is customary to begin filming at the upper left hand corner
of a large sheet and to continue from left to right in equal sections with
small overlaps. If necessary, sectioning is continued again-beginning
below the first row and continuing on until complete.

4. For any illustrations that cannot be reproduced satisfactorily by
xerography, photographic prints can be purchased at additional cost and
tipped into your xerographic copy. Requests can be made to our
Dissertations Customer Services Department.

5. Some pages in any document may have indistinct print. In all cases we
have filmed the best available copy.

MrU rtilms







3 1262 07332 021 9


I would like to express my appreciation to those people

who have assisted me in the development of this study:

Dr. Don Avila, chairperson of my doctoral committee,

whose friendship and understanding gave me the opportunity

to grow;

Dr. Walt Busby, committee member, for teaching me to

have faith in myself;

Dr. Bob Myrick, committee member, who gave me encourage-

ment when I needed it most;

Dr. Bob Jester, committee member, for his dedication

as a teacher and a friend;

Dr. Lee Rowell, Director of Student Services, Marion

County School System, for his administrative and moral support;

Marilyn Highland, Kay Castora, Jane Warren, and Jade

Maxey, the counselors who assisted me in the implementation

of the research;

Frances Smith for her dedication and perserverence in the

typing of this dissertation;

Angel and Julie Perez, my father- and mother-in-law, for

their love and personal support;

And especially, my wife, Gail, who, I found during the

development of this dissertation, loves me more than I ever

knew; and my five-year-old son, Jonathan, who insisted on

keeping me in constant supply of sharpened pencils during

the writing of the draft, and who typed the title of this


"Many people feel that ideas can be considered
true or scientific only if they can be physically measured.
Actually, it is only when science is able to make inferences
about things that it is able to progress beyond the immediate
and the palpable to deal with the abstract and remote. A
science limited to that which is directed observable would
never be able to deal with such matters as electricity,
atomic physics, or human disease."

Combs, Richards, and Richards (1976, p. 378)




Rationale for the Study . 4
Definition of Terms . 7
Statement of the Problem 8
Organization of the Remainder
of this Study . 8


Overview of the Development
of Directed Imagery .. 10
Development of the Covert Positive
Reinforcement Procedure .. 13
The Procedure of Covert
Positive Reinforcement ....... 16
Clinical Case Studies Employing
Positive Reinforcing Imagery 19
Experimental Studies Employing
Positive Reinforcing Imagery 23
Studies Specifically Investigating
Variables Related to the Presen-
tation of Positive Reinforcing
Imagery . 47


Subjects . 58
Design of the Study ... 59
Hypotheses . ... 60
Procedure . 63
Selection of the Target Behavior 69
Development of the Instrumentation 70
Development of the Directed Imagery
Treatment Tapes ... 74




Summary . .
Discussion of Results .
Conclusions . .
Implications for Clasroom and
Clinical Practice .
Limitations of the Study .
Recommendations for Further
Research . .














S 91
S 94
S 99

. 100
. 105

. 107

. 110

. 111

. 113

. 127

. 129

. 131

. 133

. 134


. 136

. 137














GROUPS 1, 3, AND 5 . .. 84

ONLY GROUPS 2, 4, and 6 ... 88

AND 5 & 6 . ... 90




William Horace Edenfield

December 1979

Chairperson: Dr. Donald L. Avila
Major Department: Foundations of Education

Directed imagery, a technique whereby a patient or sub-

ject is verbally encouraged to engage in the mental produc-

tion of suggested sensory experiences, and particularly

directed positive imagery, may be a potentially valuable

tool for future psychotherapeutic, medical, and educational

applications. However, investigation of variables related

to the presentation of directed positive imagery is rela-

tively limited, and it appears that no published research

has directly investigated the relative effects of variations

in the quality of verbal presentations of directed positive

imagery upon therapeutic gain. This study investigated the

relative effects of descriptively enhanced and non-enhanced

directed positive imagery presentations upon the rein-

forcement of effortful academic behavior in a classroom


A total of 342 students assigned to 12 fifth-grade class-

rooms at four elementary schools were randomly assigned, in

intact-classroom-groups, to the following experimental

groups: (1) no treatment control: pre- and posttested;

(2) no treatment control: posttested-only; (3) non-enhanced

imagery treatment: pre- and posttested; (4) non-enhanced

imagery treatment: posttested-only; (5) enhanced imagery

treatment: pre- and posttested; and (6) enhanced imagery

treatment: posttested-only. During the first week of this

six-week study, subjects were administered a food-treat rein-

forcer survey to identify group-reinforcers later to be

presented in directed imagery treatment sessions. Two weeks

later, all subjects were administered a pretest involving

the effortful and accurate computation of arithmetic problems.

Over the following three weeks, subjects in the four treat-

ment groups received five directed imagery treatment sessions.

During each treatment session they first engaged in a brief

relaxation exercise and a directed imagery arithmetic compu-

tation exercise which was immediately followed by the

directed imagery presentation of pre-selected food-treat

reinforcers. The relaxation and imaginary arithmetic

exercises were identical for all treatment groups. However,

subjects in the two non-enhanced imagery treatment groups

were presented with non-specific imagery descriptions of

each of the five pre-selected food reinforcers, and subjects

in the two enhanced imagery treatment groups were presented

with embellished imagery descriptions of the same five food-

treat reinforcers. During the sixth week all subjects were

posttested with the same arithmetic computation task.

Comparisons of pretest and posttest performance indi-

cated significant increases in performance for the two

treatment groups, but not for the control group. Furthermore,

analyses based upon posttest measures and pre- to posttest

gain measures indicated a significant difference between

the control group and both treatment groups, but not between

the two treatment groups. Identical relationships were found

for the three posttested-only groups. An analysis of post-

test performance measures for the three posttested-only

groups indicated a significant difference between the

posttested-only control group and both treatment groups,

but not between the two treatment groups. Finally, compari-

sons between posttest and posttest-only performance measures,

for each of the two treatment conditions, indicated signi-

ficant differences for the non-enhanced imagery treatment

condition, but not for the enhanced imagery treatment


It was concluded that the operant presentation of

directed positive imagery of pre-selected food-treat

reinforcers, in a classroom setting, positively affected

effortful academic behavior as measured by performance

on a task involving the effortful and accurate computation

of arithmetic problems. Furthermore, qualitative dif-
ferences in the verbal presentation of directed positive

imagery, based upon the degree of descriptive enhancement,

did not affect differences in the reinforcement effects of

positive-reinforcing imagery. Finally, at least for the

enhanced imagery treatment condition, it appeared that pre-

treatment overt exposure to the specific target-behavior

task had no effect upon the outcome of the imaginary

reinforcement procedure.




The purpose of this study was to investigate the

relative effects of enhanced and non-enhanced directed

positive imagery upon the group reinforcement of effort-

ful academic behavior in a classroom setting.

The use of directed imagery, whereby a patient or

subject is verbally encouraged to actively engage in the

mental production of suggested sensory experiences, was

first reported as a psychotherapeutic technique of Freud

in the late nineteenth century (Kelley, 1972). Following

Freud, other psychoanalytic practitioners later adopted

variations of the directed imagery technique. During an

approximately 40-year moratorium in the study of "inner

experiences" by American psychologists, it was the

European psychotherapists who further developed and for-

malized a variety of therapy techniques employing directed


Probably the most popularized application of directed

imagery, formalized by Wolpe (1958), appeared in America

toward the end of the behaviorist-inspired moratorium.

Ironically, Wolpe's "systematic desensitization" was based

upon the principles of counter-conditioning, and it was


other behaviorists such as Stampfl and Levis (1967) who

later developed other popularized directed imagery

techniques such as implosivee therapy."

The resurgence of American psychologists' interest in

imagery, however, was not limited to the psychoanalytic

and behavioristic camps. Over the past two decades, "third

force" practitioners have also shown an increasing use of a

variety of applications of both directed and non-directed

imagery in existential-oriented therapies and human growth


A more recent development in the behaviorally-oriented

application of directed imagery is a procedure termed

"covert positive reinforcement," formalized by Cautela

(1970). The procedure follows an operant conditioning

paradigm and is labeled "covert" because the response to

be manipulated and the reinforcing stimulus are presented

in the patients's imagination.

In the original version of this procedure, the

patient is first trained to generate imagery of a pleasant

activity (e.g., eating a favorite or preferred food).

During formal treatment, the patient is asked to imagine

performing some desired response (e.g., engaging-in or

completing a particular task), and then shift to the pre-

trained reinforcing image. The imagery shift is initially

cued by the therapists's pronouncement of the word "rein-

forcement." Cautela (1970) reports that covert posititive

reinforcement has become his main therapeutic approach,

occasionally supplemented with other behavior therapy


Directed imagery procedures, particularly those employing

positive or pleasurable imagery,appear to offer a number of

practical advantages over some of the more traditional

psychotherapeutic and behavior management techniques.

Cautela (1973) emphasizes a number of advantages in the use

of general imagery techniques: (1) the use of imagery

requires little, if any, equipment and procedures can be

applied in almost any setting; (2) the use of imagery is

not limited to the temporal, economic or other practical

restraints of reality imposed upon many other forms of

psychotherapy; and (3) imagery can easily be used as a

self-control procedure, enabling the patient to engage in

self-application outside of the clinical setting. In

addition, Kelly (1972) states that techniques employing

directed imagery appear to be especially appealing to

young children and adolescents.

Cautela further emphasizes that in terms of clinical

flexibility and minimizing resistance to treatment, positive-

reinforcing imagery in particular offers advantages over the

other two main behavioral approaches using directed

imagery. Whereas systematic desensitization (Wolpe, 1958)

and implosive therapy (Stampfl and Levis, 1967) are used

to modify maladaptive avoidance behaviors, and covert

desensitization (Cautela, 1966) is employed to modify

maladaptive approach behavior, positive reinforcing imagery

may be used to modify both maladaptive approach and mal-

adaptive avoidance behaviors. Additionally, the use of

positive imagery eliminates the necessity of using

aversive imaginary stimuli.

Rationale for the Study

Directed imagery techniques, and particularly pro-

cedures employing positive or pleasurable imagery, not

only appear to offer a number of practical advantages

over some of the other psychotherapeutic and behavior

management techniques, but furthermore may offer a

promising foundation for potential extension into a

variety of fields of application.

Jones (1968) emphasized the potential importance of

imagery in facilitating academic instruction and

developing cognitive skills. More recently Wittmer and

Myrick (1974) suggested that directed imagery is a poten-

tially powerful tool, not only for use in facilitating

classroom instruction, but in learning and improving

motor skills as well. The potential for such applications

is further supported by laboratory findings demonstrating

the importance of imagery in facilitating memory (Pavio,

1969; Pavio and Rowe, 1970) and concept learning (Pavio,

1974), in addition to the findings of a number of studies

demonstrating the positive effects of imagery upon the

improvement of a variety of motor skills (Harby, 1952;

Hammer and Natale, 1964; and Richardson, 1967).

The potential value of imagery techniques as a tool

to facilitate the enhancement of self concept has been

suggested by Combs, Richards and Richards (1976), and

supported by the findings of Krop, Calhoon, and Verrier

(1971), and Krop, Perez, and Beaudoin (1973). Directed

positive imagery has also been shown to be of promising

value in the modification of attitudes toward other per-

sons (Cautela and Wisocki, 1969; Cautela, Walsh, and Wish,

1971; and Euse, 1975) and certain activities such as

reading (Daniels, 1976).

Directed positive imagery techniques may also offer

very significant potential usefulness in the practice of

medicine and dentistry. Studies indicating autonomic and

other physiological changes accompanying imagery have

been reviewed by Sheehan (1972), and it is suggested that

directed imagery may be a potentially important tool in

the prevention, diagnosis, and/or treatment of a number of

medical disorders including allergies, pain, dermatological,

cardiovascular, gastrointestinal disorders (e.g., Kroger

and Fezler, 1976; Samuels and Bennett, 1973), tooth decay

(Nigl, 1975), and cancer (Achterberg and Lawlis, 1978;

Bry, 1978).

In spite of the previously demonstrated and potential

applications of directed imagery, it appears that the under-

lying foundations for the application of imagery techniques,

particularly those involving positive or pleasurable

imagery, lack sufficient investigation. Holt (1964), in

addressing the potential contributions of imagery applica-

tions, states that imagery has only recently been redis-

covered as a subject for systematic investigation and

clinical application, after fifty years of relative neglect.

Singer (1971) suggests that we are on the threshold of

important exploration into the image-construction and

concomitant affect-producing repertory of humans. He

further emphasizes that if the manipulation of a patient's

imagery is to become a central feature of psychotherapy,

we must develop underpinnings for the application of imagery

techniques. Travers (1972) contends that imagery has long

been a neglected area of psychological inquiry. More

recently, Wilkins (1974, p. 163) stated that, "While

investigators attest to the effectiveness of imagery

procedures, research attempting to identify the specific

parameters of therapeutic gain has not been systema-

tically undertaken."

While there seems to be strong evidence that positive

or pleasurable imagery can affect changes in clinical and

non-clinical behaviors, there appears to be little or no

research directly evaluating the manipulation of para-

meters related to the quality of verbal description in

the presentation of directed positive imagery. Such is

the purpose of the present investigation.

Definition of Terms

Imagery An individual's conscious mental production

of perceptions of sensory experiences of persons, places,

objects or events no longer present to sensory receptors.

The qualities of such perceptions are determined by the

mental recollection of real previous experience, or

learned concepts.

Directed Imagery A technique whereby a subject is

verbally encouraged to actively engage in the production

of imagery, the content of which, to varying degrees, is

suggested or prescribed by the therapist.

Directed Positive Imagery Directed imagery, the

suggested or prescribed content of which is associated

with satisfying or pleasure-arousing circumstances.

Non-Enhanced Directed Positive Imagery Directed

positive imagery, the suggested or prescribed content of

which is described by the therapist, only in the most-

essential, simple and general terms.

Enhanced Directed Positive Imagery Directed positive

imagery, the suggested or prescribed content of which is

described by the therapist, utilizing elaborated des-

criptions involving multi-sensory stimuli, the qualities

of which are provided descriptive clarity and detail

through adjectival and adverbal embellishment and directed

attention through questioning.

Effortful Academic Behavior The conscious and

deliberate engagement in sustained mental and/or physical

activity related to the acquisition of knowledge or skill

in the areas of reading, writing, or arithmetic.

Statement of the Problem

The research questions tested experimentally in this

study were: (1) Can the presentation of directed positive

imagery, positively affect effortful academic behavior, in

a classroom setting? (2) Do differences in the quality of

the verbal presentation of directed positive imagery affect

differences in the treatment effect of directed positive

imagery? and (3) Does pretreatment overt exposure to the

specific target-behavior task, as later presented in an

imaginary operant conditioning procedure, affect differences

in the outcome of the imaginary reinforcement procedure?

Organization of the Remainder of this Study

The remainder of this study is organized into four

additional chapters and appendices. Chapter II provides

an overview of the growth of directed imagery techniques,

a description of the imaginary (covert) positive rein-

forcement procedure and its underlying assumptions, a

review of the studies employing or investigating the pro-

cedure, and an overview of the importance of descriptive

enhancement to literary imagery. Chapter III includes the

design and methodology, and the development of the instru-

mentation and directed imagery treatment tapes used in

this study. The data analyses and results of this study are


presented in Chapter IV. In Chapter V, a summary and the

conclusions of this research, along with recommendations

for further investigation, are presented.



Overview of the Development of Directed Imagery

The use of directed imagery as a psychotherapeutic tech-

nique was first reported in the case of Anna 0., by Breuer

and Freud in 1895. Freud, however, later discarded the use

of directed imagery and developed psychoanalytic methods

along the related lines of free association and dream analy-

sis (Kelly, 1972).

Several years later Jung (1928) theorized that the

psychological mechanism which transmutes energy is the sym-

bol, and later developed the psychotherapeutic technique of

"active imagination" by which patients were encouraged to

visualize mental images as part of the psychoanalytic method

(Jung, 1960). Jung's "active imagination" method, however,

continued to be carried on largely in a context of anecdotal

report and discussion with little indication of formal

experimental study (Singer, 1971).

During the same period, other European psychotherapists

began to show interest in the development of directed imagery

as a therapeutic method. It was the mediative psychothera-

peutic approaches described by Kretschmer (1922) and Happich

(1938) along with the systematic mediative methods described

by Schultz (1932) for his autogenicc training technique"

which formed the foundations for later investigation and

application of directed imagery (Kelly, 1972).

Following the work of Kretschmer, Happich, and Schultz,

it was the French psychotherapist Robert Desoille who was

one of the principal developers of directed imagery tech-

niques during the American moratorium in the study of "inner

experiences." Desoille, reflecting his interest in the

Jungian concepts of the "collective unconscious" and "uni-

versal symbolism," began to develop a system that he called

"le reve eveille dirige," or the "guided daydream" (Desoille,

1938). Desoille's approach has probably become a major in-

fluence on the current group of practitioners who are active

in the use of imagery techniques (Singer, 1971).

A similar technique called "guided affective imagery"

was developed through 20 years of experimental study

conducted by Hanscarl Leuner (1968). Desoille's and Leuner's

directed imagery techniques were developed and applied in a

primarily psychoanalytically-oriented framework, where

specific mental imagery was directed by the therapist who

assisted the patient to gain insight through interpretation

of the imagined symbols and affective responses. The diag-

nostic aspect of Leuner's work termed "initiated symbol

projection" was introduced to American psychotherapy by

Swartley (1965).

The most popularized application of directed imagery

in America was formalized by Wolpe (1958). Wolpe's "system-

atic desensitization" appeared toward the end of the

behaviorists-inspired moratorium on "inner-experience" re-

search, ironically representing the application of directed

imagery based upon principles of conditioning. Essentially,

this application involves the imaginary presentation of

progressively increasing anxiety-provoking objects or situa-

tions, during a state of maximal relaxation, which is the

key feature of the counter-conditioning, arousal-reducing

procedure. A variant of Wolpe's procedure was later de-

veloped by Lazarus and Abramovitz (1962) for use with

younger children and later extended to use with adults.

Stampfl and Levis (1967) later designed a procedure

opposite to Wolpe's "systematic desensitization," called

implosivee therapy," employing directed imagery to increase

anxiety during the therapeutic paradigm. In this method the

general procedure is to have the patient imaginally enact

the worst possible consequences of a particular fear or

obsession, thereby arousing strong anxiety. In the absence

of concrete aversive consequences, the anxiety is gradually

extinguished in the course of repeated imaginary exposure.

The resurgence in the interest in, and application of,

imagery techniques has not been restricted to the psycho-

analytic and behavioristic camps. "Third force" practi-

tioners have, over the past two decades, shown an increasing

interest in, and use of, varied applications of both directed

and non-directed imagery "exercises," "experiments," and

"growth games" (e.g., Kelly, 1972; Stevens, 1971; Shorr,

1972; Schultz, 1967; Fagen and Sheppard, 1970; Perls, 1973;

and Perls, Hefferline, and Goodman, 1951). Humanistically-

oriented therapeutic applications of directed imagery have

included procedures to facilitate recollection of previous

psychodynamic experiences, generate and elevate emotional

arousal, facilitate relaxation, provide coping mechanisms,

focus concentration and attention, and conduct vicarious

rehearsal of more effective behaviors.

Development of the Covert Positive Reinforcement
(CPR) Procedure

A more recent behaviorally-oriented psychotherapeutic

technique employing directed imagery is a procedure developed

by Cautela (1970), termed "covert positive reinforcement"

(CPR). The procedure employs positive imagery as the re-

inforcing stimulus in an operant conditioning paradigm, in

which the response to be manipulated and the reinforcing

stimulus are presented in the patient's imagination.

Acknowledging that imagery has only relatively recently

become the subject of investigation by behavioristically-

oriented psychologists, Cautela (1973) credited Wolpe (1958)

for the major contribution of demonstrating that various

manipulations of covert processes in the consulting room

could affect behavior outside of the dyadic therapeutic

situation. In drawing additional support for his assump-

tions regarding the efficacy of CPR, Cautela (1970) cited

the work of Lang (1964), Paul (1966), Davidson (1965), and

Stampfl and Levis (1966), whose findings also indicated the

manipulation of imagery, when used in a desensitization

paradigm, could effectively modify maladaptive avoidance


Prior to his formalization of the CPR procedure,

Cautela's initial interest in imagery led to the develop-

ment of a variant of Wolpe's systematic desensitization

which Cautela (1966, 1967) labeled "covert sensitization."

He describes covert sensitization as a procedure for estab-

lishing an avoidance response to an undesirable stimulus

(e.g., alcohol, cigarettes, fattening foods, etc.).

In covert sensitization, the conditioned stimulus (e.g.,

a cigarette) and the aversive stimulus (e.g., vomiting over

oneself) are both presented and paired in the patient's

imagination, thereby lending to the label "covert" sensi-

tization. After a number of exposures to such imaginary

pairings, the aversive arousal associated with the original

conditioned stimulus tends to weaken or eliminate the initial

attraction. Cautela (1970) claimed that empirical validation

of his covert sensitization procedure (Stuart, 1965; Ashem

and Donner, 1968; Viernstein, 1968; and Mullen, 1968) pro-

vided further support for his speculation that presenting a

positive or pleasure-arousing stimulus in the imagination

might influence behaviors in a manner similar to the ex-

ternal presentation of a reinforcing stimulus.

Cautela's speculation regarding the reinforcing pro-

perties of positive imaginary stimuli is based upon two

major assumptions:

(1) The functional relationship between behavior and
a reinforcing stimulus presented in the imagina-
tion is similar to the functional relationship
between behavior and a reinforcing stimulus
externally presented.

(2) The manipulation of covert processes can in-
fluence overt processes in a predictable manner.

In drawing support for his first assumption, Cautela

(1970) cited Pavlov's (1955) work on the equivalence of

primary and secondary signaling systems, Guthrie's (1935)

discussion regarding "intention" and "maintaining stimuli,"

and Hull's (1952) theories regarding the relationship between

the fractional antedating reaction and its stimulus cor-

relate, in relation to processes such as planning, fore-

sight, expectation, etc. More recent support is derived

from the work of Homme (1965), who expressed the conviction,

from an operant conditioning perspective, that "operants of

the mind" obey the same laws as observable operants. Skinner

(1969) also postulated that private events obey the same

laws as overt events.

Support for the second assumption was initially drawn

from the results of previously cited clinical and case

studies on systematic desensitization, and empirical vali-

dation of the covert sensitization procedure. Additional

support was taken from Kimble's (1961) research demonstrating

stimulus generalization and Kimble's conclusion that since

thoughts are capable of providing cues for overt behavior,

they can theoretically provide the same conditions for the

control of behavior as environmental events do. Cautela

(1970) also cited Franks (1967) who concluded that conditioning

to imaginary situations could be readily acquired and gener-

alized quickly to reality levels. In final support of his

second assumption, Cautela presented a review of experimental

studies by Kimble (1961) as demonstrating the efficacy of

stimulus generalization, and experimental evidence support-

ing the concept of response generalization (Williams 1941;

Arnold, 1945; and Antonitis, 1951), concluding that on the

basis of stimulus-response generalization there can be a

transfer to overt conditioning from the imagination.

The Procedure of Covert Positive Reinforcement

Although a number of subsequent studies, reviewed

later in this chapter, have employed minor variations in

Cautela's (1970, 1972, 1973) classical CPR procedure, the

basic method involves the operant presentation of an ima-

ginary reinforcer subsequent to the imaginary presentation

of an unconditioned response or an increment of a complex

unconditioned response. Furthermore, the imaginary rein-

forcer does not necessarily have to cognitively relate to

the unconditioned response, as demonstrated in a forth-

coming example. Finally, the use of CPR does not necessi-

tate prior relaxation training nor the construction of

hierarchies as in Wolpe's (1958) systematic desensitiza-

tion procedure (Cautela, 1970).

The first step in the procedure is to identify possible

reinforcing stimuli. To do this, Cautela suggested three

possible sources: (1) administration of the Reinforcement

Survey Schedule (RSS), a 54-item survey designed for adults,

which asks a patient to rate the reinforcement value of a

variety of stimuli and experiences on a 5-point scale, in-

dicating the level of enjoyment or satisfaction associated

with each item (Cautela and Kastenbaum, 1967); (2) the

therapist merely asking the patient to suggest or identify

events or objects which would be pleasurable; (3) soliciting

information about possible reinforcing stimuli through case

histories, relatives, friends, and ward personnel (if the

patient is hospitalized). Cautela emphasized that it is

important to identify a number of potential reinforcers so

that (covert) satiation will not occur. Ayllon and Azrin

(1964) found that if a reinforcer is used too frequently it

can lose its effectiveness. For this reason Cautela often

varied the reinforcers, even within the same treatment


After the identification of one or more potential

reinforcing objects or events, the treatment phase is imple-

mented. Cautela (1970, p. 39) presented an example of the

CPR procedure, in this case involving the treatment of a

maladaptive avoidance syndrome through the imaginary posi-

tive reinforcement of an incompatible adaptive response.

In the example, a doctoral candidate who had flunked

his doctoral qualifying examination three times was re-

ferred for treatment because of extreme test anxiety. In

three months he was to be given his last chance to pass his

qualifying examination. The patient was considered an

excellent student, as indicated by his grades on papers,

semester examinations, and oral discussion. However, in

addition to suffering extreme test anxiety, he would become

anxious and could not concentrate when he attempted to

study. The following instructions were utilized in the

directed imagery procedure:

Close your eyes and try to relax. I want you to
imagine you are sitting down to study and you
feel fine. You are confident and you are relaxed.
I know you may be anxious here but try to imagine
that when you are about to study you are calm and
relaxed, as if you were acting a part. Start.
(When the S raises his finger, the therapist de-
livers the word.) "Reinforcement" which in the
case signals the image of skiing down a mountain
feeling exhilarated. Practice this twice a day
and just before you study.

Now let's work on the examination situation. It
is the day of the examination and you feel con-
fident. ("Reinforcement.") You are entering the
building in which the exam is going to be given.
("Reinforcement.") You remember that in all
these scenes you are to try to feel confident.
Now you enter the building and go into the class-
room. ("Reinforcement.") You sit down and kid
around with another student who is taking the
exam. ("Reinforcement.") The proctor comes in
with the exam. You feel good; you know you are
ready. ("Reinforcement.") The proctor hands out
the exam. ("Reinforcement.") You read the ques-
tions and you feel you can answer all of them.

Now let's do that again. This time you look the
questions over and you are not sure about one
question, but you say, "Oh well, I can still pass
the exam if I flunk this one question." ("Rein-
forcement.") All right, this time you look over
the exam, and you can see two questions about
which you are in doubt, and you say, "Well, I
can still pass this exam if I take my time and
relax." ("Reinforcement.")

The patient was then given the usual instructions to

practice the entire sequence at home, and was also told to

vary the reinforcing stimulus (e.g., another scene was

receiving his doctoral degree at graduation). As the ses-

sions continued, he reported that he felt more relaxed while

studying and was able to keep up with his study schedule.

After ten sessions he took the doctoral examination, later

reporting that he was more confident and relaxed and that he

passed the exam (Cautela, 1970).

In the treatment of such maladaptive behaviors Cautela

suggested that, after practicing the entire sequence, the

patient be told to practice once or twice a day at home,

doing the procedure twice during each practice session.

More recent clinical and experimental reports, to be re-

viewed later in this chapter, have indicated significant

covert reinforcement effects with significant variations

in the number of treatment sessions.

Clinical Case Studies Employing Positive
Reinforcing Imagery

Since the formalization of Cautela's CPR procedure,

directed positive imagery has been successfully applied to

the treatment of a variety of clinical behaviors in a number

of case studies.

In one of the earliest reported case studies employing

positive-reinforcing imagery, Wisocki (1970) used CPR in

combination with covert sensitization in treating a 27

year old obsessive-compulsive female who was excessively

compulsive about housework. After the maladaptive behavior

was eliminated using covert sensitization (i.e., aversive

imagery paired with imagery of repeated folding of clothes),

CPR was used to strengthen new behaviors intended to replace

those eliminated (i.e., reinforcing imagery paired with

imagery of folding clothes only once and putting them in

the drawer). A 6-month follow-up indicated that the

compulsive behavior had not recurred.

In a later study, Wisocki (1973) successfully used CPR

together with covert sensitization and "thought stopping"

in treating a 26 year old heroin addict. The patient had

a 3-year history of heroin abuse and had previously been

through most of the major drug programs, including Synanon

and methadone maintenance, without success. The CPR pro-

cedure was used to reinforce thoughts and behaviors antagon-

istic to the use of heroin. Thought stopping was used to

reduce thoughts of, and urges for, heroin and covert sensi-

tization was used on behaviors involved in the process of

heroin use. An 18-month follow-up showed that the patient

had not returned to heroin, had married, and was employed


Suinn (1972), in an application of CPR similar to that

of Cautela (1970), successfully treated a doctoral student

suffering extreme anxiety related to his doctoral orals.

He was instructed to imagine himself successfully and con-

fidently answering questions in front of his doctoral

committee, after which he was instructed to immediately

imagine himself experiencing positive feelings as a result

of his success.

Kendrick and McCullough (1972) reported the successful

application of CPR in the treatment of a 21 year old male

homosexual who suffered from depression and stuttering. The

CPR procedure was combined with assertiveness training in

eight weekly therapy sessions. Upon a 2-month follow-up,

the patient reported less frequent occurrences of homosexual

urges and a steady female relationship. No determination

as to the relative contribution of the CPR procedure was

made, however.

Cautela and Baron (1973) successfully combined CPR and

covert sensitization in the treatment of a 20 year old male

self-mutilator. The patient had intentionally blinded him-

self by gouging his eyes with his fingers, and had inten-

tionally bitten his lips and tongue so severely that plastic

surgery was required. He was taught relaxation as a self-

control procedure to be used whenever he felt urges to

injure himself, and was instructed to engage in the imagina-

tion of pleasant scenes following relaxation as reinforce-

ment for self-control behavior. Covert sensitization was

used as a means of reducing the actual engagement of self-

injurious behavior. Thought stopping was also employed to

block thoughts which acted as antecedents to the self-

mutilating behavior. Following treatment the patient

ceased engaging in self-mutilating behavior.

In a carefully controlled case study, Blanchard and

Draper (1973) tested CPR in the treatment of a 23 year old

rat phobic psychology student whose college career was being

severely affected as a result of her avoidance of classes,

conversation, or reading in which rats and mice might be

mentioned. The study consisted of six phases: (1) "insight

psychotherapy"; (2) a CPR procedure in which the subject

imagined a graduated series of approach behaviors related

to rats and mice, followed by an imaginary reinforcer; (3) a

control phase in which the imagined approach behavior was

not followed by the imagined reinforcer; (4) reinstatement

of the CPR procedure; (5) no treatment for 1 week; and

(6) a modeling phase in which the experimenter modeled the

approach behavior and encouraged the subject to engage in

modeling. The CPR phase consisted of six 30-minute sessions

in addition to practice at home using a tape of recorded

instructions. Results of an administration of an approach

behavior test indicated greater improvement during the CPR

phase as compared to the "psychotherapy" phase.

Without the imagined reinforcer the approach behavior

only increased slightly; however, it further increased when

CPR was reinstated. During the removal of CPR, three sub-

jective measures indicated increased fear of rats and mice.

During the no treatment phase there was a decrease in the

approach behavior. The clinical treatment lasted for 8

weeks and a 4-month follow-up indicated that improvements

made during therapy had been maintained. The authors con-

cluded that while CPR facilitated positive changes in

approach behavior, CPR was not necessary to initiate or

maintain positive changes. On the other hand, CPR did appear

necessary to initiate and maintain positive changes in

self-report measures of fear.

Pfeiffer (1977) reported a case where CPR was success-

fully used in the treatment of a 19 year old high school

student who was referred for aggressive behavior which in-

cluded the physical abuse of his girlfriend and verbal

abuse of both she and her parents. Treatment involved

weekly 45-minute sessions which included the CPR procedure.

Aggressive behavior decreased significantly upon commence-

ment of the treatment procedure, and 6- and 12-month follow-

ups indicated that only sporadic and mild verbal aggression

was exhibited.

In a more recent case study, Workman and Dickinson

(1979) demonstrated the effectiveness of CPR in the treat-

ment of hyperactivity in a 9 year old third grade student.

The CPR procedure was conducted for six sessions over a

3-week period in an effort to reduce out-of-seat behavior,

excessive noisemaking, and rocking-in-chair behaviors.

An immediate improvement in the three problem behaviors, as

measured by pre- and posttreatment baseline observations,

was indicated. Only the rocking-in-chair behavior showed

a reversal effect upon a 6-week follow-up.

Experimental Studies Employing Positive
Reinforcing Imagery

In addition to single-subject clinical case study

applications, positive reinforcing directed imagery has

received attention in a number of experimental investiga-

tions involving both clinical and non-clinical behaviors.

One of the earliest experimental investigations of CPR

attempted to change perception of circle size by cueing

subjects' imaginary production of a reinforcer contingent

upon the overestimation or underestimation of the size of

a series of circles. Wish, Cautela, and Steffan (1970)

randomly assigned 50 college students to one of five groups:

(1) subjects receiving CPR contingent upon the desired

response; (2) subjects presented with a neutral scene con-

tingent upon the desired response; (3) subjects receiving

CPR non-contingent upon the desired response; (4) subjects

receiving only the word "reinforcement" contingent upon the

desired response; and (5) a no-feedback control group.

Subjects were instructed to estimate the size of a series

of circles presented during four experimental phases, the

first of which was a baseline phase. In the second and

third phase, groups received respective experimental treat-

ments, with subjects in CPR groups being cued to imagine

scenes previously determined to be individually reinforcing

based upon answers to the Reinforcement Survey Schedule

(RSS) (Cautela and Kastenbaum, 1967). The fourth phase was

identical to the first baseline phase.

No significant differences were found among groups

during the first baseline phase. However, following experi-

mental treatment, results indicated that contingent-CPR

subjects' circle-size estimations were significnatly dif-

ferent from the neutral scene group, the non-contingent-CPR

group, and the control group. There was also significant

difference between the word-"reinforcement"-only group and

the control group, indicating that the word "reinforcement,"

in itself, may have reinforcing properties. This study

represented a variation of Cautela's (1970) original version

of CPR in that the unconditioned response was not presented

in the imagination, and the authors concluded that covertly

administered reinforcement could modify overt behavior.

Venters (1972) investigated the comparative effects of

CPR and overt reinforcement upon changes in circle-size

estimation. The CPR subjects were presented with rein-

forcing imagery following imaginary scenes of circles grow-

ing larger or smaller. Overt reinforcement subjects were

overtly rewarded for engaging in the same imagery, and

control group subjects were not given reinforcement. Analy-

sis of pre- and posttreatment circle size estimations

showed no significant effect in either treatment group,

based upon underestimation scores. An analysis of over-

estimation scores obtained among the three groups, however,

indicated a significant greater reinforcement effect among

CPR subjects as compared to overt reinforcement and control

subjects. Failure to reach significance for the under-

estimation data raised questions regarding appropriateness

of the task.

Tondo (1975), investigating the "functional similarity"

between overt and imagined reinforcement, pretrained 36

undergraduate students to earn points by carefully timing

key-press responses, with one point being delivered

contingent upon each response occurring 9-12 seconds after

the preceding response. The pretraining phase was followed

by a baseline phase, followed by an overt reinforcement

phase, a CPR phase, and a control-condition phase, presented

in a counterbalanced order for six groups.

Subjects were instructed that $5 would be awarded to

the highest scorer. Although the 9-12 second response in-

terval remained in effect during the treatment phases,

additional events occurred contingent upon responses falling

with an even smaller time-interval within the pretrained

9-12 second interval. During the overt reinforcement phase,

the additional events were the illumination of an orange

light discriminativee stimulus) followed by an "extra"

point registered on a visible counter, contingent upon the

key-press response falling within the smaller, 10-11 second

inter-response range. During the covert positive reinforce-

ment phase, the additional events were the illumination of

the orange light paired with instructions for subjects to

imagine that one extra point was added to the score on the

counter. During the feedback-only condition, the additional

events were the illumination of the orange light paired

with the instructions to simply observe its occurrence.

Results indicated that there was no significant dif-

ference among groups during the baseline phase; however,

the overt and imagined reinforcement conditions significantly

increased the frequency of target responses when compared

to the control condition. Furthermore, no significant

difference between the overt reinforcement and the imagined

reinforcement conditions was found.

In examining subject preference for imaginary rein-

forcement, compared to real reinforcement, Trent (1977)

randomly assigned 40 volunteers to four treatment groups:

CPR vs. no overt reinforcement; CPR vs. points-only rein-

forcement; CPR vs. .14 per point reinforcement; and CPR vs.

1i per point reinforcement. After establishing baseline

measures for choice preferences between two response keys,

subjects were presented with the choice dichotomy respec-

tive to their treatment group, being cued to engage in

reinforcing imagery or receiving one of the aforementioned

alternatives, depending upon the response key chosen. Re-

inforcing scenes were individually predetermined by responses

on the RSS. This phase was followed by a 5-minute extinc-

tion phase.

Results indicated a clear preference for reinforcing

imagery as compared to points-only and .1 per point. As

the value of the overt reinforcer was increased, subjects

became more accurate in matching their response rate to the

reinforcement rate. Finally, then the alternative to

reinforcing imagery was 14 per point, a clear preference

for the tangible reinforcer was indicated. The matching

of response rate with reinforcement rate was highest for

this condition. Trent concluded that the results suggest

that CPR is a more effective reinforcer than valueless,

or essentially valueless reinforcers, and that the efficacy

of imaginary reinforcers is less than a relatively small

sum of money.

Steffan (1971) used CPR to increase certain types of

words emitted from adult hospitalized schizophrenics. Sub-

jects were asked to repeat four sets of 50 words each.

Baseline measures were taken during sets one and four, and

one of the following five experimental manipulations was

administered during sets two and three: CPR contingent

upon emission of a plural noun; imagination of a previously

determined "least pleasurable" scene upon emission of a

plural noun; imagination of a reinforcing scene ten times

per set, on a pre-determined random basis; the experimenter

said the word "scene" following each plural noun emitted;

no feedback upon repeating sets of words. Results indicated

that only the CPR group demonstrated a significant increase

in the emission of plural nouns from pre- to postbaseline


Epstein and Peterson (1973) studied the relative

effects of CPR and covert sensitization. Twenty-two college

students were randomly assigned to two groups, and during a

baseline phase subjects in both groups were asked to ver-

bally emit, on cue, any number between 1 and 100. During

a visualization-training phase, subjects in both groups were

trained to imagine one standard reinforcing scene and one

standard aversive scene, based upon reported preferences

on the RSS. During the treatment phase, subjects in one

group were self-presented an imaginary reinforcing scene,

cued by the experimenter contingent upon emitting any number

ending in 1-3, and self-presented imaginary an aversive

scene, cued by the experimenter upon emitting any number

ending in 7-9. The consequences were reversed for the

other group. Because of the possibility that the cue-word

"reinforcement" might be intrinsically reinforcing, as

suggested by the findings of Wish, Cautela, and Steffen

(1970), the pleasurable and aversive scenes were cued by

the words "Scene A" and "Scene B," respectively.

The results showed that contingent exposure to pleasur-

able or aversive imaginary stimuli significantly changed

the rates of target responses. Furthermore, the frequency

of emitting a number that was followed by CPR was increased

in comparison to the emission of numbers that was followed

by aversive imagery.

In a replication study, Yager, Pace, and Tepper (1975)

failed to support the original findings of Epstein and

Peterson (1973). The authors suggested that perhaps a

valid null hypothesis had been falsely rejected in the

original study and pointed to failure to encourage cog-

nitive association between behavior and imagined scenes

as a possible reason for failure to achieve significant

results in the replication.

Ascher (1973) also attempted to modify verbal behavior

through the use of reinforcing imagery, additionally

testing for extinction effects similar to that occurring

with the over-administration of overt reinforcers. Forty

college students were divided into four groups of 10 each,

and each subject was assessed with regard to the frequency

that he or she chose one of six pronouns, when constructing

a sentence with a supplied infinitive. After baseline

measures were obtained, the two groups receiving CPR were

asked to imagine a previously determined reinforcing scene

immediately after imagining a pronoun. One of the two CPR

groups received extinction trials following this phase.

Each CPR group received fifty pairings in the conditioning

phase, 30 of one pronoun and 10 each of two other pronouns.

A control group was instructed to imagine 50 reinforcing

scenes and then imagine 50 pronouns unpaired with the re-

inforcing scenes. A fourth group received only a baseline


Results of a postexperimental baseline indicated that

subjects chose those pronouns that were reinforced more

frequently, and furthermore, there was a correlation between

the number of reinforcers and pronoun selection. Extinc-

tion, however, was not confirmed, with the author suggesting

that the study may not have included a sufficient number of

extinction trials.

Cautela and Wisocki (1969) prior to the formalization

of CPR by Cautela (1970) attempted to test the efficacy of

positive imagery in modifying attitudes held by college

students toward the elderly. Students in two general

psychology classes completed a questionnaire designed to

assess their attitudes towards elderly people. Students

receiving positive imagery were asked to imagine a scene

that they had been injured in an automobile accident, and

being saved from death by an elderly man, two times per

day. Seventeen days later, they were asked to complete

the attitude scale again. The results indicated a signi-

ficant positive change in attitudes by the experimental

group who received CPR. No significant change in attitude

was found in the control group.

A test of the classical version of CPR in the modifi-

cation of attitudes was conducted by Cautela, Walsh, and

Wish (1971), after the formalization of the CPR procedure.

Students in two abnormal psychology classes completed a

"Likert-type" attitude scale assessing their attitudes

toward mentally retarded people. One week later, students

in one of the classes were instructed to imagine a mentally

retarded person in as much detail as possible. Students in

the other class were given identical instructions, and were

additionally instructed to imagine a pleasant scene of their

own choice immediately following the initial image of a

retarded person. Both groups were instructed to practice

at home twice per day.

Classes were not significantly different as compared

on pretest measures, and class comparisons made on post-

test measures four weeks later indicated a significant and

positive change in attitude by subjects in the CPR group.

Furthermore, no significant change in attitudes was found

between pre- and posttest measures for the non-CPR subjects.

Krop, Calhoon, and Verrier (1971) attempted to modify

the reported self concept of emotionally disturbed children

while comparing overt reinforcement to CPR. Thirty-six

children diagnosed as having various behavior disorders

were given items from the Tennessee Department of Mental

Health Self Concept Scale (TSCS), and assigned to one of

three conditions for a second administration of the same

scale. One group was presented with covert positive re-

inforcement being instructed to imagine a pleasant scene

developed from the administration of the RSS following

responses associated with positive self concept. Overt

reinforcement in the form of a token and a gum drop was

contingent upon positive responses for the second group.

A control group received no reinforcement following re-

sponses. A third administration of the scale followed so

that changes from the baseline could be evaluated.

The results showed that the CPR group was the only

group that changed significantly from the first to the third

administration. A 2 week follow-up revealed that changes

were maintained. Whether or not "self concept" per se was

improved or just the frequency of agreement with positive

self statements increased isdebatable, however, not par-

ticularly relevant to evaluating the efficacy of CPR.

Later, Krop, Perez, and Beaudoin (1973) attempted to

modify the self concept of adult psychiatric patients.

Sixty-nine hospitalized adults were given items from the

TSCS, after which each subject was asked to list three

things he could wish for. Subjects were randomly assigned

to three groups: a contingent CPR group, a non-contingent

CPR group, or a control group. Subjects in the contingent

CPR group were instructed to imagine a positive scene based

upon the listing of pleasurable events, contingent upon

positive self-statements during a second administration of

the TSCS. Subjects in the non-contingent CPR group were

prompted to do the same, except on a pre-determined, non-

contingent schedule, equivalent to the number of positive

self-statements emitted during the initial baseline adminis-

tration. No reinforcement or feedback was presented to

subjects in the control group. Pre- to posttreatment

differences were obtained through a third administration

of the TSCS.

Results indicated that the contingent CPR group demon-

strated a significant positive change in the target behavior,

as compared to the non-contingent CPR and control groups.

These results supported the previous Krop et al. (1971)

study which suggested that imaginary reinforcers are effec-

tive in increasing the frequency of positive self-statements.

Euse (1975) attempted to test CPR in modifying atti-

tudes toward the physically disabled. Twenty upper-level

vocational rehabilitation students were administered the

Attitudes Toward Disabled Persons (ATDP) questionnaire,

then were matched and randomly assigned to either a CPR or

a no-contact control group. Subjects were also assessed

on the basis of "looking-time" while observing slides of

disabled persons. Subjects in the CPR group received four

1-hour treatment sessions during which directed imaginary

contacts with disabled individuals were subsequently rein-

forced with imaginary reinforcers identified on the RSS.

Results obtained through posttreatment measures on the

ADTP indicated a significant positive shift in attitudes

in CPR subjects.

Daniels (1976) attempted to modify attitudes toward

reading using positive reinforcing imagery with school

children. Daniels also compared the relative effective-

ness of imaginary scenes of varying intensity, taken from

responses on the RSS, hypothesizing that imagery of rein-

forcers receiving higher preference ratings on the RSS would

produce greater reinforcing effects than imagery of rein-

forcers receiving lower preference ratings. All subjects

completed the Estes Scale to Measure Attitudes Toward

Reading, after which each was assigned to one of four

groups: high reinforcer preference, medium reinforcer

preference, low reinforcer preference, or control. Subjects

met in groups for 1-hour sessions for 6 weeks, during

which they paired imaginary scenes of engaging in reading

activity with scenes of reinforcement. Subjects also re-

corded the number of times they practiced scenes daily

at home.

Analysis of results failed to support the hypothesis, how-

ever they indicated a significant relationship between the

amount of homework practice and improvement in reading

attitudes for the group using scenes of low preference

value. Daniels did raise question concerning the validity

and reliability of the self-report instrument used.

Stevens (1973) attempted to test the effects of CPR and

covert behavioral rehearsal on friendly assertive behavior.

Seventy-two subjects were assigned to three groups balanced

for sex, social anxiety test scores, and school attended.

Pre- and posttest measures were obtained for all subjects

on Eysenck's PEN Extraversion Scale and the Friendly Asser-

tiveness Expectations questionnaire. Measures were also

obtained on the frequency that subjects talked to each other

during the pre- and posttesting sessions. Subjects in the

CPR group received taped instructions to imagine a series

of eight scenes in which they were to imagine themselves

in a social situation, initiating a conversation with an-

other person, which evoked a friendly response from the

imaginary partner. Subjects in the CPR group followed the

same procedure, except that an unpleasant response was

evoked by the imaginary partner. Subjects in a control

group were presented with short taped lectures on Freudian


Results indicated significant treatment effects in the

CPR group. Stevens also provided a discussion on the

problems associated with studying covert behavior, con-

cluding that valuable research can be obtained from studying

covert behavior (i.e., imagery) as an intervening variable.

A study which combined covert modeling and covert re-

inforcement in an attempt to increase assertive behavior

was conducted by Kazdin (1973) who compared a covert modeling

treatment group where subjects were instructed to imagine

situations in which a model made appropriate assertive

responses, a covert modeling plus CPR where subjects imagined

situations in which a model made appropriate assertive re-

sponse, after which they imagined that the assertive re-

sponses were followed by desirable consequences, a group

where subjects were asked to imagine the model making an

assertive response, and a no-treatment control group.

Results indicated that covert modeling increased assertive-

ness, and the addition of CPR further enhanced the treat-

ment's effectiveness.

Krop, Messinger, and Reiner (1974) attempted to test

CPR in the modification of eye contact during the adminis-

tration of an anxiety arousing self-disclosure question-

naire. During the baseline phase, an interviewer asked

subjects questions designed to evoke anxiety, while looking

the subject in the eyes. Afterwards, subjects randomly

assigned to the CPR group were instructed to imagine a

pleasant scene following five or more seconds of eye con-

tact. Subjects randomly assigned to two control groups

received either no-contact during the experimental phase,

or non-contingent CPR on a predetermined schedule. A

postexperimental baseline indicated that subjects re-

ceiving contingent CPR significantly increased eye contact

during the anxiety-arousing interview; however, a 3-week

follow-up with a fourth administration of the same question-

naire indicated the increases had not been significantly


Nigl (1975) successfully applied CPR to the acquisition

of dental hygiene behavior. Seventy-two subjects were

administered a dental examination and randomly assigned to

one of four groups: CPR, covert negative reinforcement,

suggestions, and no-treatment. Treatment sessions were

conducted on five separate occasions over a 3-week

period. Results of the posttreatment analysis generally

supported the hypothesis that CPR and CNR subjects would

use dental floss significantly more often than control group

subjects. No significant differences were obtained between

treatment groups based upon posttreatment dental exams;

however, the author states that lack of significance on

this variable may be explained by the inadequacy of the

dental exam scoring system and the fact that training in

the proper use of dental floss was not included in the

treatment. A 3-month follow-up indicated the superior-

ity of CPR over covert negative reinforcement in maintaining

the use of dental floss over a period of time.

Manno and Marston (1972) compared the effects of CPR

and covert sensitization in the treatment of obesity.

Forty-one people interested in losing at least 15 pounds

and not currently under a physician's care responded to an

advertisement in a university newspaper. All subjects

completed a series of questionnaires, including a personal

history and an eating habits rating scale, deposited $15 to

insure attendance at all sessions, and were randomly assigned

to one of three groups: a CPR group, a covert sensitization

group, and a minimal contact control group. Each group met

twice per week, for 4 weeks. Subjects in the CPR group

were instructed in imaging scenes of avoiding high frequency

foods, and to pair such images with images of receiving

compliments from friends and registering less weight on a

scale. Subjects in the covert sensitization group received

imaginary pairings of eating frequently abused foods with

resultant imaginary nausea. Subjects in the minimal contact

control group met for the same number of sessions, however, they

only discussed the food data sheets and problems of


Results showed that both the CPR and the covert sensi-

tization groups demonstrated significantly greater weight

loss than the minimal contact group, both after treatment

and at a 3-month follow-up. The weight loss demonstrated

by the CPR group was not significantly greater than the

covert sensitization group.

The treatment of obesity with CPR was also studied by

Brunn and Hedberg (1974). Sixty female college students

who were at least 10 pounds overweight were matched in

blocks of four each, based upon the percentage they were

overweight, and were randomly assigned, within each block,

to one of four groups: CPR prior to the evening meal; CPR

after the evening meal; a control procedure before the

evening meal; and a control procedure after the evening

meal. Subjects in the CPR group were instructed to imagine

avoiding eating a snack food, followed immediately by the

cue "reinforcement," for the CPR scene. Subjects in the

minimal contact control groups discussed fad diets, exer-

cise, and general weight loss issues, either before or after

evening meals. Weight loss for both CPR groups and the

after meal minimal contact control group was significant.

Furthermore, subjects treated by CPR lost significantly

more weight than subjects in both control groups. Time of

treatment, before or after meals, showed no functional

significance, for the CPR groups.

Flannery (1972) compared the effectiveness of an

imagined versus an in vivo fear stimulus in treating rat

phobia with CPR. In this study 45 rat phobic

student nurses subjectively rated their fear of rodents on

the Behavioral Avoidance Test (BAT), after which they were

randomly assigned to one of two treatment groups or a con-

trol group. Students in the group receiving the standard

CPR procedure were instructed to imagine themselves suc-

cessfully engaging in behavioral responses required by the

BAT, immediately followed by the production of an imaginary

reinforcer. Students in the group receiving in vivo CPR

were prompted to engage in reinforcing response required

by the BAT. Subjects in the control group merely engaged

in discussion regarding their fear of rats. Results of a

posttest BAT, and the administration of two subjective

anxiety scales, indicated that the two treatment groups

improved significantly while the control group experienced

no significant change.

Marshall, Boutillier, and Minnes (1972), in comparing

the relative effects of systematic desensitization and CPR,

used reinforcing imagery successfully in the treatment of

snake-phobic subjects. Treatments were group-administered

using identical hierarchies of approach behavior. Both

procedures proved to be equally effective in increasing

approach behavior and in reducing anxiety as measured by


ladouceur (1974) compared CPR and a "reversed-CPR"

procedure in the treatment of rat phobia. In the "reversed-

CPR" procedure, the reinforcing image was presented prior

to the image of the target behavior scene. Comparison of

both procedures, in relation to a control group, on pre-

and posttreatment measures on the BAT, showed both pro-

cedures to be equally effective in reducing rat avoidance

behaviors. Ladouceur claimed the results of the study contra-

dict an operant conditioning explanation of CPR and support

a reciprocal inhibition explanation. However, Kearney

(1976) concluded that Ladoucer's findings do not necessarily

dispute the operant conditioning explanation of CPR, and at

best only support the possible conclusion that reciprocal

inhibition is effective as a second procedure.

Hurley (1976) attempted to compare the relative effi-

cacy of standard CPR noncontiguous CPR, and imaginary

exposure without reinforcement in the treatment of snake

phobia. Subjects in the CPR group were instructed to

imagine a series of snake-approach scenes, each followed by

reinforcing imagery. In the non-contiguous CPR group, an

attempt was made to eliminate the contiguity between the

response and reinforcing scenes by instructing subjects to

imagine consecutively two positive-imagery scenes, followed

by four target-response scenes, followed by two positive-

imaginary scenes. Subjects in the covert exposure group

were presented with neutral scenes instead of pleasant

scenes. Subjects in the attention-placebo group received

treatment similar to the CPR subjects, except that non-

snake anxiety scenes were substituted for the snake approach


Results indicated that both treatment approaches sig-

nificantly reduced snake phobia as compared to the atten-

tion-placebo group, as measured by BAT scores and self-

reported fears. Furthermore, treatment approached did not

differ among themselves. Trent (1977), however, raised

question concerning the adequacy of Hurley's experimental

procedures in eliminating contiguity in the CPR group and

in providing for "neutrality" of scenes in the covert

exposure group, concluding that methodological problems

precluded a fair test of CPR.

Kearney (1976) compared CPR and covert sensitization

in the treatment of fear of laboratory rats. Seventy-five

rat phobic female high school students were assigned to one

of five groups: systematic desensitization, CPR, CPR and

relaxation, imagery only, and no treatment. Subjects in

the four treatment groups received five sessions of treat-

ment appropriate to the group. Pre- and posttreatment fear

was assessed on three self-report measures and three be-

havioral approach measures. An analysis of fear reduction

indicated that although the covert sensitization, CPR, and

CPR and relaxation groups showed a greater behavioral change

than the imagery only and no treatment groups, changes did

not reach statistical significance on self-report measures.

Kearney concluded that systematic desensitization and CPR

can be equally effective. However, since CPR is a less

complicated procedure than desensitization, and has been

shown to be as effective using less time, CPR should be

the treatment of choice in most situations.

Wisocki (1971) conducted a variation of Cautela's

original procedure by presenting imagery instructions and

scenes via audio tape recording, in a group setting. Thirty-

six female college students who had experienced significant

test anxiety for at least 1 year, as established by pre-

treatment test anxiety scale scores, were randomly assigned

to an experimental or control group. Treatment in the CPR

group consisted of five sessions of 55 minutes each, during

a period of 4 weeks. Three different imagery scenes

were presented and practiced four times during each session.

The three imagery scenes consisted of: (1) imagining being

calm and relaxed while approaching the imaginary exam;

(2) imagining feeling calm and relaxed while taking the

exam; and (3) imagining doing well on the exam. After each

scene, subjects were directed to imagine a previously

determined reinforcing scene. Subjects were also instructed

to practice each set of scenes ten times each day at home,

between weekly treatment sessions.

Pre- and posttreatment scores on the Suinn Test

Anxiety Behavior Scale and the Otis-Lennon Mental Ability

Test were compared with the test scores of the matched,

no-contact control group. Results indicated a significant

decrease in anxiety for experimental subjects, as compared

to control group subjects; however, no significant change

in intelligence test scores of experimental subjects was


Guidry and Randolf (1974) successfully replicated the

Wisocki (1971) study, adding a placebo-control group. Sub-

jects in the placebo-control group were instructed to

visualize a series of reinforcing scenes. The CPR group

showed significant improvement over the no-treatment group

on all the self-report measures used. Furthermore, the

CPR group showed significantly less anxiety on two of the

self-report measures.

Kostka and Galassi (1974) employed CPR and desensi-

tization to treat anxiety in text-anxious subjects. Six

CPR subjects and 5 desensitization subjects received eight

and two group-administered treatment sessions, respectively.

A standard hierarchy was used for both treatments. Both

treatments were shown to be equally effective in reducing

test anxiety as compared to a no-treatment control group.

Smith (1974) in comparing CPR to covert sensitization

attempted to demonstrate a reduction in test anxiety in

undergraduate psychology students. Twelve previously

determined test-anxious subjects were randomly assigned

either a CPR group, a covert sensitization group, or a

control group. An analysis of pre- and posttest measures

indicated no significant change in either the CPR or the

covert sensitization groups. The author concluded that

insignificant results may have been due to experimental

procedures utilized.

In one of the few studies conducted with elementary-

age students, Lott (1975) tested the effects of CPR in the

treatment of test-anxious fifth graders. Twenty students

balanced for sex, who were determined to be text-anxious

on the basis of scores obtained on the Test Anxiety Scale

for Children, were assigned to three groups: a CPR group,

a music group, and a control group. The CPR group met

twice each week for 3 weeks, in 30-minute sessions,

during which they were asked to imagine a sequence of

separate scenes of going in to take a test, taking a test,

and doing well on a test, with each scene followed by

instructions to engage in imagery of a predetermined

pleasant scene. The music group was designed to control

for factors associated with group interaction, and merely

involved sessions of listening to music with subjects being

told that the session was directed toward music apprecia-

tion. The control group only received a test-retest of

the anxiety scale.

Results indicated that the CPR and music groups scored

significantly lower on the posttest than did the control

group, and there was no significant difference between the

CPR and music groups on posttest measures. Upon a 3-week

follow-up test, however, the CPR group demonstrated signifi-

cantly lower test-anxiety than all other groups.

Workman (1977) attempted to reinforce on-task behavior

in fourth-, fifth-, and sixth-graders. Fifteen students

were assigned to three groups. Group I was exposed to an ABAB

sequence of conditions, in which the A condition repre-

sented a baseline phase and the B condition represented a

CPR condition. Group II was exposed to an ABACA sequence

of conditions in which A was a baseline phase, B was a CPR

condition presented in a classroom situation, and the C

condition represented CPR presented in a small group situa-

tion. Group III was exposed to an ABACDE sequence in which

A was a baseline phase, B was a covert modeling procedure

presented in a classroom situation, C was covert modeling

presented in a small group setting, D was a CPR condition

presented in a small group setting, and E involved CPR

administered on an individual basis. On-task behavioral

measures were obtained through classroom observations.

Results of the investigation indicated no clear effect for

any of the CPR or covert modeling conditions. Lack of sig-

nificant findings, however, may have resulted from a number

of procedural factors which are discussed by the author.

Shefrin (1976) tested CPR in the maintenance of effort-

ful behavior in children, ages 7-9, indicating that most

previous research with CPR had involved adult subjects and

target behaviors that were rather effortless tasks. Effort-

ful behavior in this study was measured by a crank-turning

task. Sixty children from an inner-city elementary school

were assigned to one of three treatment conditions. Sub-

jects in the CPR condition received training in the develop-

ment of pleasant imaginal scenes, after which they were

instructed to imagine the pleasant scenes whenever they

turned a crank a predetermined number of times. Subjects

in the imagery training control group also received imagery train-

ing, however, they were not instructed to engage in pleasant

imagery during the crank-turning task. Subjects in the

no-reinforcement control group were not given imagery

training and were not instructed to use any imagery during

the crank-turning task.

Results of this study failed to show that the CPR group

outperformed the two control groups. Shefrin discussed

several factors which may have contributed to the failure

of this study to demonstrate CPR effects including: (1) the

possibility that the chosen reinforcers may not have

sufficient value to accelerate performance on effortful

tasks as compared to effortless behaviors; (2) since the

subjects were told that the task was a "game," the task may

have been so inherently reinforcing that inherent rein-

forcers may have overshadowed the effects of imaginary

reinforcers; (3) CPR subjects may have been so engrossed

in the task that they may have not employed the reinforce-

ment procedure.

Studies Specifically Investigating Variables Related
to the Presentation of Directed
Positive-Reinforcing Imagery

Only a few studies have given major attention to in-

vestigating parameters of therapeutic gain in the applica-

tion of positive imagery. Furthermore, it appears that

little or no attention has been given to investigating

variables specifically related to the quality of imagery

description in the verbal presentation of positive imagery.

Kendrick (1974) using the circle-size estimation task

investigated the influence of partial and continuous

schedules of reinforcement on the rate of target behavior

acquisition and resistance to extinctions, in a CPR para-

digm. Following baseline circle-size estimations 36 subjects

were assigned to one of three groups: a continuous covert

reinforcement group, a partial covert reinforcement group,

and a no reinforcement control group. Comparisons were made

between baseline phase estimates, acquisition phase esti-

mates, and extinction phase estimates. Results indicated

that subjects in both the partial- and continuous-schedule

groups demonstrated greater changes during the acquisition

phase than did those in the no reinforcement group. Sub-

jects in the partial schedule group made greater circle-size

estimations during the extinction phase than did subjects

in the no-reinforcment group, as well as the continuous

reinforcement group.

Pace, Ochiltree, and Yager (1975) studied the effects

of imagery selection and multiplicity of images upon covert

positive reinforcement. The variable involving method of

imagery selection was studied on three levels: RSS selec-

tion of images, individual subject-selection of images, and

standardized selections of positive images chosen by the

experimenters in advance. The multiplicity variable in-

volved two levels: the presentation of single images and

the presentation of multiple images. Forty-six subjects

were randomly assigned to a control group or one of six

treatment groups (based upon a 3 x 2 factorial design).

Results indicated no significant differences between the

control and the six treatment groups, the absence of sig-

nificant results being attributed to two possible considera-

_tians, questioning of the theoretical sarnumptions-of --t .--

covert positive reinforcement procedure, and methodological


Baron (1975) examined the parameters of quality of

reinforcing imagery and duration of reinforcing imagery

using a two-choice key-pressing task with 40 undergraduate

students. Subjects were assigned to one of four conditions:

high quality imagery vs. low quality imagery; long duration

vs. short duration imagery interval; overt (point accumula-

tion) reinforcement vs. overt plus imagery reinforcement;

and no reinforcement. Reinforcer quality was based upon

either extremely high or extremely low ratings on a rein-

forcer survey measuring self-reports on the ease, clarity,

and pleasurableness of specific reinforcer items. Short

duration imagery was presented for 5 seconds, and long

duration imagery was presented for 25 seconds. Relative

response rates were measured for subjects on the basis of

key-choices indicating their preferences between each al-

ternative in each condition.

Results indicated that a high quality image produced

significantly higher response rates than a low quality

image. A pleasant image presented for a longer duration

did not produce a significantly higher response rate than

the same image presented for a shorter duration. Finally,

combined imaginary reinforcement and overt reinforcement

did not produce a higher response rate than overt rein-

forcement alone.

Pace (1976), in comparing the effectiveness of CPR to

covert sensitization, attempted to determine the most ef-

fective manner of selecting positive and aversive images.

She also studied whether encouraging a cognitive association

between the target behavior and the content of the positive

or aversive scene would increase conditioning effects.

Forty undergraduate psychology students were adminis-

tered a pretreatment test involving verbally emitting 50

random numbers between 0 and 100, after which they were

randomly assigned to one of four groups. During the treat-

ment phase, subjects in the first group were again instructed

to randomly emit 50 numbers on a timed interval. After

numbers ending in 2, 4, or 5, subjects were cued to engage

in a self-selected pleasant scene, and after numbers ending

in 8, 9, or 0, subjects were cued to engage in a self-

selected aversive scene. They were also asked to include

in the scene content the number emitted preceding the scene

(e.g., if the number was 5, the subject may have imagined a

pleasant scene of a party with five close friends). Sub-

jects in the second group followed the same procedure,

except that no such cognitive association was made. Sub-

jects in the third group were assigned scenes determined

to be pleasant or aversive by the experimenter, and were

asked to include an association with the number emitted.

Subjects in the fourth group were also assigned scenes, how-

ever, theywere not asked to include an association.

Pre- and posttest comparisons indicated no significant

effect upon frequency of number emissions as a result of

manner of imagery selection, encouragement of cognitive

association, or use of CPR as opposed to covert sensitiza-

tion. The author indicated an insufficient number of con-

ditioning exposures, an insufficient time interval to conjure

imaginary scenes, and cognitive interference with imagery

production as possible methodological problems precluding

experimental effects.

It appears that little or no research attention has

been given to the investigation of variables directly re-

lated to the verbal presentation of positive-reinforcing

imagery, particularly variables related to the quality of

the directed imagery description. This review located only

one study that included any reference to descriptive en-

hancement in the verbal presentation of positive reinforcing


Evans (1976), in testing the effects of prior imagery

training with high and low imagery-ability subjects, in-

cluded the verbal enhancement of imagery descriptions, as

a component of his imagery training procedure. Fifty-six

adult in-patients of a Veterans Administration Hospital, on

the basis of imagery-ability scores obtained on the Sheehan

Short Form of the Bett's Questionnaire Upon Mental Imagery,

were assigned to four groups: high imagery ability with

imagery training; high imagery ability without imagery

training; low imagery ability with imagery training; and

low imagery ability without imagery training. The two

groups receiving imagery training met for 10 sessions,

during which they received massed-practice in forming

images, the descriptions of which were verbally enhanced

by the experimenter, and subjects' reports of vivid visual-

izations were verbally reinforced by the experimenter (i.e.,

saying "good"). Verbal enhancement appears to have been

attempted through the inclusion of multi-sensory (i.e.,

visual, auditory, cutaneous, kinaesthetic, gustatory, ol-

factory, and organic) references to suggested imagery stimuli.

A posttraining administration of the same imagery question-

naire revealed that subject-reported imagery ability was

significantly higher for subjects who received imagery

training. In a following phase of the same study, the CPR

treatment procedure was applied to the conditioning of

circle-size estimations.

An analysis of pre- and post-treatment circle-size

estimations revealed that reinforcement effects for the low

imagery-ability subjects who received imagery training, were

as great as the two high-imagery groups. Reinforcement

effects for the low imagery-ability subjects who did not

receive imagery training were significantly lower than those

demonstrated by the other three groups.

Descriptive enhancement of imagery, however, was not

studied as a specific variable, and remained an unfactored

component of a more-inclusive imagery-training procedure,

presented separately from the actual CPR procedure. Thus,

it appears that the descriptive enhancement of positive-

reinforcing imagery has not been investigated as a potential

and specific affector of therapeutic gain.

Descriptive Enhancement and Literary Imagery

There appears to have been no previously published

quantitative research directly investigating the functional

relationship between qualitative differences in verbal de-

scriptions of directed positive-reinforcing imagery and

therapeutic gain. Although there have been limited findings

in the psychological literature supporting the position that

images are sensory responses conditioned to word-stimuli, it

appears to have been the field of literary arts that has

given a substantial amount of attention to image-evoking

properties of verbal or written (subvocal) descriptive pre-


According to Staats (1968), sensory responses evoked

by an object, transfer to verbal stimuli occurring in con-

tiguity with that object. The idea that images are condi-

tioned sensory responses for which words function as con-

ditioned stimuli is supported by the findings of Pavio,

Yuillee, and Madigan (1968), who, in investigating the

function of images as mediators of recall in verbal learning,

found that concrete nouns are particularly effective stimuli

for the arousal of sensory images; Pavio et al. (1968)

attributed this effect to the consistent association of

concrete nouns with specific objects. Singer (1971), in

citing Pavio's work in verbal and visual memory coding

systems, concluded that the advantages of the visual (and

other sensory modality) imagery system in facilitating the

recall of complex detail, reinstating emotion, and in com-

municating experience, accounts for the fact that much great

literature is characterized by a high degree of concrete


A general purusal of literary criticism indicates that

literary critics have, for some time, recognized the impor-

tance of imagery with regard to its crucial role in the

comprehension and enjoyment of literature (Friedman, 1953).

Bacon and Breen (1959), addressing the importance of the

imaginative experience provided by literature, indicated

that writers have been called upon to become literalistss of

the imagination" (p. 5), and that the function of literature

is to "create vivid images that may develop in the reader

a disposition to accept the images as physical reality" so

that "what was first imagined becomes at last directly

sensed" (p. 6). Read (1956) described "ideal poetry" as

"pure imagery," and criticized that writers often cloud the

poem with metaphors and similies, and sentiments and ideals,

until finally the essential imagery is completely obscured.

Hester (1967) emphasized the importance of the quality of

"sensa" (p. 119), that is, the capability of the language

to generate images that evoke sensations relevant to all

sense modalities, in the production of pleasure-arousing


Although it appears that critical opinion in the field

of literary arts recognizes that the production of imagery

is a central function of poetry and prose, Sheehan (1972)

concluded that most analyses of imagery in literary materials

have been thematic or subjective, and only a few quanti-

tative studies of interest to psychology have been accom-

plished. Fogle (1949) in analyzing the breadth and frequency

of sense-imagery (sensa) in the works of two great literary

artists, John Keats and Percy Shelly, performed painstaking

frequency counts of references to sense-imagery, in collec-

tions of their respective works. Fogel's data indicated

both artists made use of visual, auditory, tactile, ol-

factory, gustatory, organic (proprioceptive), kinesthetic,

and mortoric imagery. Furthermore, there existed com-

paratively little significant disparity with regard to the

frequencies found in each category, with the exception that

Shelley used almost twice the amount of visual imagery and

almost six times the amount of mortoric imagery, than did

Keats. Spurgeon (1935) counted the number of images used

by Shakespeare in comparison with those employed by a number

of his contemporaries. This study revealed that Shakespeare

not only used much more concrete imagery, but that he made

especially greater use of taste, smell, and touch. Friendman

(1953) indicated that such analyses have shown that Browning's

imagery was mostly tactile. Dudley (1928), emphasizing the

importance of multi-sensory imagery, devised a relatively

extensive classification for literary imagery based upon

six sense modalities, and sub-categories of each, based upon

qualitative and quantitative aspects of each modality.

Psychological researchers have categorized all possible

sensory modalities as contributory components of imagery

(Lindauer, 1973; Pavio and Yarmey, 1966), and Cautela

(1973) suggested that subjects should try to imagine rein-

forcing scenes in several sense modalities in order to

heighten the effect of reinforcing imagery. However, there

appears to be no psychological research providing direct

evidence for such a relationship.

The search for literary ideals cumulated during the

second decade of this century in an Anglo-American literary

movement known as imagismm." The "Imagists" represented a

revolutionary group of poets including T. E. Hulme, Ezra

Pound, and Amy Lowell, whose literary ideals included the

presentation of poetic imagery through the exact rendering

of hard and clear particulars and avoiding vague generali-

ties (Hughes, 1960). Dudley (1928) concluded that the value

of the image to literature depends upon the clarity or

distinctness of its presentation. He also suggested that

unless the writer prepares the reader in advance, images

should always be presented with fidelity, that is their

truth to nature. Dudley contended that the power to arouse

emotions is essential to literary art, and that unless the

experience is vivid, the emotion cannot be strong. "The

emotions aroused by a poem are dependent on the images of

the poem. Insofar as they are clear and faithful, they

arouse emotions that are clearly felt. Insofar as they are

vague and confused, they arouse emotions that are vague

and confused. Insofar as images are lacking, the emotions

are also lacking" (p. 106-7). Bacon and Breen (1959)

stated that, "Even a poem or a play or a story about a

state of confusion must have clarity in the vision of con-

fusion it presents. That is, ambiguity which results simply

from inept language must be charged against the writer"

(p. 188). Richards (1925), also taking the position that

sensory quality, clarity, and the fullness of detail of the

images evoked by a poem are important to its experience,

concluded that the exact relationship between pleasure and

the presence of literary imagery is not an obvious and simple


In apparent absence of specific psychological investi-

gation into the functional relationship between variances

in verbal description and the relative effects of positive-

reinforcing imagery, there appears to be a relatively sig-

nificant amount of interest in such relationships in the

field of literary arts. Commentary and critical opinion

favoring the position that the evoking of imagery is central

to production of pleasurable literature, in general,appears

to support a number of literary variables as being important

to the stimulation of pleasure-arousing imagery. Of these

literary variables, descriptive references to multi-sensory

modalities and the rendering of descriptive clarity and

fullness of detail appear to be generally considered most

important to the resulting subjective experience.




A total of 342 fifth-grade students assigned to 12

fifth-grade classes at four elementary schools in Marion

County, Florida, were initially selected to participate

in this study. Four elementary schools were selected

on the basis of the principals' and guidance counselors'

willingness to participate in the study, and the profes-

sional qualifications of the guidance counselors. Two of

the schools contained four fifth-grade classes, one of

the schools contained six fifth-grade classes, and the

other school contained three fifth-grade classes. From

each of the four schools, three fifth-grade classes were

randomly selected for participation. Of the 342 students

initially selected, 12 were subsequently excluded from

the data analysis on the basis of having been previously

diagnosed as emotionally handicapped, learning disabled,

or cerebral palsied. An additional 50 students were later

excluded from the data analysis on the basis of excessive

(three or more) absences from the five experimental treat-

ment sessions, or absence from pre- or posttesting sessions.

Thus, a final total of 280 students provided data for this


The final sample of 280 fifth-graders serving as

subjects in this study consisted of 51 percent male, 48

percent female, 93 percent Caucasian, 6 percent Negro,

and 1 percent Oriental. Race percentages in the final

sample did not differ from the initial selection, and the

sex ratio varied only 3 percent in favor of females.

Students in the sample ranged in age from 10 to 12 years,

with a mean age of 10.55 years.

Elementary school guidance counselors conducted all

testing and treatment sessions at their respective schools.

Each counselor met the following professional requirements:

(1) employed as a full-time elementary school counselor;

(2) certified as an elementary school counselor by the

Florida State Department of Education; and (3) had

completed a minimum of at least 2 full years of counseling

experience with elementary school children.

Design of the Study

The experimental design used in this study (see Figure

1) was a six-group modification of the Solomon Four-Group

design (Campbell and Stanley, 1963). Two control and four

experimental groups were respectively numbered and desig-

nated: (1) no-treatment control: pretest and posttest;

(2) no-treatment control: posttest-only; (3) non-

enhanced imagery: pretest and posttest; (4) non-enhanced

imagery: posttest-only; (5) enhanced imagery: pretest

and posttest; and (6) enhanced imagery: posttest-only.

Using a table of random numbers, three fifth-grade

classes were randomly selected from the fifth-grade classes

at each of the four schools. From the resulting pool of

12 fifth-grade classes, two were randomly assigned to each

of the six groups designated in the experimental design.

The study ran a total of six weeks. During this time

the school counselors administered the four phases of this

study: (A) administration of a pre-experimental reinforcer

survey; (B) administration of an arithmetic computation

pretest; (C) administration of five experimental treat-

ment sessions; and (D) administration of an arithmetic

computation posttest. Each of these phases was conducted

during each counselor's regularly allotted 20-minute class-

room guidance period.


This study tested the following null hypotheses con-

cerning the reinforcing effects of directed positive

imagery, the relative reinforcement effects of enhanced

and non-enhanced directed positive imagery, and the

interactive effects of pretreatment overt exposure (pre-

testing) to the specific target-behavior task.

Null Hypothesis 1: There will be no significant

difference between pretest and posttest measures of arith-

metic computation performance for experimental Group 3,

receiving non-enhanced directed imagery reinforcement.

Null Hypothesis 2: There will be no significant

difference between pretest and posttest measures of

arithmetic computation performance for experimental Group 5,

receiving enhanced directed imagery reinforcement.

Null Hypothesis 3: There will be significant

difference between posttest measures of arithmetic computation

performance for experimental Group 3, receiving non-enhanced

directed imagery reinforcement, and control Group 1,

receiving no treatment.

Null Hypothesis 4: There will be no significant

difference between posttest measures of arithmetic computation

performance for experimental Group 5, receiving enhanced

directed imagery reinforcement, and control Group 1,

receiving no treatment.

Null Hypothesis 5: There will be no significant

difference between posttest measures of arithmetic computation

performance for experimental Group 3, receiving non-enhanced

directed imagery reinforcement, and experimental Group 5,

receiving enhanced directed imagery reinforcement.

Null Hypothesis 6: There will be a significant

difference between pre- to posttest gain measures of arithmetic

computation performance for experimental Group 3, receiving

non-enhanced directed imagery reinforcement, and control

Group 1, receiving no treatment.

Null Hypothesis 7: There will be no significant

difference between pre- to posttest gain measures of arithmetic

computation performance for experimental Group 5, receiving

enhanced directed imagery reinforcement, and control Group

1, receiving no treatment.

Null Hypothesis 8: There will be no significant

difference between pre- to posttest gain measures of arithmetic

computation performance for experimental Group 3, receiving

non-enhanced directed imagery reinforcement, and experi-

mental Group 5, receiving enhanced directed imagery


Null Hypothesis 9: There will be no significant

difference between posttest-only measures of arithmetic

computation performance, for experimental Group 4, re-

ceiving non-enhanced directed imagery reinforcers, and

control Group 2, receiving no treatment.

Null Hypothesis 10: There will be no significant

difference between posttest-only measures of arithmetic

computation performance, for experimental Group 6, re-

ceiving enhanced directed imagery reinforcement, and con-

trol Group 2, receiving no treatment.

Null Hypothesis 11: There will be no significant

difference between posttest-only measures of arithmetic

computation performance for experimental Group 4, re-

ceiving non-enhanced directed imagery reinforcement, and

experimental Group 6, receiving enhanced directed imagery


Null Hypothesis 12: There will be no significant

difference between respective posttest and posttest-only

measures of arithmetic computation performance for

experimental Group 3 and experimental Group 4, both re-

ceiving non-enhanced directed imagery reinforcement.

Null Hypothesis 13: There will be no significant

difference between respective posttest and posttest-only

measures of arithmetic computation performance for

experimental Group 5 and experimental Group 6, both

receiving enhanced directed imagery reinforcement.


Prior to the lunch period on the first day of the

first week of the experimental schedule (see Figure 1),

classes were administered the My Five Favorite Food

Treats reinforcer survey (Appendix A) in accordance with

the instructions presented in Appendix B.

Food-treat reinforcers were chosen for this study

because foods appear to be one of the more popular treat-

choices among elementary-age children. Furthermore,

foods are particularly rich in multi-sensory stimuli (e.g.,

color, shape, scent, flavor, texture, temperature,

sounds, etc.), and it would appear that most children are

more experienced with, and share a more-common knowledge

of, the sensory components of foods as compared to many

other types of reinforcers.


14 1

jJ |_ i-

= Ta ? C
E--- al 112

L4 brj J^


- -- padiaoa&a sade2j, auai-X OT PWV pzlTrW 74E9 Xahms -

- -- assaCcr0S Ca paGa4snlrY Aaam s Xraso;UTti -----------

4!t I I" : 1 1 7:

3i '^ l!Z







z z

- O


l l








Several investigations have indicated that self-

report reinforcement surveys are relatively reliable

with children and adults (Kleinknecht, McCormick, and

Thorndike, 1973; Atkins and Williams, 1972). Cautela

(1977) developed the Children's Reinforcement Survey

Schedule (CRSS), a children' version of his earlier

adult Reinforcement Survey Schedule (RSS). The CRSS,

Form C (grades 4 6), however, includes only three

food items (i.e., candy, fruit, soda), on the 75-item

forced-choice-response (i.e., "dislike," "like," "like

very much") survey. Furthermore, the CRSS appears to be

more suited to the identification of individually

administered reinforcers. In order to identify group-

preferences specific to food-reinforcers, the more sim-

plified five-item, open-choice survey schedule was designed

to use in this study. The survey was designed to identify

five different food reinforcers for imaginary presenta-

tion, in order to minimize the possibility of satiation

during treatment. Cautela (1972), citing the findings

of Ayllon and Azrin (1964), cautions that it is important

to have a number of reinforcing stimuli available so as to

minimize the possibility of over-exposure, and satiation.

In administering the My Five Favorite Food Treats

survey, students were asked to list their five food-treat

choices in order of preference. Based upon subjects'

ranked-preferences, weighted-frequency counts were developed

for each food-category within the population of foods

identified in the surveys. The five food categories

receiving the highest weighted-frequences were, in order

of highest preference: ice cream, cold (including soft)

drinks, cookies, cakes, and candies.

During the second week of this study, all children in

the school district were routinely excused from school for

a regularly scheduled Spring break. During this week,

the experimenter developed and reproduced four sets of

ten standardized directed imagery treatment tapes, based

upon two levels of descriptive-enhancement for each of

the five food reinforcers identified by the reinforcer

survey. The development and content of these tapes will

be described in a following section of this chapter.

On the first day of the third week of this study,

subjects in pretested Groups 1, 3, and 5, were administered

the Marathon Math Activity (Appendix C), a 1200-item task

requiring the computation of single-digit, simple and

complex arithmetic addition problems. The development of

this pre- and post-treatment baseline task will be

described in a following section of this chapter. After

entering the classroom and insuring that each subject was

in possession of a usable standard-size pencil, the

counselor distributed one copy of the activity booklet to

each subject. Then the counselors played an audio tape

recording of the standard instructions for the task


(Appendix D). The 147 activity booklets used in this phase

were handscored, using scoring stencils, by a group of six

college nursing-student volunteers. One point was awarded

for each correct answer obtained during the 17-minute

baseline period. All test booklets were later re-scored

by another grader. The (20) activity booklets showing a

scoring-descrepancy were scored for a third time, by still

another grader, in order to confirm the correct score.

Correct scores for subjects in pre-tested Groups 1, 3,

and 5, were recorded.

During the next two and one-half weeks, subjects in

the four treatment Groups 3, 4, 5, and 6, participated in

five directed imagery treatment sessions.

Upon entering the classroom, the counselor semi-

darkened the room to reduce extraneous and potentially-

distracting visual stimuli, by drawing existing window

curtains normally used to facilitate the viewing of

instructional films. Then the counselor played one of the

ten standard directed imagery instruction tapes developed

for this study.

Each treatment tape lasted approximately 11 minutes,

and consisted of three sequential segments: (a) an intro-

duction and relaxation segment (Appendix E) lasting appro-

ximately three minutes, during which subjects were instructed

to close their eyes, assume a comfortable position in their

seats, and engage in a series of two abbreviated muscle

relaxation exercises; (b) an imaginary target-behavior

segment (Appendix F) lasting approximately four minutes,

during which subjects were instructed to imagine themselves

engaging in and successfully completing an imaginary version

of the Marathon Math Activity; and (c) an imaginary reinforce-

ment segment, lasting approximately four minutes, during

which subjects were instructed to imagine themselves

enjoying an imaginary food-treat, the description of which

was non-enhanced (Appendices G K) or enhanced (Appendices

L P), depending upon the treatment level to which they

had been assigned.

On the first day of the sixth week of this study,

subjects in all six groups were administered the Marathon

Math Activity in order to obtain post-treatment baseline

measures. Administration of this post-treatment baseline

task was conducted on the same weekday (a Monday) as was

the pre-treatment administration, and employed the same

taped standard instructions. The 418 activity booklets

used in the posttesting phase were also handscored in an

identical fashion by 12 college nursing-student volunteers.

As in the protesting phase, all activity booklets were

rescored by another grader. The (43) activity booklets

showing a scoring-descrepancy were scored for a third time

by still another grader, in order to confirm the correct

score. The majority of scoring-descrepancies found on

pre- and posttest measures involved descrepancies of less


than five points. Correct scores for subjects in all six

groups were recorded.

Selection of the Target Behavior

In investigating the comparative reinforcement effects

of enhanced and non-enhanced directed positive imagery, an

attempt was made to select a target behavior which was

indigenous to the setting and population available for


In the classroom setting, children are often required

to actively engage in a good deal of sustained and, often

to them, laborious or aversive work in order to achieve

intermediate and long range goals. The importance of main-

taining such effortful behavior in the classroom is

apparently reflected by the large number of management

studies which have attempted to modify on-task behavior

through overt reinforcement (Williams and Anandam,1973).

The requirement of engaging in persistent and effortful

behavior, with delayed or insufficient reinforcement, is

often encountered by elementary school children, particularly

in the acquisition of basic academic skills (e.g., sight-

word and phonetic memorization in reading, letter forma-

tion and fine-motor practice in writing, and addition,

subtraction, and multiplication-facts memorization in

arithmetic computation).

Considering the increasing emphasis being placed

upon the acquisition of basic skills, resulting from recent

trends in the areas of student assessment and educational

accountability, and considering the often-encountered

necessity of maintaining sustained, effortful classroom

behavior in the acquisition of basic academic skills, a

specific attempt was made to select a target behavior:

(1) requiring sustained effort on a routine, repetitive,

and possibly aversive task; (2) required in the acquisi-

tion of a basic academic skill; (3) which, for research

purposes, could be rigorously quantified; and (4) which,

for research purposes, could be relative well-defined

and easily described in a directed imagery presentation.

Consultation with a number of professional educators,

including four classroom teachers, four school psychologists,

an elementary-math program supervisor, and an elementary-

curriculum supervisor, led to the selection of drill in

arithmetic computation as meeting the selection criteria.

Development of the Instrumentation

After an extensive examination of typical commercial

and teacher-made arithmetic drill materials, a research

instrument requiring the target-behavior task of effortful

and accurate arithmetic computation, later identified to

subjects as the Marathon Math Activity (Appendix C), was

developed for this study.

Consensus validity for the instrument, as related

to the first two selection criteria, was confirmed through

a formal survey of the previously consulted professionals.

All eleven professional educators were asked to examine

specimens of the instrument along with administration

instructions and scoring guidelines, and to provide a

written opinion with regard to the instruments relationship

to the first two selection criteria. Results indicated

consensual affirmation that the required behavior involved

the effortful and accurate additive computation of simple

and complex one-digit combinations, which represented a

task required in the acquisition of basic arithmetic


A pilot version of this instrument was developed

and tested in an attempt to establish test-retest re-

liability and a performance ceiling appropriate to the

population under study, and to pilot test administration

procedures. The initial version consisted of 500 items,

representing one-digit, simple (having a one-digit solu-

tion) and complex (having a two-digit solution) addition

problems. One-digit numbers were assumed to be simpler

to visualize, and the computational operation was re-

stricted to addition to minimize the possibility of con-

fusion through mental set, both during the imaginary

exercise and pre- and posttesting. Items were randomly

composed by using a table of random numbers to draw

number combinations from a chart containing the 100

possible combinations of the numbers 0 through 9. An

attempt was made to maximize the requirement of effortful

behavior and persistence by requiring the speeded computa-

tion of problems over a 15-minute baseline period. The

baseline task was identified as the Marathon Math Activity

for the purposes of cueing, and the title content re-

presented a compromise effort to minimize novelty effects

which may have resulted from representing the baseline

task as a "game," and to minimize unusual resistance and/or

anxiety which may have resulted from presenting it as a

"test" (all pilot subjects were being prepared for an

up-coming routine 5-day administration of the Comprehensive

Tests of Basic Skills battery).

Eighty-nine students assigned to three fifth-grade

classes were selected as subjects in the pilot study.

One class was randomly selected from each of three ran-

domly selected schools, not participating in the major

study, with the resulting sample consisting of approxi-

mately 53 percent male, 47 percent female, 89 percent

Caucasian, 10 percent Negro, 1 percent Oriental, with a

mean age of 10.78 years. Subjects were administered

identical pre- and posttests at a 3-week internal,

following the same instructions presented in Appendix D,

with the exceptions that a 15-minute baseline period

was used in the pilot study, and pilot subjects were instructed

to circle, on cue from a bell, the last item completed at the

end of the 5th- and 10th- minute. This procedure was

included to allow for a more accurate computation of per-

formance rate in the event the 500-items were not sufficient

to establish a performance ceiling.

Of the 89 subjects participating in the study, 15

were later excluded from the data analysis because of

absence from pre- or posttesting and one was excluded for

failure to follow instructions. An examination of pre-

and posttest scores for the remaining 73 subjects indi-

cated that a varying number of subjects had completed

all 500 test-items by the end of either the 10th- or 15th-

minute, which precluded an examination of representative

pilot-test means and standard deviations. Computation of

test-retest correlation coefficients for scores available

at the end of the 5th-, and 10th- and 15th-minute, however,

yielded significant reliabilities for subjects not

exceeding the testing-limits. Test-retest correlation

coefficients for 5th-, 10th-, and 15th- minute scores were

r = 0.95, p = 0.00, (n = 73); r = 0.94, p = 0.00,

(n = 69); r = 0.86, E = 0.00; (n = 42); respectively.

Pilot testing, using the 15-minute version of the

instrumentation, revealed that only about 2 minutes were

actually required for pre-administration preparations and

instructions. Therefore, in order to extend the baseline

period as far as possible within the 20-minute classroom

guidance period, a 17-minute version of the instrumenta-

tion was developed for the major study.

In establishing an appropriate test-ceiling for the

final instrumentation, an examination was made to identify

the highest response-rates obtained during the first-

interval (0 5th-minute) in either pre- or post-testing.

The two highest first-interval response-rates (368 & 315)

occurred during post-testing, representing the two sub-

jects who's scores exceeded two standard deviations above

the first-interval mean score (M = 156.93, SD = 56.29,

n = 73). Since both cases exceeded the testing-limits

by the end of the second-interval (10th-minute) of post-

testing, 17-minute projected response rates of 1,103 and

1,173, respectively, were computed, based upon their

performance during the first interval.

Although both subjects were identified by their

classroom teachers as exceptionally motivated math students,

an attempt was made to confirm the adequacy of a 1200-item

version of the instrument, under conditions of concrete

reinforcement. Both subjects were administered the 1200-

item version, being told that if they did well they would

immediately receive the food-treat of their choice, and

furthermore, the one receiving the highest score would

receive a one-dollar prize. The resulting scores were 898

and 1,135, respectively, leading to the production of a

1200-item version of the instrumentation, for use in the

major study.

Development of the Directed Imagery
Treatment Tapes

Although most of the studies reviewed did not employ

the electronic recording of directed imagery instructions,

in the present study, imagery instructions, as well as

pre- and posttest instructions, were presented via audio

tape recording in an attempt to maximize experimental

control. Additionally, it would seem that in classroom or

clinical-group settings, the electronic production of

treatment instructions might be more practical, especially

in situations involving limited staff, multiple and

repeated treatment sessions, and clinical or educational

"homework" or "self-help."

The treatment tapes developed for this study consisted

of two series of five tapes each, and each tape, approxi-

mately 11-minutes in length, was comprised of three

sequential segments: an introduction and abbreviated

relaxation exercise; an imaginary target-behavior

(unconditioned response) exercise; and an imaginary rein-

forcement segment. Although the first two segments were

identical in all tapes, the third segment varied in

that one series involved non-enhanced descriptions of

five different imaginary reinforcers, and the other series

involved enhanced descriptions of the same five imaginary


Introduction and Abbreviated Relaxation Exercise

The first segment, approximately 3 minutes in length,

consisted of a brief introduction and an abbreviated

relaxation exercise (Appendix E), during which subjects

were informed that the following exercises would help

them to relax and give them some practice in using their

imagination. They were also instructed that in order to

get the "best feelings" from the exercise, they should

try hard to do exactly as instructed.

Venters (1972) concluded that prior imagery training

is not essential to CPR, and Cautela (1972) indicated

that CPR does not necessitate training in relaxation.

Additionally, formal and systematic programs of imagery

and relaxation training, prior to the application of

directed imagery procedures, are often impractical in

educational and clinical settings. Therefore, a very

abbreviated version of a relaxation training exercise

designed for children (Koeppen, 1974), was included only

in an attempt to focus subjects' attention upon the

recorded voice, minimize psycho-motor overactivity, and

to reduce resistance to partipation resulting from possible

embarrassment or novelty effect. The abbreviated relaxa-

tion exercise encouraged subjects to engage in two

imaginary scenes requiring the alternate construction

and relaxation of muscle groups in the arms and hands,

and legs and feet, followed by verbal reinforcement for

following instructions.

Directed Imagery of the Target Behavior

The second segment, approximately 4 minutes in length,

consisted of instructions (Appendix F), encouraging sub-

jects to engage in an imaginary version of the Marathon

Math Activity. At the end of this segment, subjects were

instructed to imagine that they had done an "excellent"

job of working very hard to answer correctly as many

arithmetic problems as they were able. The imagery descrip-

tion included reference to the activity-booklet color (green)

and title, as associational cues.

Although a number of studies have successfully applied

the actual overt manifestation of the target behavior to

CPR paradigm, the imaginary version of the target behavior

was applied in this study as it would appear to offer

greater practicality in most educational and clinical


Directed Imagery of the Imaginary 'einforcer(s)

The third and last segment, approximately 4-minutes in

length, consisting of either enhanced or non-enhanced

descriptions of each of the five pre-determined food-treat

reinforcers, immediately followed the imaginary target

behavior (unconditioned response) segment.

Non-enhanced directed reinforcing imagery

The non-enhanced versions of the five imaginary rein-

forcers (Appendices G K) simply instructed subjects to

imagine "the most pleasant scene" they could of consuming

their "most favorite kind" of food-treat, respective to

each of the five pre-determined food categories.

Enhanced directed reinforcing imagery

The enhanced descriptions of the five imaginary rein-

forcers (Appendices L P) specifically encouraged sub-

jects to engage in multi-sensory imagery, the content of

which was further enhanced through adjectival and adverbal

embellishment. Encouragement to engage in multi-sensory

imagery involved a combination of specific direction

(e.g., "imagine that, in your hand, you are holding a

cone, piled-high with your most favorite kind of soft,

cold, and creamy ice cream"), suggestion (e.g., "it

smells so good"), and questioning (e.g., "what color is

it?"), relevant to at least five specific imagery modalities:

visual imagery (including shape and color): olfactory

imagery; cutaneous imagery: gustatory imagery; and pro-

prioceptive (affective) imagery. Kinesthetic imagery was

implicitly encouraged to varying degrees. Imagery des-

criptions were further enhanced through the inclusion of

adjectival (e.g., "soft, cold, creamy ice cream,"

"delicious flavor") and adverbal (e.g., "it slowly melts

in your mouth") embellishments designed to provide des-

scriptive detail and clarity, appropriate to each of the

five food categories.

In an attempt to identify appropriate descriptors

common to the perceptions of fifth-graders, a second

pilot study was conducted. Eighty-one students in three

fifth-grade classes, at a school not involved in the

major study or the first pilot study, were asked to list

the most common descriptive characteristics of each

respective food-treat category. Characteristics listed

by subjects were ranked according to frequency of choice,

and the most common of each food-treat category were

selected for inclusion in reinforcer-imagery descrip-

tions. Attention to uncommon and specific characteristics

of imaginary reinforcers was directed through direct

questioning (e.g., "what flavor is it?").

In constructing the reinforcer-imagery scripts, a

pause of at least 5-seconds was inserted after each

specific direction, suggestion, or attention-directing

question. Previous studies have demonstrated the efficacy

of reinforcing scenes of various lengths. Ascher and

Cautela (1972), Cautela, Steffan, and Wish (1975), and

Wisocki (1973) have applied visualization intervals of

5, 15, and 30-seconds, respectively, in successful

demonstration of CPR. Baron (1975), in comparing the

relative effects of 5-second exposures to 25-second

exposures, found that the reinforcing image presented for

the longer duration did not produce a significantly higher

key-press response rate than the same image presented for

the shorter duration. The findings of Haney and Euse (1976),

however, suggestedthat a visualization period of at

least 20-seconds is necessary in order for a pleasant

image to produce maximal physiological arousal. Therefore,


in addition to the insertion of at least a 5-second

pause after each initial direction, suggestion, or

attention directing question, each script included two

descriptive summaries, containing an affective (physiologi-

cal) imagery component, followed by visualization periods

of 20-seconds and 30-seconds, respectively.

The reinforcer-imagery segment of each tape closed

by preparing subjects to terminate their reinforcing

imagery. An attempt was made to positively direct the

termination of reinforcing imagery, and allow subjects

to develop imaginary closure, by including a series of

four hierarchical instructions directing the transition

to a positive non-imagery state.



The purpose of this study was to investigate the rela-

tive effects of enhanced and non-enhanced directed positive

imagery upon the group reinforcement of effortful academic

behavior in a classroom setting. Major experimental

hypotheses predicted no overall reinforcement effects

resulting from the presentation of positive-reinforcer

imagery, and no comparative difference in the reinforcement

effects of enhanced and non-enhanced presentations of

directed positive-reinforcer imagery.

The Amdahl 470 computer system in the Northeast

Reional Data Center, at the University of Florida, was used

in the processing and analysis of the data developed in

this study. This system utilized the Biomedical Computer

Program (BMDP), (Dixon and Brown, 1979); the Statistical

Package for the Social Sciences (SPSS), (Nie, Hadlai-Hull,

Jenkins, Steinbrenner and Bent, (1975), and the Statistical

Analysis System (SAS), (Barr, Goodnight and Sall, 1979),

statistical programs in the analysis of the data. The

levels of significance for statistical tests are reported

at computed probability levels, with the exception of

Tukey honestly significant difference (HSD) procedures,

for which statistical significance is reported relative

to the .05 level of probability. All hypotheses were

tested at the .05 level of probability.

Null Hypotheses 1 and 2 predicted no significant dif-

ference between pretest and posttest measures of task

performance, for non-enhanced imagery treatment Group 3

and enhanced imagery treatment Group 5, respectively.

As shown in Figure 2, pretest and posttest means

(and standard diviations) for no-treatment control Group 1,

non-enhanced imagery treatment Group 3, and enhanced

imagery treatment Group 5, were 455.21 (161.22) and 465.76

(192.72), 436.40 (130.10) and 489.30 (153.30), and 382.72

(100.40) and 423.74 (103.02), respectively. There was no

significant difference between pretest and posttest measures

for no-treatment control Group 1, t (41)= -0.82, p = .42, and

calculation of a Pearson product moment correlation

coefficient indicated a significant 3-week test-retest

reliability for the instrumentation, r (n = 42) = .90,

S= .00.

The testing of Hypotheses 1 and 2 indicated a signi-

ficant difference between pretest and posttest means for

non-enhanced imagery treatment Group 3, t (41) = -5.16,

p = .00; and enhanced imagery treatment Group 5, t (45) =

-4.91, p = .00; respectively. Thus, Null Rypothesis 1

and 2 were rejected.

Null Hypotheses 3, 4, and 5 predicted no significant dif-

ference between posttest measures of task-performance for

non-enhanced imagery treatment Group 3 and no-treatment

control Group 1, enhanced imagery treatment Group 5 and

no-treatment control Group 1, and non-enhanced imagery

treatment Group 3 and enhanced imagery treatment Group 5,


Prior to the testing of Hypotheses 3, 4, and 5, an

initial pretest analysis for pretested Groups 1, 3 and

5, was conducted, utilizing a one-way analysis of variance

of pretest measures. This analysis indicated an overall

significant difference for pretest measures, F (2, 127) =

3.61, p = .03. An a posteriori analysis utilizing

multiple t-tests and a Tukey-HSD procedure, yielded signi-

ficant pretest differences between the no-treatment control

Group 1 and the enhanced imagery treatment Group 5,

t (86) = 2.56, E = .01; d = 3.36, P = <.05; and non-

enhanced imagery treatment Group 3 and enhanced imagery

treatment Group 5, t (86) = -2.18, 2 = .03. No signi-

ficant difference between Group 1 and Group 3 was indicated,

t (82) = 0.59, 2 = .56. These pretest differences resulted

in the choice of an analysis of covarience in testing

Hypotheses 3, 4 and 5.

As shown in Figure 2, an overall analysis of covarience

of posttest measures for no-treatment control Group 1, non-

enhanced imagery treatment Group 3, and enhanced imagery

0 0

II -


/ 03"
-1 ^

It~ .It:1


II 0







00 I. 10
Cu ~'s i .,%lJ 'Iro

00 N