Citation
Trust and risk-taking as a function of pre-group training

Material Information

Title:
Trust and risk-taking as a function of pre-group training
Creator:
Peterson, Sallye Baillio, 1934-
Publication Date:
Language:
English
Physical Description:
vii, 56 leaves : ; 28 cm.

Subjects

Subjects / Keywords:
Clinical psychology ( jstor )
Group psychotherapy ( jstor )
Interpersonal interaction ( jstor )
Magnetic storage ( jstor )
Observational learning ( jstor )
Personality psychology ( jstor )
Philosophical psychology ( jstor )
Psychology ( jstor )
Social interaction ( jstor )
Social psychology ( jstor )
Clinical Psychology thesis Ph. D ( lcsh )
Dissertations, Academic -- Clinical Psychology -- UF ( lcsh )
Group psychotherapy ( lcsh )
Group relations training ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Includes bibliographical references (leaves 51-55).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Sallye B. Peterson.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Sallye Baillio Peterson. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
025830232 ( ALEPH )
03375884 ( OCLC )

Downloads

This item has the following downloads:


Full Text













TRUST AND RISK-TAKING AS A FUNCTION OF
PRE-GROUP TRAINING












By

SALLYE B. PETERSON

















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















UNIVERSITY OF FLORIDA

1977

































To Michael, Moetke,
and Megan, who
didn't understand but
loved me anyway,






and






To Jay,
for all the days and hours.

















ACKNOWLEDGEMENTS



Research is not a solitary activity. I extend my deepest gratitude

to those who have supported me in this work.

To Hugh Davis, who served as Chairman, friend, and counsel in times

of despair.

To Jim Dixon, for his philosophy and inspiration.

To Nate Perry, for his support and positive criticism.

To Barry Schlenker, for his pragmatic suggestions and friendship.

To Marvin Shaw, for his encouragement and consultation.

And to Otto von Mering, for his personal interest and support.

Thanks are also due to Mark Barnett and Randy Thomas, who spent

long hours in rating the videotapes. And to Nancy Monroe,who assisted

in the painful task of data reduction.



























iii

















TABLE OF CONTENTS


Page

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

ABSTRACT . . . . . . . . . . . . . .. vi

CHAPTER I INTRODUCTION . . . . . . . . .. 1

Pretherapy Training . . . . . . . . 3

Interpersonal Trust . . . . . . . . 6

Disclosure of Painful Secrets . . . . . 8

Establishing a Norm by Pregroup Training: Modeling 10

The Risky Shift . . . . . . . . .. 12

CHAPTER II METHOD AND PROCEDURES . . . . . . . .. 14

Subjects . . . . . . . . . . .. 14

Procedure . . . . . . . . . . . 15

Dependent Measures . . . . . . . . 21

CHAPTER III RESULTS. . . . . . . . . .. . . 25

Interrater Reliability. . . . . . . .. 25

Analysis of the Modeling Tapes. . . . . ... 26

Analysis of the Subjects' Interaction in the Groups 27

Comparison of Groups and Their Models . . . .. 29

Dependent Measures. . . . . . . . .. 29

CHAPTER IV DISCUSSION. . . ... .... .. . . . 37

Hypothesis 3: The Differential Effects of Modeling 37

Hypothesis 2: Willingness to Disclose Painful
Secrets . . . . .. . . . . . . 39





iv










TABLE OF CONTENTS
(Continued)


Page

Hypotehsis 1: Interpersonal Trust . . . .. .40

The Risky Shift. . . . . . . . . ... 41

Conclusions. . . . . . . . .. . .43

APPENDIX A Interpersonal Trust Scale (Rotter, 1967) . . .. .45

APPENDIX B Time in .5-sec. Intervals in the Six Response
Categories by Rater for the Six Groups in Each
Treatment Condition. . . . . . . . ... 48

REFERENCES . . . . . . . . ... . . . . . 51

BIOGRAPHICAL SKETCH. . . . . . . . . ... . .56











































v














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



TRUST AND RISK-TAKING AS A FUNCTION OF
PRE-GROUP TRAINING

By

Sallye B. Peterson

March 1977

Chairman: Hugh Davis, Jr.
Major Department: Clinical Psychology

Interpersonal trust and the degree to which subjects are willing

to share painful aspects of themselves were investigated as a function

of pretraining using models. Seventy-two female college sophmores were

instructed to write out privately 20 of their most painful secrets.

They were divided into eighteen four-member groups in three treatment

conditions and exposed to instructions and videotape models corresponding

to an interpersonal-openness condition, a therapy-group condition, and

a control condition. Each group then had 40 minutes of interaction and

returned one week later for what they believed to be their second group

session. At that time they chose a number of secrets to be shared with

their group and also filled out an interpersonal-trust scale. They

also discussed as a group and achieved consensus on the number of

secrets each member should share, as an investigation of the risky-shift

phenomenon. Raters analyzed the videotaped first session of each group

using six response categories. Interpersonal-openness groups differed

significantly from the other groups in their verbal interaction, in


vi











the number of secrets shared, and in their willingness to allow the

experimenter to analyze the secrets. The groups did not differ on the

trust scale or in their estimates of how painful they considered their

secrets to be. Discussion to consensus led to a "conservative shift"

in number of secrets to be shared. It was concluded that groups can be

taught to behave in therapeutically relevant ways, that a climate of

trust can thus be built up in a group, and that such a degree of trust

generalizes to some extent. Implications for group therapy and sug-

gestions for future research are discussed.














































vii
















CHAPTER I
INTRODUCTION



Every therapist who has ever undertaken to lead a therapy group

has done so with the implicit assumption that situational variables

(behavior of the members in the group situation) can alter, modify, or

in some manner influence the characteristic behaviors of the members as

individuals, and that the alterations will be in a direction beneficial

to the client. Some clients do show remarkable improvement and dramatic

alterations in their ways of relating to other persons. But clinical

data reveal that clients are not always so positively affected: their

symptoms may be exacerbated, their maladaptive behaviors more deeply

ingrained and resistant to change following the group experience.

What makes the difference? Experimental data are available to show

that prediction of a subject's behavior in an interpersonal situation is

more likely to be accurate when it is based on factors in the situation

itself than on personality traits of the subject. Some personality

variables that have been studied with these results are authoritarianism

(Worthy, Gary and Kahn, 1969); trust as a personality variable (Schlenker,

Helm & Tedeschi, 1973); levels of defensiveness and anxiety (Doster,

1975); self-esteem (Otis, 1975); and need for approval (Schofield, 1975).

Experimental findings such as these suggest that the group therapist can

maximize the probability of consistently beneficial effects for group

members by careful structuring of the situation.

The concept of an externally imposed structure for a group therapy



-1-






-2-



situation is antithetical to the traditional conceptualization of the

manner in which groups function; that is, that an ambiguous, unstructured

situation is most conducive to client self-exploration and the natural

development of the group (cf. Rabin, 1970). The necessity for ambiguity

in therapy groups is derived from theories that stress the importance

of a therapeutic relationship (individual as well as group) being unlike

the client's historical relationships in which he has learned to behave

in maladaptive ways (Singer, 1970).

But the experimental literature is replete with data on the detri-

mental effects of ambiguous situations on behavior, specifically the

relationship between ambiguity and anxiety (Lazarus, 1966). A classic

study by Dibner (1958) compared the amount of anxiety experienced by

patients who were prepared for a psychiatric interview (told how to

behave and what to expect) with that experienced by unprepared patients

(highly ambiguous situation). Behavioral data, subjective reports, and

GSR data from the patients all demonstrated that unprepared patients

experienced a significantly greater degree of anxiety. The destructive

effects of anxiety induced by ambiguity are also documented by studies

of perceptual acuity (Postman & Bruner, 1948) and visual organization

and synthesis of cues (Verville, 1946). When attention is divided, or

when the individual has difficulty knowing which elements of a situation

to attend to, the perception of speech is likely to be confused and

distorted (Broadbent, 1972). And the destructive chain continues with

distorted perceptions leading to increased anxiety (Korchin et al., 1958).

On the basis of their experience alone group therapists can testify

to the evident anxiety, inappropriate behavior, and desperate looking to

the therapist for help that occur in the initial meetings of groups.






-3-



Johnson (1963) and Yalom (1966) point out two additional experiences

familiar to group therapists, i.e., that norms and behavior patterns

established in early sessions tend to persist throughout the life of

the group, and that as many as one-third of the group members drop out

during the early (1st to 12th) group meetings. In the face of accumu-

lating evidence of the deleterious effects of situational ambiguity on

the behavior of clients in therapy, and in the light of their own ex-

periences, clinicians have begun to make some attempts at preparing

clients for the therapy situation.



Pretherapy Training


Bednar et al. (1974) have reviewed literature pertaining

to the effects of pretherapy training on outcome in both individual and

group psychotherapy. The reviewers describe the group literature as

"consistent and impressive" in substantiating the effectiveness of pre-

group training on outcome. The great majority of the studies were

carried out with patient populations and used outcome measures such as

personality inventories and therapist's ratings of clients' therapy

behavior. Experimental manipulations fell into one of two categories:

(1) verbal instructions; or (2) vicarious modeling.

Bednar et al. concluded that pretherapy training is effec-

tive when the therapist "deliberately and systematically prepares

clients for group participation at a cognitive and behavioral level."

Apparently the pretraining affects the interaction of the group members

in a therapeutic manner so that the process of group therapy leads to

client improvement. However, the relationship between the type of






-4-



training chosen and the process variable studied, and between process

and outcome is unclear. For example, it is difficult to conceptualize

how listening to excerpts of "good" patient in-therapy behavior can lead

to decreased MMPI scores (Truax & Wargo, 1969), especially when no

definition of "good" behavior is given.

Whalen (1969) conducted a study on pregroup training with a popula-

tion of college students. In her study she clearly defined the variables

which she was seeking to manipulate: particular types of verbal inter-

action, e.g., personal discussion and feedback. Whalen's study was

limited to the effects of pretraining on group interaction, and no

"outcome" variables were measured.

What seem to be lacking in the research are studies in which the

critical factors in the group situation are defined and some relation-

ship is established among pretraining, group process, and outcome.

Obviously the parameters of the group situation seen as crucial to

client improvement depend on one's theoretical orientation; the critical

process and outcome variables and their relationship to factors in the

group situation are likewise dependent on one's conception of how

therapeutic change takes place; and the most effective means of in-

fluencing the crucial group parameters depends on the researcher's

conceptions in the areas stated above plus one's evaluation of the

evidence as to how learning occurs.

Critical factors in the group situation. Most modern systems of

psychotherapy are interpersonally based. Stemming from the basic theses

propounded by Harry Stack Sullivan in his interpersonal theory of

psychiatry, current psychotherapies view the origins of psychopathology

as existing in disordered interpersonal relations; thus, the appropriate






-5-



method of treatment is also seen as essentially interpersonal. Even

behavior theorists such as Wolpe regard "improved interpersonal

relationships" as among the primary criteria for evaluating therapeutic

change (Wolpe, 1969). Thus it would appear that the major characteris-

tic of the group over which the therapist must exert some control is

that of the ways in which the members behave toward each other. Such

a statement seems both simplistic in its obviousness and antagonistic

to the pusposes of group therapy: the idea that the therapist can or

should control the members' behavior. Yet, when no consistent in-

fluences are brought to bear on the members' behavior in the situation

one may observe the relative lack of structure and norm-ambiguity which

lead to heightened anxiety and low risk-taking on the part of the group

members.

What particular aspects of interpersonal relations have therapists

considered relevant? One of the most frequently cited is what might be

summarized as "interpersonal openness," or the "readiness to convey

personal information about one's life, one's feelings, and one's im-

pressions of other individuals" (Whalen, 1969, p. 510). Rogers (1959)

emphasizes the importance of such behavior for both client and therapist,

as do Jourard (1971), Gendlin (1970), and Mowrer (1964), and the en-

counter movement therapists (e.g., Bradford, 1964; Schutz, 1973). Even

analytically oriented therapists such as Franz Alexander (1946) have

recognized the importance of the client's experiencing in the present

and honestly reporting that experience in order to undergo a "corrective

emotional experience."

What are the factors which might contribute to a group situation

in which the members consistently behave in an interpersonally open






-6-



fashion? The question of whether setting of norms can facilitate

specific behaviors has been investigated with uniformly positive results.

Ribner (1974) found that groups receiving an explicit group contract

calling for self-disclosure increased significantly both their frequency

and depth of self-disclosure.

The clearest demonstration of building into a group a norm of

interpersonal openness is Whalen's study, in which the combination of a

film demonstration modeling interpersonal openness and detailed instruc-

tions of expected behavior produced significantly greater effects than

either independent variable alone. In explaining her results Whalen

relied on an attention-information format detailed by Thibaut & Kelley

(1959). The norm-sending process involved rule statements (filmed

modeling plus instructions), surveillance-evaluation (presence of other

group members), and application of sanctions (applied by the subject as

well as the other group members). Such a conceptualization stresses

the learning of complex verbal response classes, a type of learning

requiring that the subject attend to the relevant variables in the

situation rather than simply imitate a discrete response.



Interpersonal Trust


How can interpersonal openness contribute to the therapeutic effec-

tiveness of the group for its members? One of the means by which group

therapy is considered to bring about changes in an individual's habits

and attitudes is through allowing him to try out new and different ways

of relating to others in the group. Yalom (1970) has emphasized the

importance of group feedback in modifying interpersonal distortions and






-7-



clarifying interpersonal relationships. However, an individual is un-

likely either to risk such new behavior or, once having done so, to

accept the feedback he receives except where group members share a

fairly high degree of interpersonal trust.

Deutsch (1962) defined a trusting situation, as differentiated from

a gambling situation, as one in which one "has much to lose or little

to gain," probably the view most new group members hold of the therapy

group. He reasoned that one needs "considerable confidence in a posi-

tive outcome" to trust. From his research on cooperation, competition,

and trust, Deutsch defined trust in motivational terms: cooperative or

competitive motives will determine the subject's behavior in the situa-

tion; but an individualistic orientation, neither cooperating nor

competing but simply trying to maximize one's gains, resulted in the

subject's behavior being much more influenced by the specific experi-

mental conditions. In terms of Deutsch's model, if a group member

exposes himself to possible criticism, ridicule, rejection, or even

potential blackmail by other members, such behavior is indicative of

trust -- "much to lose." Whether he engages in such behavior will be

strongly influenced by situational variables; in this case the behavior

of the other group members.

Schlenker, Helm, and Tedeschi (1973) offer a definition of inter-

personal trust as "a reliance upon information received from another

person about uncertain environmental states and their accompanying

outcomes in a risky situation" (emphasis supplied). In that definition

the importance of situational variables in the operation of interpersonal

trust is even more central. And Rotter (1967, 1971) likewise emphasized

reliance upon communications from other individuals in developing his






-8-



Interpersonal Trust Scale (1967). Interpersonal trust cannot be said

to exist unless the person has some intelligible communication from

another as to how the other will respond to his actions.

In a group therapy situation, how does a member come to trust, rely

on, his fellow group members? Clearly the credibility of those relative

strangers is a primary determinant (Gahagan & Tedeschi, 1968). And

such credibility can only be based on past behavior in the group of the

group members, since in a group therapy situation members typically

know little, at least initially, of each other's behavior outside the

group. In order to maximize the degree of interpersonal trust in a

group, therefore, the members need some demonstration of what behaviors

they can mutually expect under specified conditions. Interpersonal

openness as a norm is the most likely vehicle to produce such a demon-

stration. Under such.conditions members of the group may discern over

a relatively brief period of time (1) to what degree each adheres to

group norms, and (2) how each is likely to respond to candid admissions

about one's feelings, experiences, or expectations. Such information

is clearly more likely to produce interpersonal trust among the members

than is the typical initial meeting of a therapy group in which social

chitchat, exchanges of demographic data, and other such "low-risk"

behaviors are the rule:

Hypothesis 1: Groups in which interpersonal openness is the

norm will develop a higher degree of interpersonal trust

than those in which no such clear norm is established.



Disclosure of Painful Secrets


Therapy-patients bring with them to the group certain painful






-9-



secrets which they regard as (1) extremely important to their present

problems and (2) too dangerous or embarrassing to reveal to anyone else.

With the possible exception of psychoanalysts, most therapists do not

regard the client's revelation of such secrets as crucial to his

improved functioning. Thus, "self-disclosure" to the extent of re-

vealing in detail one's greatest worries, shames, and fears has not been

shown to be an effective or even desirable goal of psychotherapy.

Although Jourard (1971) and other humanistic psychologists have

maintained that "Self-disclosure is a symptom of personality health and

a means of ultimately achieving healthy personality," research findings

have been contradictory. Cozby (1973) pointed out in his review of the

self-disclosure literature that one reason for the difficulty in com-

paring results across studies is the lack of consistency in defining

"mental health." And Allen (1973), reviewing the methodological

limitations of self-disclosure research, has listed as primary the fact

that the research is based on relatively nonintimate disclosures. Thus,

the extension of findings from such studies requires substantial con-

ceptual leaps. Particularly with regard to groups there is reason to

question the therapeutic benefits to be gained from self-disclosure,

since the majority of findings are based on self-report or dyadic

interactions. Finally, "self-disclosure" has been defined in virtually

all of the research in terms of the Jourard questionnaire or variants

of it (Allen, 1973), and the validity of these measures has yet to be

established.

Despite the lack of evidence establishing the therapeutic relevance

of actual self-disclosure thus defined, an index of the extent to which

we trust each other is our willingness to have them know our "painful






-10-



secrets" as we ourselves define them. We do not give our most powerful

weapons into hands that are likely to turn them against us. One of the

functions of the therapist in a group is to protect the individual

group member who has attained such a degree of trust in the group from

revealing such secret data at a time when the other members may not be

ready to receive it. That is, when one has reached the stage of being

willing to share, the actual sharing may be anticlimactic.

Hypothesis 2: Members of groups in which interpersonal

openness is the norm will show a greater willingness to

risk disclosing painful secrets than will members of groups

in which no such norm has been established.



Establishing a Norm by Pregroup Training: Modeling


The studies by Deutsch and by Schlenker et al., indicate that

subjects in a group situation will be likely to perform in accord with

the norms of the situation. How can such a norm be established, and

how can such complex verbal behavior as "interpersonal openness" be

"built into" group members? Whalen'sand Otis's research findings con-

cerning the powerful effects of modeling on a subject's later behavior

can be interpreted in terms of Bandura's social learning theory, which

can be summarized as follows:

Behavior acquired through observational learning is coded into

mediating images or words which are stored until appropriate environ-

mental cues suggest that performance of the observed behavior is likely

to be rewarded. The effects of modeling on the observer may be

exercised through (1) acquisition of novel responses, (2) inhibition






-11-



or disinhibition of previously learned responses, depending on the

consequences which the model is observed to experimence, or (3) facili-

tation of responses through focusing of the observer's attention on

discriminative stimuli in the environment (Bandura, 1969).

Especially in the case of vicarious learning of complex verbal

response patterns, the observer's attention must be focused on the

essential elements of the model's behavior. As Whalen pointed out, what

the subject must learn is a set of rules and regulations, rather than

discrete responses; hence the importance of orienting instructions to

the observer in facilitating the coding of verbal mediators. The effects

of observational learning are maximized by such variables as a moderate

degree of similarity of the model to the observer and the observer's

arousal and motivational levels.

There is experimental evidence in support of Bandura's theory as

it relates to observational learning of responses which deviate from

the patterns of behavior which the observer would ordinarily expect in

the situation in which the model is viewed. Apparently, the disinhibit-

ing effects of the modeling occur through (1) establishing a norm for

such behavior; and (2) providing discriminative stimuli that inform the

observer as to when the norm is in operation (Freed, Chandler, Blake &

Mouton, 1955; Grosser, Polansky & Lippit, 1951; Kimbrell & Blake, 1958;

Walters & Parke, 1964).

According to Bandura's theory of observational learning and the

experimental evidence supporting that theory, Hypothesis 3 is stated as

follows:

Hypothesis 3: Groups of individuals exposed to explicit,

detailed instructions and to a videotape of a group in which






-12-



the members model interpersonal openness will engage in

significantly more interpersonally open behavior than will

groups who do not receive such pretraining.


In summary, it was hypothesized that pretraining via modeling and

detailed instructions would lead to greater interpersonal openness on

the part of group members; that such openness would in turn result in

more interpersonal trust among the members; and that a greater degree

of interpersonal trust would also be reflected in a greater willingness

to risk disclosure of painful secrets to the group.



The Risky Shift


Studies using the Choice Dilemma Questionnaire developed by Kogan

and Wallach (1964) have often found a shift in decision-making in a

"riskier" direction following group discussion. However, the shift is

not uniform for all items (some shifting consistently in a more con-

servative direction), and stronger risky shifts are associated with

items having the least serious consequences (Dion, Baron & Miller,

1970). Explanations of the phenomenon have been many and varied, in-

cluding the decision-making model of Pruitt (1971); a social-comparison

interpretation (Jellison & Riskind, 1970, 1971); and the expected-

utility model of Vinokur (1971), recently amplified in the subjective

expected utility model of Kahan (1975).

As pointed out by Dion et al., however, a strong reservation in

interpreting the results of any risky-shift study using the CDQ is the

extent to which the tasks lack any (much less severe) consequences for

the group members. And Freedman (1969) argues cogently against the







-13-


generalizability of such studies on the basis that people are very poor

predictors of their own behavior.

Those few recent studies using designs having real consequences for

the subjects have obtained mixed results (Blascovich, Ginsberg & Veach,

1975; Blascovich, Veach & Ginsberg, 1973; McCauley, Stitt, Woods &

Lipton, 1973; Yinon, 1975). It seems reasonable at this point that

theoristic explanation should wait upon further research involving con-

sequences that are real and personal to the subjects.

The present study has as one of its aims an investigation of shift

in risk-taking as it pertains to the actual behavior of group members

in a situation where they have much to lose in the form of sharing

painful aspects of themselves with relative strangers. A shift in the

"risky" direction would presumably indicate an expected positive outcome

to the group members of sharing their secrets which would outweigh an

expected loss, the possibility of embarrassment or ridicule.
















CHAPTER II
METHOD AND PROCEDURES



Subjects


The subjects were 72 female undergraduate students enrolled in

introductory psychology courses at the University of Florida, who

volunteered to participate in order to fulfill one of the requirements

of the course. The subjects were told that two experimental sessions,

one week apart, would be held and that participation in both was re-

quired. Any subject who indicated she would be unable to attend the

second session was eliminated from the study.

As a pretreatment condition each subject was seen individually in a

private room and given written instructions as follows: "Here are 20

slips of paper and an envelope. You are to write on each slip of paper a

secret about yourself that you would not be willing to share with a group

of female strangers. For most people such secrets center around subjects

such as sex, violence, failures, loneliness, mental health, stealing,

cheating, and so on. However, these are only examples, and you should

not be limited to them. After you have finished, fold the slips of pa-

per, place them in the envelope, and seal it. Then write your name across

the flaR of the envelope like this. After you are

finished, I will lock your envelope in this file cabinet. I have the

only key, and no one will ever read what you have written unless you

decide on your own to show it to them. The envelope will be returned

to you after the study is over. When you have finished, please call




-14-







-15-



me." The subject areas for the secrets were taken from data on intimate

secrets obtained by Norton, Feldman & Tayofa (1974). When a subject

had finished the pretreatment phase she was randomly assigned to one

of three treatment conditions and to one of six groups of four within

that condition. An exception to the randomness was that no subject was

assigned to a group in which she was acquainted with the members.



Procedure


Condition 1. Interpersonal openness. Condition 1 was an attempt

to structure the six groups in that condition in such a way as to maxi-

mize interpersonal openness. The videotape model and instructions were

adapted from those used by Whalen and described in detail in her study

as "Film-detailed instructions condition."

The videotape model consisted of a twelve-minute, untitled video-

tape recording (VTR) produced for this study. Data from previous

studies (e.g., Bandura, Ross & Ross, 1963; Otis, 1975)

indicate that videotaped models are as potent as live models in elicit-

ing imitative behavior. In the VTR a group of four females conversed

after implying that they had just met. Two of the participants spent

a few seconds conveying biographical information, and then a third

member, interrupting, maintained that the group members were not using

the best method of getting to know each other. For the remainder of

the time the participants talked on a more personal level, describing

their anxieties and other feelings, and feeding back their impressions

-- whether positive, negative, or neutral, of the other group members.

An attempt was made to make the VTR group appear as similar as possible

in composition to the subject groups, e.g., in age, sex, and dress.







-16-



Similarity between subjects and models was maximized because of the

possibility that large subject-model differences might lead the sub-

jects to conclude that behavior matching was inappropriate, unnecessary,

or impossible (Goldstein, Heller & Sechrest, 1966; Hollander, 1958).

Detailed instructions to the interpersonal-openness groups were

given in written form and also by audiotape as follows: "This is a

group dynamics laboratory. It is a rare opportunity to improve your

ability to communicate with others by learning how you appear to them;

that is, how you present yourself and what impressions they form about

you. To accomplish this, it is important that you be open and honest

with other members of the group and discuss freely your immediate

feelings, attitudes, and opinions, both about yourself and others in

the group. The information used to form your impressions of the other

group members will come from their appearance, words, and behavior. We

want you to feed back your impressions to the others as soon as the

impression is formed in your mind. This will be difficult, but it is

essential to obtain the maximum benefits from the experience. The amount

you learn about yourself will depend on the extent to which you par-

ticipate in the group.

"To show you what we mean by open discussion of feelings and

feedback of impressions, we are going to show you a videotape of a

group similar in composition to yours. The group members in this tape

provide excellent examples of desirable types of group behavior. You

will note that they are open in talking about themselves. In addition,

they freely express their immediate impressions of other members of the

group, and each member makes an honest effort to accept and examine the

truth or falsity of the impression as it applies to her.







-17-



"After the tape you will form your own group, and we would like you

to try and interact in a manner similar to that of the videotape group.

No experimenter will participate in your group, but an observer will be

watching through the one-way mirror in order to record the interaction

variables we are examining.

"After you have seen the tape you may begin. You will have 40

minutes, and you will be notified when your time is up. Remember, it

is important that you openly discuss your feelings and immediate im-

pressions of others in the group, as well as invite similar comments

about yourself.

"Next week at this same time you are to return to this room for

your second group session."

Condition 2. Therapy group. Condition 2 was an attempt to have

the six groups in that condition vicariously experience a typical in-

itial meeting of a therapy group. Participants in the videotape model

were the same as those on the videotape used in Condition 1. However,

interaction among the group members in the tape was limited to the

following categories: (1) exchange of demographic data; (2) comments

about the physical environment (building, rooms, chairs, etc.);

(3) wondering aloud about how to proceed (to "get to know each other");

(4) extragroup process (Whalen's term for impersonal discussion un-

related to the group or the experiment); (5) laughter, silences, grunts,

etc.

Instructions to groups in the therapy-group condition were given

as follows: "This is a group dynamics laboratory. We are interested

in examining some variables important in the development of group

interaction. To help us to study these variables we will ask you to






-18-



interact as a group and to try to get to know each other. In order to

help you gain a clearer understanding of how a group like yours might

function, we will show you a videotape of four people in a situation

similar to the one you will find yourselves in. After the tape is over

you may begin your own interaction. No experimenter will participate

in your group, but an observer will be watching through the one-way

mirror in order to record the interaction variables we are examining.

You will have 40 minutes, and you will be notified when your time is up.

"Next week at this same time you are to return to this room for

your second group session."

Condition 3. Control group. Condition 3 was an attempt to examine

the effects of a group's simply meeting together (after viewing a

videotape not depicting group interaction) on the dependent variables

under investigation. The videotape shown was of the same length as

those shown to groups in Conditions 1 and 2 and with the same partici-

pants, but depicting simply a variety of modes of interaction (teaching,

working at a task together, drinking and eating together).

Instructions to the groups in the control-group condition were the

same as those for Condition 2 groups, except that description of the VTR

was altered as follows: "This is a group dynamics laboratory. We are

interested in examining some variables important in the development of

group interaction. To help us study these variables we will ask you to

interact as a group and to try to get to know each other. In order to

help you get in the mood for your group experience, we will show you a

videotape of people interacting in various ways. After the tape is

over you may begin your own interaction. No experimenter will partici-

pate in your group, but an observer will be watching through the one-way






-19-


mirror in order to record the interaction variables we are examining.

You will have 40 minutes, and you will be notified when your time is up.

"Next week at this same time you are to return to this room for

your second group session."

Each of the groups in the three conditions saw their VTR and re-

ceived their instructions as a group but apart from the other seventeen

groups. Each group met in a room containing four chairs located around

a circular table with each chair partly facing a one-way mirror.

Observing and videotaping was done from the adjoining room. The sub-

jects were met by the experimenter when they entered the experimental

room and were asked not to talk to the other group members until the

experiment began. When all four subjects had arrived, the experimenter

passed out the written and played the taped instructions, and then

showed the VTR appropriate to the particular treatment condition.

After the subjects had seen the tape and before the experimenter

left the room she handed out a 5 x 8 index card and a marking pen to

each subject, requesting that she write her first name on it, fold it

in half, and stand it up on the table in front of her so that the

other members could see it. The experimenter then left the room. After

40 minutes had elapsed, the experimenter re-entered the room to end

the interaction, to remind the group that they were due back for the

second session one week later, and to request that they not discuss

the experiment with their friends.

The second session. One week after the first group meetings

all groups returned for a "second session." Before the second

session began, each subject was left alone in a private room to

fill out Rotter's Interpersonal Trust Scale. After she had com-

pleted the Scale (reproduced in Appendix A),






-20-



she was handed an empty envelope, her original envelope containing her

20 secrets, and a typed instruction sheet which read as follows:

"Please take out of your envelope at random -- that means, without

looking into the envelope -- as many, from zero to twenty, of the slips

of paper as you are willing to put in the pot for discussion in the

group tonight. The actual discussion session of your group will not be

recorded tonight. Please put the slips of paper you select into this

second envelope, write your name across the front of it, and drop it in

the bowl on the table. Hold on to your original envelope and call me."

When the four members of a group were assembled the experimenter

instructed them as follows: "Before you begin your group tonight I

would like you to discuss among yourselves the number of secrets which

each member of your group should contribute to discussion. You are not

to discuss the content of the secrets but only the number, and you must

agree as a group on one number." Discussion to consensus was video-

taped. This portion of the experimental manipulation was.a check on

whether the risky-shift phenomenon was operating in these groups.

After consensus was reached, the experiment was over, and the subjects

were debriefed regarding their expectations and hypotheses about the

experimental manipulations.

Each subject was seen individually at this point and given a slip

of paper on which she was asked to rate on a scale from 1 to 7 how

painful or embarrassing her secrets were (a rating of 1 representing

"not embarrassing or painful at all" and a rating of 7 representing

"extremely embarrassing or painful"). She was then given a typewritten

slip of paper with directions as follows:

"You were told that no one would ever read your secrets






-21-



unless you decided on your own to share them. That is still

true. However, if you have no objections, I would like to

analyze the content of your secrets. This could be done by

your tearing open both your envelopes and dumping the contents

into a bag with other secrets so that I would never know to

whom a secret belonged. However, this is not an integral part

of the experiment and I want you to feel free to refuse if you

have any reservations."



Dependent Variables


The dependent measures consisted of (1) mean score for each group

on the Interpersonal Trust Scale; (2) behavioral data in the form of

mean number of secrets per group selected to share with the group;

(3) ratings of the total verbalizations of each group on a revised

version of Whalen's rating scale; and (4) differences between (a) mean

number of secrets chosen for sharing by group members prior to dis-

cussion to consensus and (b) number agreed upon by each group.

Interpersonal openness rating scale. For purposes of this study

the six broad response categories of Whalen's scheme were used. They

are described in detail below. The descriptions following each response

category were taken as criteria upon which the judges based their

ratings.

Response categories:

1. Personal discussion. (a) Personal self-disclosure: Discussion

of feelings, attitudes, and behaviors unrelated to current group inter-

action which are non-public, not ordinarily readily volunteered, and






-22-



which may make the individual vulnerable to negative evaluations from

others. Examples of actual comments categorized as personal self-

disclosure by Whalen's raters are "I'm a status seeker like a lot of

people,""[ might even try homosexuality," or "I just broke down and

started yelling and screaming." (b) Immediate feelings: Discussion of

feelings and attitudes related to current group interaction which are

relatively private, not ordinarily volunteered by most people, and

which may make the individual vulnerable to negative evaluations from

others (e.g., "I feel alienated here" or "Now I'm embarrassed").

(c) Personal questions: Questions relating to an individual's opinions,

feelings, or actions which are usually considered to be private in-

formation: "Have you always been heavy?" or "I wondered how you felt

when we were so silent."

2. Feedback. (a) Positive feedback: The conveying of an indi-

vidual's positively valenced impressions of or reactions to the comments,

appearance, or actions of another group member. This category includes

compliments, flattery, etc. Examples are "I like the way you said

that," or "You seem like an alert, intelligent person." (b) Negative

feedback: Negative evaluation of the personality, intelligence,

actions, etc. of another group member. This category includes hos-

tility, ridicule, criticism and implications that the individual is not

telling the truth. Examples of comments rated as negative feedback

are "You were a fool," "Big deal!" or "Your argument isn't very ration-

al." (c) Neutral feedback: This category includes comments about a

group member that cannot be evaluated as either positive or negative,

as well as direct advice from one group member to another. Some

examples are "You seem like the silent type" or "You ought to go up






-23-



and tell her." (d) Acceptance of feedback: Acceptance by a group

member of the positive, negative, or neutral feedback she receives,

for example, "Yeah, I guess you're right." (e) Reject feedback:

Refusal to accept the feedback about oneself provided by other group

members, for example, "No, you see it's really like this ."

(f) Request feedback: A direct request that the group or particular

members provide the subject with an evaluation of herself, for example,

"What do you think of me?"

(These first two responses classes comprised the primary types

of interaction modeled in the videotape recording for the

interpersonal-openness groups.)

3. Impersonal discussion. (a) Impersonal self-disclosure:

Nonpersonal biographical information about oneself which is either

generally accessible or readily volunteered in appropriate contexts by

most people. Examples: "I come from Chicago" or "I don't know the

outskirts of the city very well." (b) Extragroup process: Nonpersonal

topics unrelated to the experiment or the group interactions -- "All

universities don't accept correspondence courses" or "I know a girl who

is in the Navy now." (c) Impersonal questions: Questions regarding

public, easily accessible, readily volunteered information about a

group member -- "Have you ever seen the San Diego campus?" or "How many

hours are you taking?"

4. Group process. Nonpersonal discussion relevant to the experi-

ment such as comments regarding the experimental manipulations, the

purposes of the experiment, the behavior of the experimenter or par-

ticipants in the VTR, a plan of action, etc. Examples: "Are they

watching us?" or "Let's all talk at the same time -- that ought to






-24-



goof them up."

5. Descriptive aspects of communicative speech: (a) Agreement:

Any verbal indication of agreement, for example, "Uh huh," and "Yeah,

that's right." Head nods are not scored. (b) Disagreement: Any

verbal indication of disagreement, for example, "NO," and "That's not

true!" Head nods or shakes are not scored. (c) Laughter: Any amount

of audible giggling or laughing by one or more group members. Smiles

and grins are not scored. (d) Silence: This category is scored only

if the duration of the silent period exceeds 3 seconds. (e) Inter-

ruption: Scored whenever one group member stops another from talking

or when two or more individuals are speaking simultaneously.

6. Unscoreable utterances. A scoring category for those remarks

which are either inaudible, unintelligible, or too difficult to

categorize immediately.

Rating procedure. The middle twenty minutes of discussion for

each of the eighteen groups was rated independently by two trained

raters using an Esterline-Angus event recorder. The resultant scores

described total group interaction in terms of number of seconds spent

in discussion rated as falling into each of the response categories

(no ratings of individual members).

The raters were advanced graduate students in clinical psychology

who were trained on practice tapes to a criterion of 94% agreement on

a tape. After training, the raters also analyzed the three modeling

tapes in terms of time spent in each of the six response categories.

















CHAPTER III
RESULTS



Interrater Reliability


To determine whether the raters were reliable in their judgments

of the total number of seconds spent in each category of discussion,

Spearman's rank correlation coefficient was computed for the 108 pairs

of scores. (Appendix A shows the raters' scores.) The resulting test

statistic of r = .986 is highly significant (p < .001). Individual

correlation coefficients were also computed over the 18 groups for

each of the six categories. The results are shown in Table 1.



Table 1

Interrater Reliability


Response Category Spearman's r P

Personal discussion .999 .01
Feedback .955 .01
Impersonal discussion .981 .01
Group process .944 .01
Descriptive aspects of
communicative speech .964 .01
Unscoreable utterances .844 .01



Since interrater reliability was judged to be satisfactory, the

average of the two raters' scores was used in further computations.








-25-






-26-



Analysis of the Modeling Tapes


To determine whether the three 12-minute tapes used in the three

treatment conditions actually depicted the types of verbal interaction

which they were designed to model, the judges rated each of the tapes

separately. Each verbalization was rated as falling into one of the

six response categories.

The modeling tape for the interpersonal-openness condition was

designed to depict primarily the categories of personal discussion and

feedback, since those were the types of verbalizations which the inter-

personal-openness groups were supposed to model. Tapes for the therapy-

group and control-group conditions were designed to model primarily

impersonal discussion. As can be seen from Table 2, the tapes did

model the types of verbal interaction for which they were designed.

(The interpersonal-openness condition tape included some personal dis-

cussion as an example of the kind of interaction which was not to be

modeled by the subjects.)



Table 2

Analysis of Modeling Tapes: Time (in seconds) and Percent of
Total Time Spent in Each of the Six Response Categories

Inter-
Response personal "Therapy" Control
Category Openness % Group % Group %

Personal Discussion 417 58 0 0 0 0
Feedback 83 12 0 0 0 0
Impersonal
Discussion 87.5 12 680.5 95 575 80
Group Process 78.5 11 0 0 0 0
Descriptive aspects
of Communicative
Speech 54.0 07 39.5 5 145 20
Unscoreable Utterances 0 0 0 0 0 0
720.0 100 720.0 100 720.0 100







-27-



Analysis of the Subjects' Interaction in the Groups


From the forty minutes of interaction had by each group, a video-

tape of the middle twenty minutes was analyzed by the raters with each

verbalization judged as falling into one of the six response categories

(personal discussion, feedback, impersonal discussion, group process,

descriptive aspects of communicative speech, and unscoreable utterances).

The Kruskal-Wallis one-way analysis of variance was used to determine

whether significant differences existed among the groups in the three

treatment conditions with regard to the number of seconds spent in each

category. Results of the analysis showed that the groups differed

among themselves significantly in three of the response categories:

personal discussion (p < .01), feedback (p < .05), and impersonal dis-

cussion (p < .01). Table 3 shows the results of the analysis.

For each of the response categories in which a significant dif-

ference among the groups was found, the Mann-Whitney U Test was performed

to determine where the differences lay. As can be seen from Table 4,

groups in the interpersonal-openness condition spent a significantly

greater amount of their time in personal discussion and a significantly

lesser amount of time in impersonal discussion than did groups in the

therapy-group and control-group conditions, which did not differ sig-

nificantly from each other. In the category of feedback, the only

significant difference was between the interpersonal-openness and the

therapy-group conditions.









Table 3

Mean Number of Seconds, Percent of Total Time, and Kruskal-Wallis Analysis
of Variance for the Three Treatment Conditions




Condition Means and Percent

Interpersonal Therapy Control
Openness Group Group

Response Mean Percent Mean Percent Mean Percent H
Category

Pers Disc 1112.83 45.1 8.00 .3 12.58 .5 12.51**
Feedback 107.92 4.5 0.00 .0 9.92 .4 6.12*
Impers
Disc 1002.42 42.5 2178.08 90.2 2206.42 91.5 9.58**
Group
Process 59.08 2.4 44.83 1.9 28.00 1.2 .07
Descrip 116.63 4.9 165.50 6.9 140.33 5.8 4.16
Unscore 13.75 .6 16.92 .7 13.33 .6 .26
2412.63 100.0 2413.33 100.0 2410.58 100.0



*P < .05
** < .01






-29-



Table 4

Differences between Pairs of Treatment Conditions on the
Three Significantly Different Response Categories


Treatments (in Increasing
Response Category Order of the Means of the Ranks)


Personal Discussion T-G C-G I-0

Feedback T-G C-G I-0

Impersonal Discussion I-0 T-G C-G


Note: Treatments connected by a line are not significantly different
at the .05 level.





Comparison of Groups and Their Models


Figures 1 through 3 represent comparisons of the percent of time

spent by group members and by their models in the six categories of

discussion. Groups in the therapy-group and control-group conditions

most closely matched their models. In the interpersonal-openness con-

dition the groups matched their models except in the category of

impersonal discussion, where the group members spent 42.5% of their

time, compared to the models' 12%.



Dependent Measures


To determine whether the groups in the three treatment conditions

differed among themselves with regard to the five separate response

variables, the Kruskal-Wallis one-way analysis of variance was computed

for each of the dependent variables. Reference to Table 5 shows that

the groups did differ significantly on three of the dependent variables:






-30-













100 ]Models


90 E Group
Members

80


70


60


50


40


o 30


20


10


0
Pers Feedback Impers Group Descrip Unscore
Disc Disc Process




Figure 1. Comparison of models and groups in the interpersonal-
openness condition on percent of time spent in each of
the six categories of response.






-31-












100 l Models


90 Group
Members

80


70


60


50


u 40


30


20


10


0 E
Pers Feedback Impers Group Descrip Unscore
Disc Disc Process



Figure 2. Comparison of models and groups in the therapy-group
condition on percent of time spent in each of the six
categories of response.






-32-














100 I Models


90 M Group
Members

80


70


60


50


40
U


30


20


10



Pers Feedback Impers Group Descrip Unscore
Disc Disc Process



Figure 3. Comparison of models and groups in the control-group
condition on percent of time spent in each of the six
categories of response.







-33-



mean number of secrets which they were willing to share (p < .05);

mean number of group members willing to allow the experimenter to

analyze the content of their secrets (p < .01); and mean number of

secrets which the group as a group agreed to reveal after discussion

to consensus (p < .05). The groups did not differ on their individual

ratings of how painful or embarrassing they considered their secrets

to be or on their mean scores on Rotter's Interpersonal Trust Scale.








Table 5

Differences Among the Treatment Conditions on
the Five Dependent Variables


Interpersonal Therapy Control
Dependent Openness Group Group
Variable Mean Rank Mean Mean Rank Mean Mean Rank Mean H

Secrets shared 14.33 7.16 7.00 2.96 7.17 2.78 7.38*
Analysis of
secrets
permitted 15.00 4.00 7.30 2.16 6.17 2.0 10.50**
Painfulness of
secrets
(1-7) 10.50 4.80 10.25 4.68 7.75 4.53 .98
Trust-Scale
Scores 6.22 66.75 12.25 69.53 6.92 64.66 3.02
Secrets chosen
by consensus 14.00 2.70 8.50 1.00 6.00 .37 7.67*



*p < .05
**p < .01







-34-



For each of the dependent variables on which the groups differed

significantly, the Mann-Whitney U-Test was performed for each pair of

treatment conditions. As Table 6 shows, groups in the interpersonal-

openness condition were willing to share a significantly greater number

of their secrets with other members of their group, and more of the

group members were willing to permit the experimenter to analyze their

secrets, than were groups in the other treatment conditions, who did

not differ significantly from each other. As to the mean number of

secrets which the groups agreed to share after discussion to consensus,

the significant difference (p < .05) occurred between groups in the

interpersonal-openness condition and those in the control-group

condition.






Table 6

Differences Between Pairs of Treatment Conditions on
the Three Significantly Different Dependent Variables


Treatments (in decreasing order
Dependent Variable of the means of the ranks)

Secrets shared I-0 C-G T-G
Analysis of secrets
permitted I-0 T-G C-G
Secrets chosen by
consensus I-0 T-G C-G



Note: Treatments connected by a line are not significantly different
at the .05 level.






-35-



Table 7 is a further analysis of the "risky-shift" data. By

reference to that table it can be seen that the groups in the three

treatment conditions differed significantly in the number of secrets

which they were willing to share both before and after discussion to

consensus (p < .05). However, after discussion to consensus the only

significant difference was between the interpersonal-openness and

control-group conditions. The differences between the number of secrets

subjects were willing to share before and after discussion were signi-

ficant in each of the treatment conditions and were all in the direction

of a "conservative" rather than a "risky" direction (Sign Test).



Table 7

Comparison of Mean Number of Secrets Subjects Were
Willing to Reveal Before and After Discussion to Consensus



Interpersonal Therapy Control
Openness Group Group P

Before
Discussion 7.16 2.96 2.78 .05

After
Discussion 2.70 1.00 .37 .05

Shift 5.46 1.96 2.41

p .02 .03 .02


Note: Treatments connected by a line are not significantly different
at the .05 level.



To discover whether the groups differed among themselves with

reference to the amount of time they spent in their discussion to con-

sensus, the Kruskal-Wallis test was performed. The mean amount of time







-36-



in minutes for groups in each treatment condition was as follows:

interpersonal-openness, 6.46; therapy-group, 3.28; control-group, 10.34.

The groups did not differ significantly at the .05 level.

















CHAPTER IV
DISCUSSION



The results of the study are discussed in the order in which the

data were obtained and analyzed. The effects of modeling (Hypothesis

3) are discussed first, followed by their influence on group interaction

and the outcome variables (Hypotheses 2 and 1, respectively).



Hypothesis 3: The Differential Effects of Modeling


As Whalen and others have shown, complex verbal response patterns

can be taught to groups by the use of descriptive and exhortative in-

structions and the use of videotape models of interpersonally open

verbal interaction. The data from this experiment support this finding

and additionally indicate the need for specific instructions and un-

ambiguous models to facilitate a particular verbal behavior.

Apparently, personal discussion as defined herein is a more dif-

ficult behavior to inculcate in group members than is impersonal dis-

cussion. The data show that, even with the use of models whose

interaction included very minimal impersonal discussion, the subjects

produced more than three times as much impersonal discussion as the

models exhibited. In the groups viewing models of primarily impersonal

discussion the interaction matched the models much more closely. How-

ever, the only condition in which personal discussion predominated was

that in which the models exhibited it as the primary means of interaction.




-37-






-38-



Thus, while interpersonally open behavior is difficult to induce, the

interpersonal-openness models and instructions used in this study were

effective in producing a significantly greater amount of that behavior

as compared to the other two conditions, the therapy-group and the

control-group.

An additional essential element in effective modeling other than

the clarity of the behavior to be matched and the explicit instructions

appears to be motivation. Only groups in the interpersonal-openness

condition received as part of their instructions the statement that

"[This] is a rare opportunity to improve your ability to communicate

with others by learning how you appear to them; that is, how you present

yourself and what impressions they form about you." A motivational

factor was also present in the modeling tape in that a member who was

open was reinforced by the other group members with positive feedback.

It seems probable that the treatment effects in the present study were

exerted through Bandura's principle of disinhibition of a previously

learned response due to the positive consequences experienced by the

models. Lakin and Carson (1966) have observed that people most often

avoid engaging in personal discussion because of the fear of being per-

ceived as weak and inadequate. To counteract such a fear and disinhibit

the modeled response, the subject must have been motivated by an

expectation of positive consequences.

Thus, Hypothesis 3 is supported with the qualification that suf-

ficient motivation in the subjects to model the behavior must also be

induced.







-39-


Hypothesis 2: Willingness to Disclose Painful Secrets


Hypothesis 2 is supported by the finding that members of groups in

the interpersonal-openness treatment condition were willing to disclose

a significantly greater number of secrets to their fellow members than

were subjects in the other two treatment conditions. The fact that the

subjects did not differ across conditions in terms of how painful or

embarrassing they considered their secrets to be indicates that the same

degree of risk obtained for all the group members.

Hypothesis 2 predicted that the readiness to risk disclosing secrets

would be related to the presence or absence of a norm of interpersonal

openness. Analyzing the treatment condition in terms of the norm-sending

process detailed by Thibaut and Kelley (1959), one finds that the three

components of norm-sending were clearly present only in the interpersonal-

openness condition. The components and their counterparts in the

present study are (1) rule statements (instructions and videotape);

(2) surveillance-evaluation (presence of other group members and the

experimenter's observation from the next room); and (3) application of

sanctions (applied by the subject and the other group members in the

form of rewards for interpersonal openness; the expected reward promised

in the instructions of "improving your ability to communicate with

others ."). Additionally, the finding that groups in the inter-

personal-openness condition actually spent approximately half of their

time in personal discussion and feedback indicates that a norm of

interpersonal openness was operating.

There is also some support for the idea that a norm of interper-

sonal openness generalized outside the group. Such support is found in

the fact that a significantly greater number of members of groups in






-40-



the interpersonal-openness condition were willing to permit the ex-

perimenter to analyze the content of their secrets. The readiness to

allow a stranger to read what one had previously identified as secrets

one would not be willing to share, even when promised anonymity,

indicates at least a state variable of being open about one's self.

It is also worth emphasizing that there was no pressure on any of the

subjects to permit analysis of the content of their secrets; indeed,

one-third of the subjects in the therapy-group and in the control-group

conditions declined to allow such analysis.



Hypothesis 1: Interpersonal Trust


Members of groups in the three treatment conditions did not differ

significantly on the variable of interpersonal trust as measured by

Rotter's Scale. The means for all of the groups were well within one

standard deviation of the mean for the standardization group of females

and also for a group of 411 females at Ohio University recently studied

by Wright & Tedeschi (1975), indicating that the population from which

our samples were drawn does not diverge markedly from the norms for

this scale.

Rotter speaks of interpersonal trust as measured by his scale as

a "generalized expectancy that the oral or written statements of other

people can be relied upon." In thus describing this variable he dif-

ferentiates it from expectancies specific to the situation in which an

individual finds himself. In other research such a differentiation is

sometimes described in terms of "trait" versus "state" variables. Thus

Rotter would predict that scores on his scale would remain relatively

stable over time, since generalized expectancies are based on long-term







-41-


life experiences with parents, teachers, and peers.

As measured by Rotter's scale, then, interpersonal trust is an

attitude that would not be expected to fluctucate over such a short

period of time as was involved in the present study unless some very

dramatic change occurred in the lives of the group members. Apparently,

the experimental manipulation, while it was effective in altering the

expectancies of the subjects relative to the specific situation of the

experiment (witness the significant differences on other dependent

variables) was not potent enough to effect a change in what might be

thought of as a personality or trait variable of trust. Such an effect

as personality change might be obtained over a longer period of time

with repeated experiences in groups such as those in the present study.

One would expect that a successful therapy-group experience might pro-

duce such a shift in attitude.

Alternatively, if interpersonal trust is less narrowly defined,

the data on the subjects' willingness to share secrets and to allow them

to be analyzed by the experimenter could be taken as an indication that

trust in other persons was affected differentially by the treatment

conditions. Certainly, some degree of trust is indicated by a readiness

to disclose painful aspects of one's self to others. However, such a

conclusion is limited in its applicability until further research

demonstrates that the effects endure over time and across other groups

or individuals. Hypothesis 1 is not clearly supported by the data

obtained in the present study.



The Risky Shift


Groups in all three treatment conditions shifted significantly in







-42-



a conservative direction following discussion to consensus on number of

secrets to be shared. However, after discussion the difference between

groups in the interpersonal-openness condition and those in the therapy-

group condition disappeared. Groups in the control-group condition

were not significantly different from those in the therapy-group

condition.

The reasons for such a uniformly conservative shift are difficult

to discover. Each group member decided on her own to reveal a certain

number of secrets. Yet, apparently she was not willing to impose on

other group members a greater degree of risk than she had accepted for

herself. One could also theorize that the less trusting members of the

group prevailed; an element of distrust in the person of one reluctant

member might have dissipated the atmosphere of trust built up in the

group. In sum, being unsure of the reasons for a member or members'

reluctance to disclose her self to any degree, the other members may

have concluded (a) that they should not try to persuade her (or them)

otherwise; or (b) that perhaps that person or persons was not really

so trustworthy as they had thought. Certainly, in therapy groups a

client is often heard to say, after the session is over, that he or she

had planned to talk about such-and-such a topic, but had lost the courage

to do so when other members seemed reluctant to expose themselves.

Unlike studies using the Choice Dilemma Questionnaire, the present

study involved a real and personal risk for the group members. It would

seem that the gain to be expected from taking such a risk was, in this

case, outweighed by other factors which became operative only in the

context of the discussion to consensus. What those factors were is a

matter of speculation. However, further research involving analysis of







-43-



the videotaped discussions might possibly reveal the answers. Such an

analysis is beyond the scope of the present investigation.



Conclusions


The results of the present study suggest some important conclusions

for the theory and practice of group psychotherapy. If the data herein

can be viewed tentatively as applicable to a clinic population, we

might conclude:

1. That members of a group can be taught to behave in thera-

peitically relevant ways (e.g., relating in an interpersonally open

manner);

2. That such training can be carried out in a simple and practical

manner through the use of brief instructions and videotape models;

3. That, when members of a group engage in such interpersonally

open behavior, a climate of trust is built up wherein the group members

become willing to share with each other painful and embarrassing aspects

of their lives;

4. That, while the actual disclosure of such secrets is not

necessarily a sine qua non of therapy, the readiness to do so may be;

and

5. Finally, that trust in one's fellow group members does gener-

alize, at least to the extent in the present study of trusting the

experimenter.

Necessarily, generalization of the results of this study to a

clinic population must wait upon replication with actual therapy groups.

However, such research seems clearly justified on the basis of the data

obtained herein.







-44-



Whalen's rating scale as modified in the present study has shown

itself to be highly reliable; the raters reported no great difficulty

in distinguishing among the categories. Hence it would seem to be an

instrument of choice for further research in this area.

No conclusions can be drawn regarding the reasons for the "con-

servative shift" obtained from our subjects. The present research

fortunately included videotaping of the actual discussion-to-consensus

sessions. Analysis of the group interaction in such sessions would

seem to be a natural outgrowth of this study and perhaps have a bearing

on the interactional aspects of the risky shift phenomenon.

















APPENDIX A

Interpersonal Trust Scale (Rotter, 1967)



General Opinion Survey


This is a questionnaire to determine the attitudes and beliefs of

different people on a variety of statements. Please answer the state-

ments by giving as true a picture of your own beliefs as possible. Be

sure to read each item carefully and show your beliefs by marking the

appropriate number of your answer sheet.

If you strongly agree with an item, circle number 1. Circle number

2 if you mildly agree with the item. That is, circle number 2 if you

think the item is generally more true than untrue according to your

beliefs. Circle number 3 if you feel the item is about equally true as

untrue. Circle number 4 if you mildly disagree with the item. That is,

circle number 4 if you feel the item is more untrue than true. If you

strongly disagree with an item, circle number 5.

1. Strongly agree

2. Mildly agree

3. Agree and disagree equally

4. Mildly disagree

5. Strongly disagree

1. Most people would rather live in a climate that is mild all year
around than in one in which the winters are cold.

2. Hypocrisy is on the increase in our society.



-45-






-46-



3. In dealing with strangers one is better off to be cautious until
they have provided evidence that they are trustworthy.

4. This country has a dark future unless we can attract better people
into politics.

5. Fear of social disgrace or punishment rather than conscience pre-
vents most people from breaking the law.

6. Parents usually can be relied upon to keep their promises.

7. The advice of elders is often poor because the older person doesn't
recognize how times have changed.

8. Using the honor system of not having a teacher present during
exams would probably result in increased cheating.

9. The United Nations will never be an effective force in keeping
world peace.

10. Parents and teachers are likely to say what they believe themselves
and not just what they think is good for the child to hear.

11. Most people can be counted on to do what they say they will do.

12. As evidenced by recent books and movies morality seems on the
downgrade in this country.

13. The judiciary is a place where we can all get unbiased treatment.

14. It is safe to believe that in spite of what people say, most people
are primarily interested in their own welfare.

15. The future seems very promising.

16. Most people would be horrified if they knew how much news the
public hears and sees is distorted.

17. Seeking advice from several people is more likely to confuse than
it is to help one.

18. Most elected public officials are really sincere in their campaign
promises.

19. There is no simple way of deciding who is telling the truth.

20. This country has progressed to the point where we can reduce the
amount of competitiveness encouraged by schools and parents.

21. Even though we have reports in newspapers, radio and television,
it is hard to get objective accounts of public events.

22. It is more important that people achieve happiness than that they
achieve greatness.






-47-



23. Most experts can be relied upon to tell the truth about the limits
of their knowledge.

24. Most parents can be relied upon to carry out their threats of
punishment.

25. One should not attack the political beliefs of other people.

26. In these competitive times one has to be alert or someone is likely
to take advantage of you.

27. Children need to be given more guidance by teachers and parents
than they now typically get.

28. Most rumors usually have a strong element of truth.

29. Many major national sport contests are fixed in one way or another.

30. A good leader molds the opinions of the group he is leading rather
than merely following the wishes of the majority.

31. Most idealists are sincere and usually practice what they preach.

32. Most salesmen are honest in describing their products.

33. Education in this country is not really preparing young men and
women to deal with the problems of the future.

34. Most students in school would not cheat even if they were sure of
getting away with it.

35. The hordes of students now going to college are going to find it
more difficult to find good jobs when they graduate than did the
college graduates of the past.

36. Most repairmen will not overcharge even if they think you are
ignorant of their specialty.

37. A large share of accident claims filed against insurance companies
are phony.

38. One should not attack the religious beliefs of other people.

39. Most people answer public opinion polls honestly.

40. If we really knew what was going on in international politics, the
public would have more reason to be frightened than they now seem
to be.









APPENDIX B

Time in .5-sec. Intervals in the Six Response Categories by Rater for
the Six Groups in Each Treatment Condition

1. Personal Discussion

Interpersonal-Openness Therapy-Group Control Group
Rater Rater Rater
R Group Group R M Group R M

1 948 997 3 29 30 2 19 21
4 708 934 6 0 0 5 0 0
7 1071 1073 9 0 0 8 0 0
11 107 110 12 0 0 10 0 0
14 1395 1470 15 19 18 13 0 0
17 1591 1630 18 0 0 16 0 0
5820 6214 48 48 19 21



2. Feedback

Interpersonal-Openness Therapy-Group Control-Group
Rater Rater Rater
Group R M Group R M Group R M

1 0 0 3 0 0 2 48 52
4 30 36 6 0 0 5 6 5
7 493 501 9 0 0 8 0 0
11 0 0 12 0 0 10 4 0
14 70 80 15 0 0 13 0 0
17 46 39 18 0 0 16 0 0
639 656 0 0 58 57











3. Impersonal Discussion

Interpersonal-Openness Therapy-Group Control-Group
Rater Rater Rater
Group R M Grou p R M Group R M

1 1483 1428 3 2260 2223 2 2088 2072
4 1419 1218 6 2291 2316 5 2244 2274
7 419 446 9 2065 2150 8 2191 2159
11 2168 2169 12 2220 2257 10 2316 2336
14 727 698 15 1958 1979 13 2194 2177
17 360 334 18 2221 2197 16 2216 2211
6576 6293 13,015 13,122 13,249 13,229



4. Group Process

Interpersonal-Openness Therapy-Group Control-Group
Rater Rater Rater
Group R M Group R M Group R M

1 0 0 3 0 0 2 35 30
4 54 10 6 0 10 5 25 26
7 192 160 9 50 44 8 0 0
11 0 0 12 0 0 10 7 6
14 148 145 15 216 208 13 0 0
17 0 0 18 0 0 16 71 71
394 315 266 262 131 127










5. Descriptive Aspects of
Communicative Speech

Interpersonal-Openness Therapy-Group Control-Group
Rater Rater Rater
Group R Group R M Group R M

1 87 70 3 157 154 2 97 93
4 113 116 6 116 119 5 145 130
7 135 131 9 266 233 8 175 174
11 115 109 12 198 183 10 111 123
14 91 89 15 151 166 13 175 175
17 185 158 18 117 126 16 187 194
726 673 1005 981 890 889



6. Unscoreable Utterances

Interpersonal-Openness Therapy-Group Control-Group
Rater Rater Rater
Group R M Group R M Group R M

1 0 0 3 0 0 2 25 35
4 46 35 6 9 0 5 0 0
7 41 16 9 0 0 8 0 0
11 12 15 12 0 0 10 0 20
14 0 0 15 26 26 13 0 0
17 0 0 18 63 70 16 0 0
99 66 98 96 25 55
















REFERENCES



Alexander, F. Psychoanalytic therapy: Principles and applications.
New York: Ronald Press, 1946.

Allen, J.G. Implications of research in self-disclosure for group
psychotherapy. International Journal of Group Psychotherapy,
1973, 23, 306-321.

Bandura, A. Principles of behavior modification. New York: Holt,
Reinhart, 1969.

Bandura, A., Ross, D., & Ross, S. Imitation of film-mediated aggressive
models. Journal of Abnormal and Social Psychology, 1963, 66,
3-11.

Bednar, R.L., Weet, C., Evensen, P., Lanier, D., & Melnick, J. Empirical
guidelines for group therapy: Pretraining, cohesion and modeling.
Journal of Applied Behavioral Science, 1974, 10 (2), 149-165.

Blascovich, J., Ginsberg, G.P., & Veach, T.L. A pluralistic explanation
of choice shifts on the risk dimension. Journal of Personality
and Social Psychology, 1975, 31, 422-429.

Blascovich, J., Veach, T.L., & Ginsberg, G.P. Blackjack and the risky
shift. Sociometry, 1973, 36, 42-45.

Bradford, L.P. Membership and the learning process. In L.P. Bradford,
J.R. Gibb & K.D. Berne (Eds.), T-group therapy and laboratory
method. New York: Wiley, 1964.

Broadbent, D. Attention and the perception of speech. In Perception:
Mechanisms and models. San Francisco: W.H. Freeman, 1972.

Cozby, P.C. Self-disclosure: A literature review. Psychological
Bulletin, 1973, 79, 73-91.

Deutsch, M. Cooperation and trust: Some theoretical notes. Nebraska
symposium on motivation, 1962.

Dibner, A.S. Ambiguity and anxiety. Journal of Abnormal and Social
Psychology, 1958, 56, 165-174.

Dion, K., Baron, R., & Miller, N. Why do groups make riskier decisions
than individuals? In L. Berkowitz (Ed.), Advances in Experimental
Psychology, vol. 5. New York: Academic Press, 1970.



-51-







-52-



Doster, J.A. Individual differences affecting interviewee expectancies
and perceptions of self-disclosure. Journal of Counseling
Psychology, 1975, 22, 192-198.

Freed, A., Chandler, P.J., Blake, R.R., & Mouton, J. Stimulus and back-
ground factors in sign violation. Journal of Personality, 1955,
23, 499.

Freedman, J.L. Role playing: Psychology by consensus. Journal of
Personality and Social Psychology, 1969, 13, 107-114.

Gahagan, J.P., & Tedeschi, J.T. Strategy and the credibility of promises
in the Prisoner's Dilemma game. Journal of Conflict Resolution,
1968, 12, 224-234.

Gendlin, E.T. A theory of personality change. In J.T. Hart & T.M.
Tomlinson (Eds.), New directions in client-centered therapy.
Boston: Houghton Mifflin, 1970.

Goldstein, A.P., Heller, K., & Sechrest, L.B. Psychotherapy and the
psychology of behavior change. New York: Wiley, 1966.

Grosser, D., Polansky, N., & Lippit, R. A laboratory study of behavioral
contagion. Human Relations, 1951, 4, 115-142.

Hollander, E.P. Conformity, status and idiosyncrasy credit.
Psychological Review, 1958, 65, 117-127.

Jellison, J.M., & Riskind, J. A social comparison of abilities inter-
pretation of risk-taking behavior. Journal of Personality and
Social Psychology, 1970, 15, 375-390.

Jellison, J.M., & Riskind, J. Attribution of risk to others as a
function of their ability. Journal of Personality and Social
Psychology, 1971, 20, 413-415.

Johnson, J.A. Group therapy: A practical approach. New York: McGraw-
Hill, 1963.

Jourard, S.M. The transparent self. New York: Van Nostrand, 1971.

Kahan, J.P. A subjective probability interpretation of the risky shift.
Journal of Personality and Social Psychology, 1975, 31 (6),
977-982.

Kimbrell, D., & Blake, R.R. Motivational factors in the violation of
a prohibition. Journal of Abnormal and Social Psychology, 1958,
56, 132-133.







-53-


Kogan, N., & Wallach, M.A. Risk taking: A study in cognition and
personality. New York: Harper, 1964.

Korchin, S.J., Basowitz, H., Grinker, R., Hamburg, D., Persky, H.,
Sabshin, M., Heath, H., & Board, F. Experience of perceptual
distortion as a source of anxiety. American Medical Association
Archives of Neurology and Psychiatry, 1958, 80, 98-113.

Lakin, M., & Carson, R.C. A therapeutic vehicle in search of a theory
of therapy. Journal of Applied Behavioral Science, 1966, 2,
27-40.

Lazarus, R. Psychological stress and the coping process. New York:
McGraw-Hill, 1966.

McCauley, C., Stitt, C.L., Woods, K., & Lipton, D. Group shift to
caution at the race track. Journal of Experimental Social
Psychology, 1973, 9, 80-86.

Mowrer, O.H. The new group therapy. Princeton: Van Nostrand, 1964.

Norton, R., Feldman, C., & Tayofa, D. Risk parameters across type of
secrets. Journal of Counseling Psychology, 1974, 21, 450-454.

Otis, M.R. The effects of interpersonal skills training with videotaped
and live models on verbal and nonverbal components of behavior.
Unpublished master's thesis. University of Florida, 1975.

Postman, L., & Bruner, J. Perception under stress. Psychological
Review, 1948, 55, 314-323.

Pruitt, D.G. Conclusions: Toward an understanding of choice shifts in
group discussion. Journal of Personality and Social Psychology,
1971, 20, 404-510.

Rabin, H.M. Preparing patients for group psychotherapy. International
Journal of Group Psychotherapy, 1970, 20, 135-145.

Ribner, N.G. Effects of an explicit group contract on self-disclosure
and group cohesiveness. Journal of Counseling Psychology, 1974,
21 (2), 116-120.

Rogers, C.R. Client-centered therapy. In Silvano Arieti (Ed.),
Ameridan handbook of psychiatry, vol. 3. New York: Basic Books,
1959.

Rotter, J.B. A new scale for the measurement of interpersonal trust.
Journal of Personality, 1967, 35, 651-665.

Rotter, J.B. Generalized expectancies for interpersonal trust.
American Psychologist, 1971, 26, 443-452.







-54-


Schlenker, B.R., Helm, B., & Tedeschi, J.T. The effects of personality
and situational variables on behavioral trust. Journal of Per-
sonality and Social Psychology, 1973, 25 (3), 419-427.

Schofield, J.W. Effect of norms, public disclosure, and need for
approval on volunteering behavior consistent with attitudes.
Journal of Personality and Social Psychology, 1975, 31 (6),
1126-1133.

Schutz, W.C. Encounter. In Raymond Corsini (Ed.), Current
psychotherapies. Itasca, Ill: F.E. Peacock, Inc., 1973.

Singer, E. Key concepts in psychotherapy. New York: Basic Books,
1970.

Thibaut, J.W., & Kelley, H.H. The social psychology of groups.
New York: Wiley, 1959.

Truax, C.B., & Wargo, D.B. Effects of therapeutic conditions, alternate
sessions, vicarious therapy pretraining, and patient self-explora-
tion on hospitalized mental patients during group therapy.
Journal of Consulting and Clinical Psychology, 1969, 33, 440-447.

Verville, E. The effect of emotional and motivational sets on the
perception of incomplete pictures. Journal of General Psychology,
1946, 69, 133-145.

Vinokur, A. Cognitive and affective processes influencing risk taking
in groups: An expected utility approach. Journal of Personality
and Social Psychology, 1971, 20, 472-486.

Walters, R.H., & Parke, R.D. Influence of response consequences to a
social model on resistance to deviation. Journal of Experimental
Child Psychology, 1964, 1, 269-280.

Whalen, C. Effects of a model and instructions on group verbal be-
haviors. Journal of Consulting and Clinical Psychology, 1969, 33,
509-521.

Wolpe, J. The practice of behavior therapy. New York: Pergamon
Press, 1969.

Worthy, G., Gary, A.L., & Kahn, G. Self-disclosure as an exchange
process. Journal of Personality and Social Psychology, 1969,
13 (1), 59-63.

Wright, T., & Tedeschi, R. Factor analysis of the Interpersonal Trust
Scale. Journal of Consulting and Clinical Psychology, 1975, 43,
470-477.

Yalom, I.D. A study of group dropouts. Archives of General Psychiatry,
1966, 14, 393-414.







-55-



Yalom, I.D. The theory and practice of group psychotherapy. New York:
Basic Books, 1970.

Yinon, Y. Risky aggression in individuals and groups. Journal of
Personality and Social Psychology, 1975, 31, 808-815.


















BIOGRAPHICAL SKETCH



Sallye B. Peterson was born in Shreveport, Louisiana, on October 31,

1934, and grew up there, graduating from Byrd High School in 1951. She

received her B.A. magna cum laude from Louisiana State University in

1965 and her M.A. in Clinical Psychology in 1967.

Between 1967 and 1973 she taught at colleges in Rhode Island and

Tennessee and worked as a staff psychologist at the Helen R. McNabb

Community Mental Health Center in Knoxville, Tennessee. She is presently

employed as Coordinator of School Consultation at the Child, Youth, and

Family Center in Gainesville, Florida, where she has been working since

April, 1975. She is the mother of three children: Michael Andrew,

aged seven; Moetke Laren, aged six; and Megan Kathleen, aged two.





























-56-













I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.





Hugh C. is, Jr., Chairman
Professor of Clinical Psychology





I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.





James C. Dixon
Professor of Psychology





I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.





Nathan W. Perry
Professor of Clinical Psychology
Acting Chairman of the Department of
Clinical Psychology


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.



/ n *' /4 // d

Barry R. Schlenker
Associate Professor of Psychology













I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.


z, A /


Marvin E. Shaw
Professor of Psychology





I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

-------



Otto von Mering
Professor of AnthropoTbg








This dissertation was submitted to the Graduate Faculty of the Department
of Clinical Psychology in the College of Arts and Sciences and to the
Graduate Council, and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.

March 1977





Dean, Graduate School




Full Text
-11-
or disinhibition of previously learned responses, depending on the
consequences which the model is observed to experimence, or (3) facili
tation of responses through focusing of the observer's attention on
discriminative stimuli in the environment (Bandura, 1969).
Especially in the case of vicarious learning of complex verbal
response patterns, the observer's attention must be focused on the
essential elements of the model's behavior. As Whalen pointed out, what
the subject must learn is a set of rules and regulations, rather than
discrete responses; hence the importance of orienting instructions to
the observer in facilitating the coding of verbal mediators. The effects
of observational learning are maximized by such variables as a moderate
degree of similarity of the model to the observer and the observer's
arousal and motivational levels.
There is experimental evidence in support of Bandura's theory as
it relates to observational learning of responses which deviate from
the patterns of behavior which the observer would ordinarily expect in
the situation in which the model is viewed. Apparently, the disinhibit-
ing effects of the modeling occur through (1) establishing a norm for
such behavior; and (2) providing discriminative stimuli that inform the
observer as to when the norm is in operation (Freed, Chandler, Blake &
Mouton, 1955; Grosser, Polansky & Lippit, 1951; Kimbrell & Blake, 1958;
Walters 6 Parke, 1964).
According to Bandura's theory of observational learning and the
experimental evidence supporting that theory, Hypothesis 3 is stated as
follows:
Hypothesis 3: Groups of individuals exposed to explicit,
detailed instructions and to a videotape of a group in which


Table 3
Mean Number of Seconds, Percent of Total Time, and Kruskal-Wallis Analysis
of Variance for the Three Treatment Conditions
Condition Means and Percent
Interpersonal
Therapy
Control
Openness
Group
Group
Response
Category
Mean
Percent
Mean
Percent
Mean
Percent
H
Pers Disc
1112.83
45.1
8.00
.3
12.58
.5
12.51**
Feedback
Impers
107.92
4.5
0.00
.0
9.92
.4
6.12*
Disc
1002.42
42.5
2178.08
90.2
2206.42
91.5
9.58**
Group
Process
59.08
2.4
44.83
1.9
28.00
1.2
.07
Descrip
116.63
4.9
165.50
6.9
140.33
5.8
4.16
Unscore
13.75
.6
16.92
.7
13.33
.6
.26
2412.63
100.0
2413.33
100.0
2410.58
100.0
*£ < .05
**£ < .01


-55-
Yalom, I.D. The theory and practice of group psychotherapy. New York
Basic Books, 1970.
Yinon, Y. Risky aggression in individuals and groups. Journal of
Personality and Social Psychology, 1975, 31, 808-815.


-43-
the videotaped discussions might possibly reveal the answers. Such an
analysis is beyond the scope of the present investigation.
Conclusions
The results of the present study suggest some important conclusions
for the theory and practice of group psychotherapy. If the data herein
can be viewed tentatively as applicable to a clinic population, we
might conclude:
1. That members of a group can be taught to behave in thera
peutically relevant ways (e.g., relating in an interpersonally open
manner);
2. That such training can be carried out in a simple and practical
manner through the use of brief instructions and videotape models;
3. That, when members of a group engage in such interpersonally
open behavior, a climate of trust is built up wherein the group members
become willing to share with each other painful and embarrassing aspects
of their lives;
4. That, while the actual disclosure of such secrets is not
necessarily a sine qua non of therapy, the readiness to do so may be;
and
5. Finally, that trust in one's fellow group members does gener
alize, at least to the extent in the present study of trusting the
experimenter.
Necessarily, generalization of the results of this study to a
clinic population must wait upon replication with actual therapy groups.
However, such research seems clearly justified on the basis of the data
obtained herein.


-46-
3. In dealing with strangers one is better off to be cautious until
they have provided evidence that they are trustworthy.
4. This country has a dark future unless we can attract better people
into politics.
5. Fear of social disgrace or punishment rather than conscience pre
vents most people from breaking the law.
6. Parents usually can be relied upon to keep their promises.
7. The advice of elders is often poor because the older person doesn't
recognize how times have changed.
8. Using the honor system of not having a teacher present during
exams would probably result in increased cheating.
9. The United Nations will never be an effective force in keeping
world peace.
10. Parents and teachers are likely to say what they believe themselves
and not just what they think is good for the child to hear.
11. Most people can be counted on to do what they say they will do.
12. As evidenced by recent books and movies morality seems on the
downgrade in this country.
13. The judiciary is a place where we can all get unbiased treatment.
14. It is safe to believe that in spite of what people say, most people
are primarily interested in their own welfare.
15. The future seems very promising.
16. Most people would be horrified if they knew how much news the
public hears and sees is distorted.
17. Seeking advice from several people is more likely to confuse than
it is to help one.
18. Most elected public officials are really sincere in their campaign
promises.
19. There is no simple way of deciding who is telling the truth.
20. This country has progressed to the point where we can reduce the
amount of competitiveness encouraged by schools and parents.
21. Even though we have reports in newspapers, radio and television,
it is hard to get objective accounts of public events.
22. It is more important that people achieve happiness than that they
achieve greatness.


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Professor of Clinical Psychology
1 certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
James C. Dixon
Professor of Psychology
* I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
y
/( /7 c><-4y
Nathan W. Perry
Professor of Clinical Psychology
Acting Chairman of the Department of
Clinical Psychology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Barry R. Schlenker
Associate Professor of Psychology


-54-
Schlenker, B.R., Helm, B., & Tedeschi, J.T. The effects of personality
and situational variables on behavioral trust. Journal of Per
sonality and Social Psychology, 1973, 15 (3), 419-427.
Schofield, J.W. Effect of norms, public disclosure, and need for
approval on volunteering behavior consistent with attitudes.
Journal of Personality and Social Psychology, 1975, 31 (6),
1126-1133.
Schtz, W.C. Encounter. In Raymond Corsini (Ed.), Current
psychotherapies. Itasca, Ill: F.E. Peacock, Inc., 1973.
Singer, E. Key concepts in psychotherapy. New York: Basic Books,
19 70.
Thibaut, J.W., & Kelley, H.H. The social psychology of groups.
New York: Wiley, 1959.
Truax, C.B., & Wargo, D.B. Effects of therapeutic conditions, alternate
sessions, vicarious therapy pretraining, and patient self-explora
tion on hospitalized mental patients during group therapy.
Journal of Consulting and Clinical Psychology, 1969, 33, 440-447.
Verville, E. The effect of emotional and motivational sets on the
perception of incomplete pictures. Journal of General Psychology,
1946, 69, 133-145.
Vinokur, A. Cognitive and affective processes influencing risk taking
in groups: An expected utility approach. Journal of Personality
and Social Psychology, 1971, 2(), 472-486.
Walters, R.H., & Parke, R.D. Influence of response consequences to a
social model on resistance to deviation. Journal of Experimental
Child Psychology, 1964, 1, 269-280.
Whalen, C. Effects of a model and instructions on group verbal be
haviors. Journal of Consulting and Clinical Psychology, 1969, 33,
509-521.
Wolpe, J. The practice of behavior therapy. New York: Pergamon
Press, 1969.
Worthy, G., Gary, A.L., & Kahn, G. Self-disclosure as an exchange
process. Journal of Personality and Social Psychology, 1969,
13 (1), 59-63.
Wright, T., & Tedeschi, R. Factor analysis of the Interpersonal Trust
Scale. Journal of Consulting and Clinical Psychology, 1975, 43,
470-477.
Yalom, I.D. A study of group dropouts. Archives of General Psychiatry,
1966, 14, 393-414.


-18-
interact as a group and to try to get to know each other. In order to
help you gain a clearer understanding of how a group like yours might
function, we will show you a videotape of four people in a situation
similar to the one you will find yourselves in. After the tape is over
you may begin your own interaction. No experimenter will participate
in your group, but an observer will be watching through the one-way
mirror in order to record the interaction variables we are examining.
You will have 40 minutes, and you will be notified when your time is up.
"Next week at this same time you are to return to this room for
your second group session."
Condition 3. Control group. Condition 3 was an attempt to examine
the effects of a group's simply meeting together (after viewing a
videotape not depicting group interaction) on the dependent variables
under investigation. The videotape shown was of the same length as
those shown to groups in Conditions 1 and 2 and with the same partici
pants, but depicting simply a variety of modes of interaction (teaching,
working at a task together, drinking and eating together).
Instructions to the groups in the control-group condition were the
same as those for Condition 2 groups, except that description of the VTR
was altered as follows: "This is a group dynamics laboratory. We are
interested in examining some variables important in the development of
group interaction. To help us study these variables we will ask you to
interact as a group and to try to get to know each other. In order to
help you get in the mood for your group experience, we will show you a
videotape of people interacting in various ways. After the tape is
over you may begin your own interaction. No experimenter will partici
pate in your group, but an observer will be watching through the one-way


-39-
Hypothesis 2: Willingness to Disclose Painful Secrets
Hypothesis 2 is supported by the finding that members of groups in
the interpersonal-openness treatment condition were willing to disclose
a significantly greater number of secrets to their fellow members than
were subjects in the other two treatment conditions. The fact that the
subjects did not differ across conditions in terms of how painful or
embarrassing they considered their secrets to be indicates that the same
degree of risk obtained for all the group members.
Hypothesis 2 predicted that the readiness to risk disclosing secrets
would be related to the presence or absence of a norm of interpersonal
openness. Analyzing the treatment condition in terms of the norm-sending
process detailed by Thibaut and Kelley (1959), one finds that the three
components of norm-sending were clearly present only in the interpersonal
openness condition. The components and their counterparts in the
present study are (1) rule statements (instructions and videotape);
(2) surveillance-evaluation (presence of other group members and the
experimenter's observation from the next room); and (3) application of
sanctions (applied by the subject and the other group members in the
form of rewards for interpersonal openness; the expected reward promised
in the instructions of "improving your ability to communicate with
others . ."). Additionally, the finding that groups in the inter
personal-openness condition actually spent approximately half of their
time in personal discussion and feedback indicates that a norm of
interpersonal openness was operating.
There is also some support for the idea that a norm of interper
sonal openness generalized outside the group. Such support is found in
the fact that a significantly greater number of members of groups in


Percent
-31-
Disc Disc
. B
Group Descrip
Process

Models
Group
Members
Unscore
Figure 2. Comparison of models and groups in the therapy-group
condition on percent of time spent in each of the six
categories of response.


-41-
life experiences with parents, teachers, and peers.
As measured by Rotter's scale, then, interpersonal trust is an
attitude that would not be expected to fluctucate over such a short
period of time as was involved in the present study unless some very
dramatic change occurred in the lives of the group members. Apparently,
the experimental manipulation, while it was effective in altering the
expectancies of the subjects relative to the specific situation of the
experiment (witness the significant differences on other dependent
variables) was not potent enough to effect a change in what might be
thought of as a personality or trait variable of trust. Such an effect
as personality change might be obtained over a longer period of time
with repeated experiences in groups such as those in the present study.
One would expect that a successful therapy-group experience might pro
duce such a shift in attitude.
Alternatively, if interpersonal trust is less narrowly defined,
the data on the subjects' willingness to share secrets and to allow them
to be analyzed by the experimenter could be taken as an indication that
trust in other persons was affected differentially by the treatment
conditions. Certainly, some degree of trust is indicated by a readiness
to disclose painful aspects of one's self to others. However, such a
conclusion is limited in its applicability until further research
demonstrates that the effects endure over time and across other groups
or individuals. Hypothesis 1 is not clearly supported by the data
obtained in the present study.
The Risky Shift
Groups in all three treatment conditions shifted significantly in


-2-
situation is antithetical to the traditional conceptualization of the
manner in which groups function; that is, that an ambiguous, unstructured
situation is most conducive to client self-exploration and the natural
development of the group (cf. Rabin, 1970). The necessity for ambiguity
in therapy groups is derived from theories that stress the importance
of a therapeutic relationship (individual as well as group) being unlike
the client's historical relationships in which he has learned to behave
in maladaptive ways (Singer, 1970). i
But the experimental literature is replete with data on the detri
mental effects of ambiguous situations on behavior, specifically the
relationship between ambiguity and anxiety (Lazarus, 1966). A classic
study by Dibner (1958) compared the amount of anxiety experienced by
patients who were prepared for a psychiatric interview (told how to
behave and what to expect) with that experienced by unprepared patients
(highly ambiguous situation). Behavioral data, subjective reports, and
GSR data from the patients all demonstrated that unprepared patients
experienced a significantly greater degree of anxiety. The destructive
effects of anxiety induced by ambiguity are also documented by studies
of perceptual acuity (Postman & Bruner, 1948) and visual organization
and synthesis of cues (Verville, 1946). When attention is divided, or
when the individual has difficulty knowing which elements of a situation
to attend to, the perception of speech is likely to be confused and
distorted (Broadbent, 1972). And the destructive chain continues with
distorted perceptions leading to increased anxiety (Korchin et al., 1958).
On the basis of their experience alone group therapists can testify
to the evident anxiety, inappropriate behavior, and desperate looking to
the therapist for help that occur in the initial meetings of groups.


ACKNOWLEDGEMENTS
Research is not a solitary activity. I extend my deepest gratitude
to those who have supported me in this work.
To Hugh Davis, who served as Chairman, friend, and counsel in times
of despair.
To Jim Dixon, for his philosophy and inspiration.
To Nate Perry, for his support and positive criticism.
To Barry Schlenker, for his pragmatic suggestions and friendship.
To Marvin Shaw, for his encouragement and consultation.
And to Otto von Mering, for his personal interest and support.
Thanks are also due to Mark Barnett and Randy Thomas, who spent
long hours in rating the videotapes. And to Nancy Monroe, who assisted
in the painful task of data reduction.
iii


-5-
raethod of treatment is also seen as essentially interpersonal. Even
behavior theorists such as Wolpe regard "improved interpersonal
relationships" as among the primary criteria for evaluating therapeutic
change (Wolpe, 1969). Thus it would appear that the major characteris
tic of the group over which the therapist must exert some control is
that of the ways in which the members behave toward each other. Such
a statement seems both simplistic in its obviousness and antagonistic
to the pusposes of group therapy: the idea that the therapist can or
should control the members' behavior. Yet, when no consistent in
fluences are brought to bear on the members' behavior in the situation
one may observe the relative lack of structure and norm-ambiguity which
lead to heightened anxiety and low risk-taking on the part of the group
members.
What particular aspects of interpersonal relations have therapists
considered relevant? One of the most frequently cited is what might be
summarized as "interpersonal openness," or the "readiness to convey
personal information about one's life, one's feelings, and one's im
pressions of other individuals" (Whalen, 1969, p. 510). Rogers (1959)
emphasizes the importance of such behavior for both client and therapist,
as do Jourard (1971), Gendlin (1970), and Mowrer (1964), and the en
counter movement therapists (e.g., Bradford, 1964; Schtz, 1973). Even
analytically oriented therapists such as Franz Alexander (1946) have
recognized the importance of the client's experiencing in the present
and honestly reporting that experience in order to undergo a "corrective
emotional experience."
What are the factors which might contribute to a group situation
in which the members consistently behave in an interpersonally open


-7-
clarifying interpersonal relationships. However, an individual is un
likely either to risk such new behavior or, once having done so, to
accept the feedback he receives except where group members share a
fairly high degree of interpersonal trust.
Deutsch (1962) defined a trusting situation, as differentiated from
a gambling situation, as one in which one "has much to lose or little
to gain," probably the view most new group members hold of the therapy
group. He reasoned that one needs "considerable confidence in a posi
tive outcome" to trust. From his research on cooperation, competition,
and trust, Deutsch defined trust in motivational terms: cooperative or
competitive motives will determine the subject's behavior in the situa
tion; but an individualistic orientation, neither cooperating nor
competing but simply trying to maximize one's gains, resulted in the
subject's behavior being much more influenced by the specific experi
mental conditions. In terms of Deutsch's model, if a group member
exposes himself to possible criticism, ridicule, rejection, or even
potential blackmail by other members, such behavior is indicative of
trust "much to lose." Whether he engages in such behavior will be
strongly influenced by situational variables; in this case the behavior
of the other group members.
Schlenker, Helm, and Tedeschi (1973) offer a definition of inter
personal trust as "a reliance upon information received from another
person about uncertain environmental states and their accompanying
outcomes in a risky situation" (emphasis supplied). In that definition
the importance of situational variables in the operation of interpersonal
trust is even more central. And Rotter (1967, 1971) likewise emphasized
reliance upon communications from other individuals in developing his


CHAPTER III
RESULTS
Interrater Reliability
To determine whether the raters were reliable in their judgments
of the total number of seconds spent in each category of discussion,
Spearman's rank correlation coefficient was computed for the 108 pairs
of scores. (Appendix A shows the raters' scores.) The resulting test
statistic of r^ = .986 is highly significant (j) < .001). Individual
correlation coefficients were also computed over the 18 groups for
each of the six categories. The results are shown in Table 1.
Table 1
Interrater Reliability
Response Category
Spearman's r
£
Personal discussion
.999
.01
Feedback
.955
.01
Impersonal discussion
.981
.01
Group process
.944
.01
Descriptive aspects of
communicative speech
.964
.01
Unscoreable utterances
.844
.01
Since interrater reliability was judged to be satisfactory, the
average of the two raters' scores was used in further computations.
-25-


Abstract of Dissertation Presented to the Graduate Council
of the University for Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
TRUST AND RISK-TAKING AS A FUNCTION OF
PRE-GROUP TRAINING
By
Sallye B. Peterson
March 1977
Chairman: Hugh Davis, Jr.
Major Department: Clinical Psychology
Interpersonal trust and the degree to which subjects are willing
to share painful aspects of themselves were investigated as a function
of pretraining using models. Seventy-two female college sophmores were
instructed to write out privately 20 of their most painful secrets.
They were divided into eighteen four-member groups in three treatment
conditions and exposed to instructions and videotape models corresponding
to an interpersonal-openness condition, a therapy-group condition, and
a control condition. Each group then had 40 minutes of interaction and
returned one week later for what they believed to be their second group
session. At that time they chose a number of secrets to be shared with
their group and also filled out an interpersonal-trust scale. They
also discussed as a group and achieved consensus on the number of
secrets each member should share, as an investigation of the risky-shift
phenomenon. Raters analyzed the videotaped first session of each group
using six response categories. Interpersonal-openness groups differed
significantly from the other groups in their verbal interaction, in
vi


-29-
Table 4
Differences between Pairs of Treatment Conditions on the
Three Significantly
Different Response Categories
Response
Category
Treatments (in Increasing
Order of the Means of the Ranks)
Personal
Discussion
T-G C-G
1-0
Feedback
T-G C-G
1-0
Impersonal Discussion 1-0 T-G C-G
Note: Treatments connected by a line are not significantly different
at the .05 level.
Comparison of Groups and Their Models
Figures 1 through 3 represent comparisons of the percent of time
spent by group members and by their models in the six categories of
discussion. Groups in the therapy-group and control-group conditions
most closely matched their models. In the interpersonal-openness con
dition the groups matched their models except in the category of
impersonal discussion, where the group members spent 42.5% of their
time, compared to the models' 12%.
Dependent Measures
To determine whether the groups in the three treatment conditions
differed among themselves with regard to the five separate response
variables, the Kruskal-Wallis one-way analysis of variance was computed
for each of the dependent variables. Reference to Table 5 shows that
the groups did differ significantly on three of the dependent variables:


TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS . iii
ABSTRACT vi
CHAPTER I INTRODUCTION. 1
Pretherapy Training 3
Interpersonal Trust 6
Disclosure of Painful Secrets 8
Establishing a Norm by Pregroup Training: Modeling 10
The Risky Shift 12
CHAPTER II METHOD AND PROCEDURES 14
Subjects 14
Procedure 15
Dependent Measures. 21
CHAPTER III RESULTS 25
Interrater Reliability 25
Analysis of the Modeling Tapes 26
Analysis of the Subjects' Interaction in the Groups 27
Comparison of Groups and Their Models 29
Dependent Measures 29
CHAPTER IV DISCUSSION 37
Hypothesis 3: The Differential Effects of Modeling 37
Hypothesis 2: Willingness to Disclose Painful
Secrets 39
iv


-6-
fashion? The question of whether setting of norms can facilitate
specific behaviors has been investigated with uniformly positive results.
Ribner (1974) found that groups receiving an explicit group contract
calling for self-disclosure increased significantly both their frequency
and depth of self-disclosure.
The clearest demonstration of building into a group a norm of
interpersonal openness is Whalen's study, in which the combination of a
film demonstration modeling interpersonal openness and detailed instruc
tions of expected behavior produced significantly greater effects than
either independent variable alone. In explaining her results Whalen
relied on an attention-information format detailed by Thibaut & Kelley
(1959). The norm-sending process involved rule statements (filmed
modeling plus instructions), surveillance-evaluation (presence of other
group members), and application of sanctions (applied by the subject as
well as the other group members). Such a conceptualization stresses
the learning of complex verbal response classes, a type of learning
requiring that the subject attend to the relevant variables in the
situation rather than simply imitate a discrete response.
Interpersonal Trust
How can interpersonal openness contribute to the therapeutic effec
tiveness of the group for its members? One of the means by which group
therapy is considered to bring about changes in an individual's habits
and attitudes is through allowing him to try out new and different ways
of relating to others in the group. Yalom (1970) has emphasized the
importance of group feedback in modifying interpersonal distortions and


-23-
and tell her." (d) Acceptance of feedback: Acceptance by a group
member of the positive, negative, or neutral feedback she receives,
for example, "Yeah, I guess you're right." (e) Reject feedback:
Refusal to accept the feedback about oneself provided by other group
members, for example, "No, you see it's really like this ..."
(f) Request feedback: A direct request that the group or particular
members provide the subject with an evaluation of herself, for example,
"What do you think of me?"
(These first two responses classes comprised the primary types
of interaction modeled in the videotape recording for the
interpersonal-openness groups.)
3. Impersonal discussion. (a) Impersonal self-disclosure:
Nonpersonal biographical information about oneself which is either
generally accessible or readily volunteered in appropriate contexts by
most people. Examples: "I come from Chicago" or "I don't know the
outskirts of the city very well." (b) Extragroup process: Nonpersonal
topics unrelated to the experiment or the group interactions "All
universities don't accept correspondence courses" or "I know a girl who
is in the Navy now." (c) Impersonal questions: Questions regarding
public, easily accessible, readily volunteered information about a
group member "Have you ever seen the San Diego campus?" or "How many
hours are you taking?"
4. Group process. Nonpersonal discussion relevant to the experi
ment such as comments regarding the experimental manipulations, the
purposes of the experiment, the behavior of the experimenter or par
ticipants in the VTR, a plan of action, etc. Examples: "Are they
watching us?" or "Let's all talk at the same time that ought to


-9-
secrets which they regard as (1) extremely important to their present
problems and (2) too dangerous or embarrassing to reveal to anyone else.
With the possible exception of psychoanalysts, most therapists do not
regard the client's revelation of such secrets as crucial to his
improved functioning. Thus, "self-disclosure" to the extent of re
vealing in detail one's greatest worries, shames, and fears has not been
shown to be an effective or even desirable goal of psychotherapy.
Although Jourard (1971) and other humanistic psychologists have
maintained that "Self-disclosure is a symptom of personality health and
a means of ultimately achieving healthy personality," research findings
have been contradictory. Cozby (1973) pointed out in his review of the
self-disclosure literature that one reason for the difficulty in com
paring results across studies is the lack of consistency in defining
"mental health." And Allen (1973), reviewing the methodological
limitations of self-disclosure research, has listed as primary the fact
that the research is based on relatively nonintimate disclosures. Thus,
the extension of findings from such studies requires substantial con
ceptual leaps. Particularly with regard to groups there is reason to
question the therapeutic benefits to be gained from self-disclosure,
since the majority of findings are based on self-report or dyadic
interactions. Finally, "self-disclosure" has been defined in virtually
all of the research in terms of the Jourard questionnaire or variants
of it (Allen, 1973), and the validity of these measures has yet to be
established.
Despite the lack of evidence establishing the therapeutic relevance
of actual self-disclosure thus defined, an index of the extent to which
we trust each other is our willingness to have them know our "painful


-53-
Kogan, N., & Wallach, M.A. Risk taking: A study in cognition and
personality. New York: Harper, 1964.
Korchin, S.J., Basowitz, H., Grinker, R., Hamburg, D., Persky, H.,
Sabshin, M., Heath, H., & Board, F. Experience of perceptual
distortion as a source of anxiety. American Medical Association
Archives of Neurology and Psychiatry, 1958, ^0, 98-113.
Lakin, M., & Carson, R.C. A therapeutic vehicle in search of a theory
of therapy. Journal of Applied Behavioral Science, 1966, 2^,
27-40.
Lazarus, R. Psychological stress and the coping process. New York:
McGraw-Hill, 1966.
McCauley, C., Stitt, C.L., Woods, K., & Lipton, D. Group shift to
caution at the race track. Journal of Experimental Social
Psychology, 1973, _9, 80-86.
Mowrer, O.H. The new group therapy. Princeton: Van Nostrand, 1964.
Norton, R., Feldman, C., & Tayofa, D. Risk parameters across type of
secrets. Journal of Counseling Psychology, 1974, _21, 450-454.
Otis, M.R. The effects of interpersonal skills training with videotaped
and live models on verbal and nonverbal components of behavior.
Unpublished master's thesis. University of Florida, 1975.
Postman, L., & Bruner, J. Perception under stress. Psychological
Review, 1948, 55, 314-323.
Pruitt, D.G. Conclusions: Toward an understanding of choice shifts in
group discussion. Journal of Personality and Social Psychology,
1971, 20, 404-510.
Rabin, H.M. Preparing patients for group psychotherapy. International
Journal of Group Psychotherapy, 1970, jM), 135-145.
Ribner, N.G. Effects of an explicit group contract on self-disclosure
and group cohesiveness. Journal of Counseling Psychology, 1974,
21 (2), 116-120.
Rogers, C.R. Client-centered therapy. In Silvano Arieti (Ed.),
Ameridan handbook of psychiatry, vol. 3. New York: Basic Books,
1959.
Rotter, J.B. A new scale for the measurement of interpersonal trust.
Journal of Personality, 1967, 35, 651-665.
Rotter, J.B. Generalized expectancies for interpersonal trust.
American Psychologist, 1971, 2jj, 443-452.


1
4
7
11
14
17
OUP
1
4
7
11
14
17
3. Impersonal Discussion
nterpersonal-Openness
Therapy
-Group
Control1
-Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
1483
1428
3
2260
2223
2
2088
2072
1419
1218
6
2291
2316
5
2244
2274
419
446
9
2065
2150
8
2191
2159
2168
2169
12
2220
2257
10
2316
2336
727
698
15
1958
1979
13
2194
2177
360
334
18
2221
2197
16
2216
2211
6576
6293
13,015
13,122
13,249
13,229
4. Group Process
Interpersonal-Openness
Therapy-
Group
Control-
-Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
0
0
3
0
0
2
35
30
54
10
6
0
10
5
25
26
192
160
9
50
44
8
0
0
0
0
12
0
0
10
7
6
148
145
15
216
208
13
0
0
0
0
18
0
0
16
71
71
394
315
266
262
131
127
-49-


-19-
mirror in order to record the interaction variables we are examining.
You will have 40 minutes, and you will be notified when your time is up.
"Next week at this same time you are to return to this room for
your second group session."
Each of the groups in the three conditions saw their VTR and re
ceived their instructions as a group but apart from the other seventeen
groups. Each group met in a room containing four chairs located around
a circular table with each chair partly facing a one-way mirror.
Observing and videotaping was done from the adjoining room. The sub
jects were met by the experimenter when they entered the experimental
room and were asked not to talk to the other group members until the
experiment began. When all four subjects had arrived, the experimenter
passed out the written and played the taped instructions, and then
showed the VTR appropriate to the particular treatment condition.
After the subjects had seen the tape and before the experimenter
left the room she handed out a 5 x 8 index card and a marking pen to
each subject, requesting that she write her first name on it, fold it
in half, and stand it up on the table in front of her so that the
other members could see it. The experimenter then left the room. After
40 minutes had elapsed, the experimenter re-entered the room to end
the interaction, to remind the group that they were due back for the
second session one week later, and to request that they not discuss
the experiment with their friends.
The second session. One week after the first group meetings
all groups returned for a "second session." Before the second
session began, each subject was left alone in a private room to
fill out Rotters Interpersonal Trust Scale. After she had com
pleted the Scale (reproduced in Appendix A),


BIOGRAPHICAL SKETCH
Sallye B. Peterson was born in Shreveport, Louisiana, on October 31,
1934, and grew up there, graduating from Byrd High School in 1951. She
received her B.A. magna cum laude from Louisiana State University in
1965 and her M.A. in Clinical Psychology in 1967.
Between 1967 and 1973 she taught at colleges in Rhode Island and
Tennessee and worked as a staff psychologist at the Helen R. McNabb
Community Mental Health Center in Knoxville, Tennessee. She is presently
employed as Coordinator of School Consultation at the Child, Youth, and
Family Center in Gainesville, Florida, where she has been working since
April, 1975. She is the mother of three children: Michael Andrew,
aged seven; Moetke Laren, aged six; and Megan Kathleen, aged two.
-56-


-4-
training chosen and the process variable studied, and between process
and outcome is unclear. For example, it is difficult to conceptualize
how listening to excerpts of "good" patient in-therapy behavior can lead
to decreased MMPI scores (Truax & Wargo, 1969), especially when no
definition of "good" behavior is given.
Whalen (1969) conducted a study on pregroup training with a popula
tion of college students. In her study she clearly defined the variables
which she was seeking to manipulate: particular types of verbal inter
action, e.g., personal discussion and feedback. Whalen's study was
limited to the effects of pretraining on group interaction, and no
"outcome" variables were measured.
What seem to be lacking in the research are studies in which the
critical factors in the group situation are defined and some relation
ship is established among pretraining, group process, and outcome.
Obviously the parameters of the group situation seen as crucial to
client improvement depend on one's theoretical orientation; the critical
process and outcome variables and their relationship to factors in the
group situation are likewise dependent on one's conception of how
therapeutic change takes place; and the most effective means of in
fluencing the crucial group parameters depends on the researcher's
conceptions in the areas stated above plus one's evaluation of the
evidence as to how learning occurs.
Critical factors in the group situation. Most modern systems of
psychotherapy are interpersonally based. Stemming from the basic theses
propounded by Harry Stack Sullivan in his interpersonal theory of
psychiatry, current psychotherapies view the origins of psychopathology
as existing in disordered interpersonal relations; thus, the appropriate


Percent
-30-
100
Models
90
80
Group
Members
70
60
50
40
30
20
10
0
Pers Feedback
Disc
Impers
Disc
Group Descrip Unscore
Process
Figure 1. Comparison of models and groups in the interpersonal
openness condition on percent of time spent in each of
the six categories of response.


-12-
the members model Interpersonal openness will engage in
significantly more interpersonally open behavior than will
groups who do not receive such pretraining.
In summary, it was hypothesized that pretraining via modeling and
detailed instructions would lead to greater interpersonal openness on
the part of group members; that such openness would in turn result in
more interpersonal trust among the members; and that a greater degree
of interpersonal trust would also be reflected in a greater willingness
to risk disclosure of painful secrets to the group.
The Risky Shift
Studies using the Choice Dilemma Questionnaire developed by Kogan
and Wallach (1964) have often found a shift in decision-making in a
"riskier" direction following group discussion. However, the shift is
not uniform for all items (some shifting consistently in a more con
servative direction), and stronger risky shifts are associated with
items having the least serious consequences (Dion, Baron & Miller,
1970). Explanations of the phenomenon have been many and varied, in
cluding the decision-making model of Pruitt (1971) ; a social-comparison
interpretation (Jellison & Riskind, 1970, 1971) ; and the expected-
utility model of Vinokur (1971), recently amplified in the subjective
expected utility model of Kahan (1975) .
As pointed out by Dion et al., however, a strong reservation in
interpreting the results of any risky-shift study using the CDQ is the
extent to which the tasks lack any (much less severe) consequences for
the group members. And Freedman (1969) argues cogently against the


-20-
she was handed an empty envelope, her original envelope containing her
20 secrets, and a typed instruction sheet which read as follows:
"Please take out of your envelope aj^ random that means, without
looking into the envelope as many, from zero to twenty, of the slips
of paper as you are willing to put in the pot for discussion in the
group tonight. The actual discussion session of your group will not be
recorded tonight. Please put the slips of paper you select into this
second envelope, write your name across the front of it, and drop it in
the bowl on the table. Hold on to your original envelope and call me."
When the four members of a group were assembled the experimenter
instructed them as follows: "Before you begin your group tonight I
would like you to discuss among yourselves the number of secrets which
each member of your group should contribute to discussion. You are not
to discuss the content of the secrets but only the number, and you must
agree as a group on one number." Discussion to consensus was video
taped. This portion of the experimental manipulation was a check on
whether the risky-shift phenomenon was operating in these groups.
After consensus was reached, the experiment was over, and the subjects
were debriefed regarding their expectations and hypotheses about the
experimental manipulations.
Each subject was seen individually at this point and given a slip
of paper on which she was asked to rate on a scale from 1 to 7 how
painful or embarrassing her secrets were (a rating of 1 representing
"not embarrassing or painful at all" and a rating of 7 representing
"extremely embarrassing or painful"). She was then given a typewritten
slip of paper with directions as follows:
"You were told that no one would ever read your secrets


APPENDIX A
Interpersonal Trust Scale (Rotter, 1967)
General Opinion Survey
This is a questionnaire to determine the attitudes and beliefs of
different people on a variety of statements. Please answer the state
ments by giving as true a picture of your own beliefs as possible. Be
sure to read each item carefully and show your beliefs by marking the
appropriate number of your answer sheet.
If you strongly agree with an item, circle number 1. Circle number
2 if you mildly agree with the item. That is, circle number 2 if you
think the item is generally more true than untrue according to your
beliefs. Circle number 3 if you feel the item is about equally true as
untrue. Circle number 4 if you mildly disagree with the item. That is,
circle number 4 if you feel the item is more untrue than true. If you
strongly disagree with an item, circle number 5.
1. Strongly agree
2. Mildly agree
3. Agree and disagree equally
4. Mildly disagree
5. Strongly disagree
1. Most people would rather live in a climate that is mild all year
around than in one in which the winters are cold.
2. Hypocrisy is on the increase in our society.
-45-


-38-
Thus, while interpersonally open behavior is difficult to induce, the
interpersonal-openness models and instructions used in this study were
effective in producing a significantly greater amount of that behavior
as compared to the other two conditions, the therapy-group and the
control-group.
An additional essential element in effective modeling other than
the clarity of the behavior to be matched and the explicit instructions
appears to be motivation. Only groups in the interpersonal-openness
condition received as part of their instructions the statement that
"[This] is a rare opportunity to improve your ability to communicate
with others by learning how you appear to them; that is, how you present
yourself and what impressions they form about you." A motivational
factor was also present in the modeling tape in that a member who was
open was reinforced by the other group members with positive feedback.
It seems probable that the treatment effects in the present study were
exerted through Bandura's principle of disinhibition of a previously
learned response due to the positive consequences experienced by the
models. Lakin and Carson (1966) have observed that people most often
avoid engaging in personal discussion because of the fear of being per
ceived as weak and inadequate. To counteract such a fear and disinhibit
the modeled response, the subject must have been motivated by an
expectation of positive consequences.
Thus, Hypothesis 3 is supported with the qualification that suf
ficient motivation in the subjects to model the behavior must also be
induced.


-35-
Table 7 is a further analysis of the "risky-shift" data. By
reference to that table it can be seen that the groups in the three
treatment conditions differed significantly in the number of secrets
which they were willing to share both before and after discussion to
consensus (p < .05). However, after discussion to consensus the only
significant difference was between the interpersonal-openness and
control-group conditions. The differences between the number of secrets
subjects were willing to share before and after discussion were signi
ficant in each of the treatment conditions and were all in the direction
of a "conservative" rather than a "risky" direction (Sign Test).
Table 7
Comparison of Mean Number of Secrets Subjects Were
Willing to Reveal Before and After Discussion to Consensus
Interpersonal Therapy Control
Openness
Group
Group
£
Before
Discussion
7.16
2.96
2.78
.05
After
Discussion
2. 70
1.00
.37
.05
Shift
5.46
1.96
2.41
£
.02
.03
.02
Note: Treatments connected by a line are not significantly different
at the .05 level.
To discover whether the groups differed among themselves with
reference to the amount of time they spent in their discussion to con
sensus, the Kruskal-Wallis test was performed. The mean amount of time


-3-
Johnson (1963) and Yalom (1966) point out two additional experiences
familiar to group therapists, i.e., that norms and behavior patterns
established in early sessions tend to persist throughout the life of
the group, and that as many as one-third of the group members drop out
during the early (1st to 12th) group meetings. In the face of accumu
lating evidence of the deleterious effects of situational ambiguity on
the behavior of clients in therapy, and in the light of their own ex
periences, clinicians have begun to make some attempts at preparing
clients for the therapy situation.
Pretherapy Training
Bednar et al. (1974) have reviewed literature pertaining
to the effects of pretherapy training on outcome in both individual and
group psychotherapy. The reviewers describe the group literature as
"consistent and impressive" in substantiating the effectiveness of pre
group training on outcome. The great majority of the studies were
carried out with patient populations and used outcome measures such as
personality inventories and therapist's ratings of clients' therapy
behavior. Experimental manipulations fell into one of two categories:
(1) verbal instructions; or (2) vicarious modeling.
Bednar et al. concluded that pretherapy training is effec
tive when the therapist "deliberately and systematically prepares
clients for group participation at a cognitive and behavioral level."
Apparently the pretraining affects the interaction of the group members
in a therapeutic manner so that the process of group therapy leads to
client improvement. However, the relationship between the type of


-8-
Interpersonal Trust Scale (1967). Interpersonal trust cannot be said
to exist unless the person has some intelligible communication from
another as to how the other will respond to his actions.
In a group therapy situation, how does a member come to trust, rely
on, his fellow group members? Clearly the credibility of those relative
strangers is a primary determinant (Gahagan & Tedeschi, 1968). And
such credibility can only be based on past behavior in the group of the
group members, since in a group therapy situation members typically
know little, at least initially, of each other's behavior outside the
group. In order to maximize the degree of interpersonal trust in a
group, therefore, the members need some demonstration of what behaviors
they can mutually expect under specified conditions. Interpersonal
openness as a norm is the most likely vehicle to produce such a demon
stration. Under such.conditions members of the group may discern over
a relatively brief period of time (1) to what degree each adheres to
group norms, and (2) how each is likely to respond to candid admissions
about one's feelings, experiences, or expectations. Such information
is clearly more likely to produce interpersonal trust among the members
than is the typical initial meeting of a therapy group in which social
chitchat, exchanges of demographic data, and other such "low-risk"
behaviors are the rule:
Hypothesis 1: Groups in which interpersonal openness is the
norm will develop a higher degree of interpersonal trust
than those in which no such clear norm is established.
Disclosure of Painful Secrets
Therapy patients bring with them to the group certain painful


the number of secrets shared, and in their willingness to allow the
experimenter to analyze the secrets. The groups did not differ on the
trust scale or in their estimates of how painful they considered their
secrets to be. Discussion to consensus led to a "conservative shift"
in number of secrets to be shared. It was concluded that groups can be
taught to behave in therapeutically relevant ways, that a climate of
trust can thus be built up in a group, and that such a degree of trust
generalizes to some extent. Implications for group therapy and sug
gestions for future research are discussed.
vii


-34-
For each of the dependent variables on which the groups differed
significantly, the Mann-Whitney U-Test was performed for each pair of
treatment conditions. As Table 6 shows, groups in the interpersonal
openness condition were willing to share a significantly greater number
of their secrets with other members of their group, and more of the
group members were willing to permit the experimenter to analyze their
secrets, than were groups in the other treatment conditions, who did
not differ significantly from each other. As to the mean number of
secrets which the groups agreed to share after discussion to consensus,
the significant difference (p < .05) occurred between groups in the
interpersonal-openness condition and those in the control-group
condition.
Table 6
Differences Between Pairs of Treatment Conditions on
the Three Significantly Different Dependent Variables
Treatments (in decreasing order
Dependent Variable
of the
means of the
ranks)
Secrets shared
1-0
C-G
T-G
Analysis of secrets
permitted
1-0
T-G
C-G
Secrets chosen by
consensus
1-0
T-G
C-G
Note: Treatments connected by a line are not significantly different
at the .05 level.


-33-
mean number of secrets which they were willing to share (p < .05);
mean number of group members willing to allow the experimenter to
analyze the content of their secrets (p < .01); and mean number of
secrets which the group as a group agreed to reveal after discussion
to consensus (p < .05). The groups did not differ on their individual
ratings of how painful or embarrassing they considered their secrets
to be or on their mean scores on Rotter's Interpersonal Trust Scale.
Table 5
Differences Among the Treatment Conditions on
the Five Dependent Variables
Dependent
Variable
Interpersonal
Openness
Mean Rank
Mean
Therapy
Group
Mean Rank
Mean
Control
Group
Mean Rank
Mean
H
Secrets shared
14.33
7.16
7.00
2.96
7.17
2.78
7.38*
Analysis of
secrets
permitted
15.00
4.00
7.30
2.16
6.17
2.0
10.50**
Painfulness of
secrets
(1-7)
10.50
4.80
10.25
4.68
7. 75
4.53
.98
Trust-Scale
Scores
6.22
66.75
12.25
69.53
6.92
64.66
3.02
Secrets chosen
by consensus
14.00
2.70
8.50
1.00
6.00
.37
7.67*
*p < .05
**p < .01


-17-
"After the tape you will form your own group, and we would like you
to try and interact in a manner similar to that of the videotape group.
No experimenter will participate in your group, but an observer will be
watching through the one-way mirror in order to record the interaction
variables we are examining.
"After you have seen the tape you may begin. You will have 40
minutes, and you will be notified when your time is up. Remember, it
is important that you openly discuss your feelings and immediate im
pressions of others in the group, as well as invite similar comments
about yourself.
"Next week at this same time you are to return to this room for
your second group session."
Condition 2. Therapy group. Condition 2 was an attempt to have
the six groups in that condition vicariously experience a typical in
itial meeting of a therapy group. Participants in the videotape model
were the same as those on the videotape used in Condition 1. However,
interaction among the group members in the tape was limited to the
following categories: (1) exchange of demographic data; (2) comments
about the physical environment (building, rooms, chairs, etc.);
(3) wondering aloud about how to proceed (to "get to know each other");
(4) extragroup process (Whalen's term for impersonal discussion un
related to the group or the experiment); (5) laughter, silences, grunts,
etc.
Instructions to groups in the therapy-group condition were given
as follows: "This is a group dynamics laboratory. We are interested
in examining some variables important in the development of group
interaction. To help us to study these variables we will ask you to


CHAPTER I
INTRODUCTION
Every therapist who has ever undertaken to lead a therapy group
has done so with the implicit assumption that situational variables
(behavior of the members in the group situation) can alter, modify, or
in some manner influence the characteristic behaviors of the members as
individuals, and that the alterations will be in a direction beneficial
to the client. Some clients do show remarkable improvement and dramatic
alterations in their ways of relating to other persons. But; clinical
data reveal that clients are not always so positively affected: their
symptoms may be exacerbated, their maladaptive behaviors more deeply
ingrained and resistant to change following the group experience.
What makes the difference? Experimental data are available to show
that prediction of a subject's behavior in an interpersonal situation is
more likely to be accurate when it is based on factors in the situation
itself than on personality traits of the subject. Some personality
variables that have been studied with these results are authoritarianism
(Worthy, Gary and Kahn, 1969); trust as a personality variable (Schlenker,
Helm & Tedeschi, 1973); levels of defensiveness and anxiety (Doster,
1975); self-esteem (Otis, 1975); and need for approval (Schofield, 1975).
Experimental findings such as these suggest that the group therapist can
maximize the probability of consistently beneficial effects for group
members by careful structuring of the situation.
The concept of an externally imposed structure for a group therapy
-1-


1
4
7
11
14
17
oup
1
4
7
11
14
17
APPENDIX B
Time in .5-sec. Intervals in the Six Response Categories by Rater for
the Six Groups in Each Treatment Condition
1. Personal Discussion
Interpersonal-Openness
Therapy-
Group
Control
Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
948
997
3
29
30
2
19
21
708
934
6
0
0
5
0
0
1071
1073
9
0
0
8
0
0
107
110
12
0
0
10
0
0
1395
1470
15
19
18
13
0
0
1591
1630
18
0
0
16
0
0
5820
6214
48
48
19
21
2.
Feedback
Interpersonal-Openness
Therapy-Group
Control-
-Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
0
0
3
0
0
2
48
52
30
36
6
0
0
5
6
5
493
501
9
0
0
8
0
0
0
0
12
0
0
10
4
0
70
80
15
0
0
13
0
0
46
39
18
0
0
16
0
0
639
656
0
0
58
57
i
-p-
oo
i


1
4
7
11
14
17
oup
1
4
7
11
14
17
5. Descriptive Aspects of
Communicative Speech
Interpersonal-Openness
Therapy-Group
Control-
Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
87
70
3
157
154
2
97
93
113
116
6
116
119
5
145
130
135
131
9
266
233
8
175
174
115
109
12
198
183
10
111
123
91
89
15
151
166
13
175
175
185
158
18
117
126
16
187
194
726
673
1005
981
890
889
i
Ln
O
I
6. Unscoreable Utterances
Interpersonal-Openness
Therapy-
Group
Control-
Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
0
0
3
0
0
2
25
35
46
35
6
9
0
5
0
0
41
16
9
0
0
8
0
0
12
15
12
0
0
10
0
20
0
0
15
26
26
13
0
0
_0
_0
18
63
20
16
_0
_0
99
66
98
96
25
55


-10-
secrets" as we ourselves define them. We do not give our most powerful
weapons into hands that are likely to turn them against us. One of the
functions of the therapist in a group is to protect the individual
group member who has attained such a degree of trust in the group from
revealing such secret data at a time when the other members may not be
ready to receive it. That is, when one has reached the stage of being
willing to share, the actual sharing may be anticlimactic.
Hypothesis 2: Members of groups in which interpersonal
openness is the norm will show a greater willingness to
risk disclosing painful secrets than will members of groups
in which no such norm has been established.
Establishing a Norm by Pregroup Training: Modeling
The studies by Deutsch and by Schlenker et al., indicate that
subjects in a group situation will be likely to perform in accord with
the norms of the situation. How can such a norm be established, and
how can such complex verbal behavior as "interpersonal openness" be
"built into" group members? Whalen's and Otis's research findings con
cerning the powerful effects of modeling on a subject's later behavior
can be interpreted in terms of Bandura's social learning theory, which
can be summarized as follows:
Behavior acquired through observational learning is coded into
mediating images or words which are stored until appropriate environ
mental cues suggest that performance of the observed behavior is likely
to be rewarded. The effects of modeling on the observer may be
exercised through (1) acquisition of novel responses, (2) inhibition


-36-
in minutes for groups in each treatment condition was as follows:
interpersonal-openness, 6.46; therapy-group, 3.28; control-group, 10.34.
The groups did not differ significantly at the .05 level.


TABLE OF CONTENTS
(Continued)
Page
Hypotehsis 1: Interpersonal Trust 40
The Risky Shift 41
Conclusions 43
APPENDIX A Interpersonal Trust Scale (Rotter, 1967) 45
APPENDIX B Time in .5-sec. Intervals in the Six Response
Categories by Rater for the Six Groups in Each
Treatment Condition 48
REFERENCES 51
BIOGRAPHICAL SKETCH 56
v


-15-
me." The subject areas for the secrets were taken from data on intimate
secrets obtained by Norton, Feldman & Tayofa (1974). When a subject
had finished the pretreatment phase she was randomly assigned to one
of three treatment conditions and to one of six groups of four within
that condition. An exception to the randomness was that no subject was
assigned to a group in which she was acquainted with the members.
Procedure
Condition 1. Interpersonal openness. Condition 1 was an attempt
to structure the six groups in that condition in such a way as to maxi
mize interpersonal openness. The videotape model and instructions were
adapted from those used by Whalen and described in detail in her study
as "Film-detailed instructions condition."
The videotape model consisted of a twelve-minute, untitled video
tape recording (VTR) produced for this study. Data from previous
studies (e.g., Bandura, Ross & Ross, 1963; Otis, 1975)
indicate that videotaped models are as potent as live models in elicit
ing imitative behavior. In the VTR a group of four females conversed
after implying that they had just met. Two of the participants spent
a few seconds conveying biographical information, and then a third
member, interrupting, maintained that the group members were not using
the best method of getting to know each other. For the remainder of
the time the participants talked on a more personal level, describing
their anxieties and other feelings, and feeding back their impressions
whether positive, negative, or neutral, of the other group members.
An attempt was made to make the VTR group appear as similar as possible
in composition to the subject groups, e.g., in age, sex, and dress.


I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
'Y } 7-
Marvin E. Shaw
Professor of Psychology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
This dissertation was submitted to the Graduate Faculty of the Department
of Clinical Psychology in the College of Arts and Sciences and to the
Graduate Council, and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
March 1977
Dean, Graduate School


-13-
generalizability of such studies on the basis that people are very poor
predictors of their own behavior.
Those few recent studies using designs having real consequences for
the subjects have obtained mixed results (Blascovich, Ginsberg & Veach,
1975; Blascovich, Veach & Ginsberg, 1973; McCauley, Stitt, Woods &
Lipton, 1973; Yinon, 1975). It seems reasonable at this point that
theoristic explanation should wait upon further research involving con
sequences that are real and personal to the subjects.
The present study has as one of its aims an investigation of shift
in risk-taking as it pertains to the actual behavior of group members
in a situation where they have much to lose in the form of sharing
painful aspects of themselves with relative strangers. A shift in the
"risky" direction would presumably indicate an expected positive outcome
to the group members of sharing their secrets which would outweigh an
expected loss, the possibility of embarrassment or ridicule.


-21-
unless you decided on your own to share them.' That is still
true. However, if you have no objections, I would like to
analyze the content of your secrets. This could be done by
your tearing open both your envelopes and dumping the contents
into a bag with other secrets so that 1 would never know to
whom a secret belonged. However, this is not an integral part
of the experiment and I want you to feel free to refuse if you
have any reservations."
Dependent Variables
The dependent measures consisted of (1) mean score for each group
on the Interpersonal Trust Scale; (2) behavioral data in the form of
mean number of secrets per group selected to share with the group;
(3) ratings of the total verbalizations of each group on a revised
version of Whalen's rating scale; and (4) differences between (a) mean
number of secrets chosen for sharing by group members prior to dis
cussion to consensus and (b) number agreed upon by each group.
Interpersonal openness rating scale. For purposes of this study
the six broad response categories of Whalen's scheme were used. They
are described in detail below. The descriptions following each response
category were taken as criteria upon which the judges based their
ratings.
Response categories:
1. Personal discussion. (a) Personal self-disclosure: Discussion
of feelings, attitudes, and behaviors unrelated to current group inter
action which are non-public, not ordinarily readily volunteered, and


To Michael, Moetke,
and Megan, who
didn't understand but
loved me anyway,
and
To Jay,
for all the days and hours.


REFERENCES
Alexander, F. Psychoanalytic therapy: Principles and applications.
New York: Ronald Press, 1946.
Allen, J.G. Implications of research in self-disclosure for group
psychotherapy. International Journal of Group Psychotherapy,
1973, 23, 306-321.
Bandura, A. Principles of behavior modification. New York: Holt,
Reinhart, 1969.
Bandura, A., Ross, D., & Ross, S. Imitation of film-mediated aggressive
models. Journal of Abnormal and Social Psychology, 1963, 66,
3-11.
Bednar, R.L., Weet, C., Evensen, P., Lanier, D., & Melnick, J. Empirical
guidelines for group therapy: Pretraining, cohesion and modeling.
Journal of Applied Behavioral Science, 1974, 10_ (2), 149-165.
Blascovich, J., Ginsberg, G.P., & Veach, T.L. A pluralistic explanation
of choice shifts on the risk dimension. Journal of Personality
and Social Psychology, 1975, JU, 422-429.
Blascovich, J., Veach, T.L., & Ginsberg, G.P. Blackjack and the risky
shift. Sociometry, 1973, _36, 42-45.
Bradford, L.P. Membership and the learning process. In L.P. Bradford,
J.R. Gibb & K.D. Berne (Eds.), T-group therapy and laboratory
method. New York: Wiley, 1964.
Broadbent, D. Attention and the perception of speech. In Perception:
Mechanisms and models. San Francisco: W.H. Freeman, 1972.
Cozby, P.C. Self-disclosure: A literature review. Psychological
Bulletin, 1973, 79, 73-91.
Deutsch, M. Cooperation and trust: Some theoretical notes. Nebraska
symposium on motivation, 1962.
Dibner, A.S. Ambiguity and anxiety. Journal of Abnormal and Social
Psychology, 1958, 56, 165-174.
Dion, K., Baron, R., & Miller, N. Why do groups make riskier decisions
than individuals? In L. Berkowitz (Ed.), Advances in Experimental
Psychology, vol. 5. New York: Academic Press, 1970.
-51-


-32-
100
| Models
90
80
70
60
c
QJ
U
u
0)
P-.
50
40
30
20
10
0
Pers Feedback Impers
Disc Disc
|jjfl Group
Members
Group Descrip Unscore
Process
Figure 3. Comparison of models and groups in the control-group
condition on percent of time spent in each of the six
categories of response.


-42-
a conservative direction following discussion to consensus on number of
secrets to be shared. However, after discussion the difference between
groups in the interpersonal-openness condition and those in the therapy-
group condition disappeared. Groups in the control-group condition
were not significantly different from those in the therapy-group
condition.
The reasons for such a uniformly conservative shift are difficult
to discover. Each group member decided on her own to reveal a certain
number of secrets. Yet, apparently she was not willing to impose on
other group members a greater degree of risk than she had accepted for
herself. One could also theorize that the less trusting members of the
group prevailed; an element of distrust in the person of one reluctant
member might have dissipated the atmosphere of trust built up in the
group. In sum, being unsure of the reasons for a member or members'
reluctance to disclose her self to any degree, the other members may
have concluded (a) that they should not try to persuade her (or them)
otherwise; or (b) that perhaps that person or persons was not really
so trustworthy as they had thought. Certainly, in therapy groups a
client is often heard to say, after the session is over, that he or she
had planned to talk about such-and-such a topic, but had lost the courage
to do so when other members seemed reluctant to expose themselves.
Unlike studies using the Choice Dilemma Questionnaire, the present
study involved a real and personal risk for the group members. It would
seem that the gain to be expected from taking such a risk was, in this
case, outweighed by other factors which became operative only in the
context of the discussion to consensus. What those factors were is a
matter of speculation. However, further research involving analysis of


TRUST AND RISK-TAKING AS A FUNCTION OF
PRE-GROUP TRAINING
By
SALLYE B. PETERSON
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1977


CHAPTER II
METHOD AND PROCEDURES
Sub j ects
The subjects were 72 female undergraduate students enrolled in
introductory psychology courses at the University of Florida, who
volunteered to participate in order to fulfill one of the requirements
of the course. The subjects were told that two experimental sessions,
one week apart, would be held and that participation in both was re
quired. Any subject who indicated she would be unable to attend the
second session was eliminated from the study.
As a pretreatment condition each subject was seen individually in a
private room and given written instructions as follows: "Here are 20
slips of paper and an envelope. You are to write on each slip of paper a
secret about yourself that you would not be willing to share with a group
of female strangers. For most people such secrets center around subjects
such as sex, violence, failures, loneliness, mental health, stealing,
cheating, and so on. However, these are only examples, and you should
not be limited to them. After you have finished, fold the slips of pa
per, place them in the envelope, and seal it. Then write your name across
After you are
finished, I will lock your envelope in this file cabinet. I have the
only key, and iio one will ever read what you have written unless you
decide on your own to show it to them. The envelope will be returned
to you after the study is over. When you have finished, please call
the flap of the envelope like this.
-14-


-44-
Whalen's rating scale as modified in the present study has shown
itself to be highly reliable; the raters reported no great difficulty
in distinguishing among the categories. Hence it would seem to be an
instrument of choice for further research in this area.
No conclusions can be drawn regarding the reasons for the "con
servative shift" obtained from our subjects. The present research
fortunately included videotaping of the actual discussion-to-consensus
sessions. Analysis of the group interaction in such sessions would
seem to be a natural outgrowth of this study and perhaps have a bearing
on the interactional aspects of the risky shift phenomenon.


-24-
goof them up."
5. Descriptive aspects of communicative speech: (a) Agreement:
Any verbal indication of agreement, for example, "Uh huh," and "Yeah,
that's right." Head nods are not scored. (b) Disagreement: Any
verbal indication of disagreement, for example, "NO," and "That's not
true!" Head nods or shakes are not scored. (c) Laughter: Any amount
of audible giggling or laughing by one or more group members. Smiles
and grins are not scored. (d) Silence: This category is scored only
if the duration of the silent period exceeds 3 seconds. (e) Inter
ruption : Scored whenever one group member stops another from talking
or when two or more individuals are speaking simultaneously.
6. Unscoreable utterances. A scoring category for those remarks
which are either inaudible, unintelligible, or too difficult to
categorize immediately.
Rating procedure. The middle twenty minutes of discussion for
each of the eighteen groups was rated independently by two trained
raters using an Esterline-Angus event recorder. The resultant scores
described total group interaction in terms of number of seconds spent
in discussion rated as falling into each of the response categories
(no ratings of individual members).
The raters were advanced graduate students in clinical psychology
who were trained on practice tapes to a criterion of 94% agreement on
a tape. After training, the raters also analyzed the three modeling
tapes in terms of time spent in each of the six response categories.


-52-
Doster, J.A. Individual differences affecting interviewee expectancies
and perceptions of self-disclosure. Journal of Counseling
Psychology, 1975, 2/2, 192-198.
Freed, A., Chandler, P.J., Blake, R.R., & Mouton, J. Stimulus and back
ground factors in sign violation. Journal of Personality, 1955,
23, 499.
Freedman, J.L. Role playing: Psychology by consensus. Journal of
Personality and Social Psychology, 1969, 13, 107-114.
Gahagan, J.P., & Tedeschi, J.T. Strategy and the credibility of promises
in the Prisoner's Dilemma game. Journal of Conflict Resolution,
1968, 12, 224-234.
Gendlin, E.T. A theory of personality change. In J.T. Hart & T.M.
Tomlinson (Eds.), New directions in client-centered therapy.
Boston: Houghton Mifflin, 1970.
Goldstein, A.P., Heller, K., & Sechrest, L.B. Psychotherapy and the
psychology of behavior change. New York: Wiley, 1966.
Grosser, D., Polansky, N., & Lippit, R. A laboratory study of behavioral
contagion. Human Relations, 1951, 4^, 115-142.
Hollander, E.P. Conformity, status and idiosyncrasy credit.
Psychological Review, 1958, 6J3, 117-127.
Jellison, J.M., & Riskind, J. A social comparison of abilities inter
pretation of risk-taking behavior. Journal of Personality and
Social Psychology, 1970, 1_5, 375-390.
Jellison, J.M., & Riskind, J. Attribution of risk to others as a
function of their ability. Journal of Personality and Social
Psychology, 1971, 20, 413-415.
Johnson, J.A. Group therapy: A practical approach. New York: McGraw-
Hill, 1963.
Jourard, S.M. The transparent self. New York: Van Nostrand, 1971.
Kahan, J.P. A subjective probability interpretation of the risky shift.
Journal of Personality and Social Psychology, 1975, 31 (6),
977-982.
Kimbrell, D., & Blake, R.R. Motivational factors in the violation of
a prohibition. Journal of Abnormal and Social Psychology, 1958,
56, 132-133.


-22-
which may make the individual vulnerable to negative evaluations from
others. Examples of actual comments categorized as personal self
disclosure by Whalen's raters are "I'm a status seeker like a lot of
people," "I might even try homosexuality," or "I just broke down and
started yelling and screaming." (b) Immediate feelings: Discussion of
feelings and attitudes related to current group interaction which are
relatively private, not ordinarily volunteered by most people, and
which may make the individual vulnerable to negative evaluations from
others (e.g., "I feel alienated here" or "Now I'm embarrassed").
(c) Personal questions: Questions relating to an individual's opinions,
feelings, or actions which are usually considered to be private in
formation: "Have you always been heavy?" or "I wondered how you felt
when we were so silent."
2. Feedback. (a) Positive feedback: The conveying of an indi
vidual's positively valenced impressions of or reactions to the comments,
appearance, or actions of another group member. This category includes
compliments, flattery, etc. Examples are "I like the way you said
that," or "You seem like an alert, intelligent person." (b) Negative
feedback: Negative evaluation of the personality, intelligence,
actions, etc. of another group member. This category includes hos
tility, ridicule, criticism and implications that the individual is not
telling the truth. Examples of comments rated as negative feedback
are "You were a fool," "Big deal!" or "Your argument isn't very ration
al." (c) Neutral feedback: This category includes comments about a
group member that cannot be evaluated as either positive or negative,
as well as direct advice from one group member to another. Some
examples are "You seem like the silent type" or "You ought to go up


-4 7-
23. Most experts can be relied upon to tell the truth about the limits
of their knowledge.
24. Most parents can be relied upon to carry out their threats of
punishment.
25. One should not attack the political beliefs of other people.
26. In these competitive times one has to be alert or someone is likely
to take advantage of you.
27. Children need to be given more guidance by teachers and parents
than they now typically get.
28. Most rumors usually have a strong element of truth.
29. Many major national sport contests are fixed in one way or another.
30. A good leader molds the opinions of the group he is leading rather
than merely following the wishes of the majority.
31. Most idealists are sincere and usually practice what they preach.
32. Most salesmen are honest in describing their products.
33. Education in this country is not really preparing young men and
women to deal with the problems of the future.
34. Most students in school would not cheat even if they were sure of
getting away with it.
35. The hordes of students now going to college are going to find it
more difficult to find good jobs when they graduate than did the
college graduates of the past.
36. Most repairmen will not overcharge even if they think you are
ignorant of their specialty.
37. A large share of accident claims filed against insurance companies
are phony.
38. One should not attack the religious beliefs of other people.
39. Most people answer public opinion polls honestly.
40. If we really knew what was going on in international politics, the
public would have more reason to be frightened than they now seem
to be.


-26-
Analysis of the Modeling Tapes
To determine whether the three 12-minute tapes used in the three
treatment conditions actually depicted the types of verbal interaction
which they were designed to model, the judges rated each of the tapes
separately. Each verbalization was rated as falling into one of the
six response categories.
The modeling tape for the interpersonal-openness condition was
designed to depict primarily the categories of personal discussion and
feedback, since those were the types of verbalizations which the inter-
personal-openness groups were supposed to model. Tapes for the therapy-
group and control-group conditions were designed to model primarily
impersonal discussion. As can be seen from Table 2, the tapes did
model the types of verbal interaction for which they were designed.
(The interpersonal-openness condition tape included some personal dis
cussion as an example of the kind of interaction which was not to be
modeled by the subjects.)
Table 2
Analysis of Modeling Tapes: Time (in seconds) and Percent of
Total Time Spent in Each of the Six Response Categories
Response
Category
Inter
personal
Openness
%
"Therapy"
Group
%
Control
Group
/
Personal Discussion
417
58
0
0
0
0
Feedback
83
12
0
0
0
0
Impersonal
Discussion
87.5
12
680.5
95
575
80
Group Process
78.5
11
0
0
0
0
Descriptive aspects
of Communicative
Speech
54.0
07
39.5
5
145
20
Unscoreable Utterances
0
0
0
0
0
0
720.0
100
720.0
100
720.0
100


-27-
Analysis of the Subjects' Interaction in the Groups
From the forty minutes of interaction had by each group, a video
tape of the middle twenty minutes was analyzed by the raters with each
verbalization judged as falling into one of the six response categories
(personal discussion, feedback, impersonal discussion, group process,
descriptive aspects of communicative speech, and unscoreable utterances).
The Kruskal-Wallis one-way analysis of variance was used to determine
whether significant differences existed among the groups in the three
treatment conditions with regard to the number of seconds spent in each
category. Results of the analysis showed that the groups differed
among themselves significantly in three of the response categories:
personal discussion (p < .01), feedback (p < .05), and impersonal dis
cussion (p < .01). Table 3 shows the results of the analysis.
For each of the response categories in which a significant dif
ference among the groups was found, the Mann-Whitney U Test was performed
to determine where the differences lay. As can be seen from Table A,
groups in the interpersonal-openness condition spent a significantly
greater amount of their time in personal discussion and a significantly
lesser amount of time in impersonal discussion than did groups in the
therapy-group and control-group conditions, which did not differ sig
nificantly from each other. In the category of feedback, the only
significant difference was between the interpersonal-openness and the
therapy-group conditions.


-40-
the interpersonal-openness condition were willing to permit the ex
perimenter to analyze the content of their secrets. The readiness to
allow a stranger to read what one had previously identified as secrets
one would not be willing to share, even when promised anonymity,
indicates at least a state variable of being open about one's self.
It is also worth emphasizing that there was no pressure on any of the
subjects to permit analysis of the content of their secrets; indeed,
one-third of the subjects in the therapy-group and in the control-group
conditions declined to allow such analysis.
Hypothesis 1: Interpersonal Trust
Members of groups in the three treatment conditions did not differ
significantly on the variable of interpersonal trust as measured by
Rotter's Scale. The means for all of the groups were well within one
standard deviation of the mean for the standardization group of females
and also for a group of 411 females at Ohio University recently studied
by Wright & Tedeschi (1975), indicating that the population from which
our samples were drawn does not diverge markedly from the norms for
this scale.
Rotter speaks of interpersonal trust as measured by his scale as
a "generalized expectancy that the oral or written statements of other
people can be relied upon." In thus describing this variable he dif
ferentiates it from expectancies specific to the situation in which an
individual finds himself. In other research such a differentiation is
sometimes described in terms of "trait" versus "state" variables. Thus
Rotter would predict that scores on his scale would remain relatively
stable over time, since generalized expectancies are based on long-term


TRUST AND RISK-TAKING AS A FUNCTION OF
PRE-GROUP TRAINING
By
SALLYE B. PETERSON
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1977

To Michael, Moetke,
and Megan, who
didn't understand but
loved me anyway,
and
To Jay,
for all the days and hours.

ACKNOWLEDGEMENTS
Research is not a solitary activity. I extend my deepest gratitude
to those who have supported me in this work.
To Hugh Davis, who served as Chairman, friend, and counsel in times
of despair.
To Jim Dixon, for his philosophy and inspiration.
To Nate Perry, for his support and positive criticism.
To Barry Schlenker, for his pragmatic suggestions and friendship.
To Marvin Shaw, for his encouragement and consultation.
And to Otto von Mering, for his personal interest and support.
Thanks are also due to Mark Barnett and Randy Thomas, who spent
long hours in rating the videotapes. And to Nancy Monroe, who assisted
in the painful task of data reduction.
iii

TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS . iii
ABSTRACT vi
CHAPTER I INTRODUCTION. 1
Pretherapy Training 3
Interpersonal Trust 6
Disclosure of Painful Secrets 8
Establishing a Norm by Pregroup Training: Modeling 10
The Risky Shift 12
CHAPTER II METHOD AND PROCEDURES 14
Subjects 14
Procedure 15
Dependent Measures. 21
CHAPTER III RESULTS 25
Interrater Reliability 25
Analysis of the Modeling Tapes 26
Analysis of the Subjects' Interaction in the Groups 27
Comparison of Groups and Their Models 29
Dependent Measures 29
CHAPTER IV DISCUSSION 37
Hypothesis 3: The Differential Effects of Modeling 37
Hypothesis 2: Willingness to Disclose Painful
Secrets 39
iv

TABLE OF CONTENTS
(Continued)
Page
Hypotehsis 1: Interpersonal Trust 40
The Risky Shift 41
Conclusions 43
APPENDIX A Interpersonal Trust Scale (Rotter, 1967) 45
APPENDIX B Time in .5-sec. Intervals in the Six Response
Categories by Rater for the Six Groups in Each
Treatment Condition 48
REFERENCES 51
BIOGRAPHICAL SKETCH 56
v

Abstract of Dissertation Presented to the Graduate Council
of the University for Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
TRUST AND RISK-TAKING AS A FUNCTION OF
PRE-GROUP TRAINING
By
Sallye B. Peterson
March 1977
Chairman: Hugh Davis, Jr.
Major Department: Clinical Psychology
Interpersonal trust and the degree to which subjects are willing
to share painful aspects of themselves were investigated as a function
of pretraining using models. Seventy-two female college sophmores were
instructed to write out privately 20 of their most painful secrets.
They were divided into eighteen four-member groups in three treatment
conditions and exposed to instructions and videotape models corresponding
to an interpersonal-openness condition, a therapy-group condition, and
a control condition. Each group then had 40 minutes of interaction and
returned one week later for what they believed to be their second group
session. At that time they chose a number of secrets to be shared with
their group and also filled out an interpersonal-trust scale. They
also discussed as a group and achieved consensus on the number of
secrets each member should share, as an investigation of the risky-shift
phenomenon. Raters analyzed the videotaped first session of each group
using six response categories. Interpersonal-openness groups differed
significantly from the other groups in their verbal interaction, in
vi

the number of secrets shared, and in their willingness to allow the
experimenter to analyze the secrets. The groups did not differ on the
trust scale or in their estimates of how painful they considered their
secrets to be. Discussion to consensus led to a "conservative shift"
in number of secrets to be shared. It was concluded that groups can be
taught to behave in therapeutically relevant ways, that a climate of
trust can thus be built up in a group, and that such a degree of trust
generalizes to some extent. Implications for group therapy and sug
gestions for future research are discussed.
vii

CHAPTER I
INTRODUCTION
Every therapist who has ever undertaken to lead a therapy group
has done so with the implicit assumption that situational variables
(behavior of the members in the group situation) can alter, modify, or
in some manner influence the characteristic behaviors of the members as
individuals, and that the alterations will be in a direction beneficial
to the client. Some clients do show remarkable improvement and dramatic
alterations in their ways of relating to other persons. But; clinical
data reveal that clients are not always so positively affected: their
symptoms may be exacerbated, their maladaptive behaviors more deeply
ingrained and resistant to change following the group experience.
What makes the difference? Experimental data are available to show
that prediction of a subject's behavior in an interpersonal situation is
more likely to be accurate when it is based on factors in the situation
itself than on personality traits of the subject. Some personality
variables that have been studied with these results are authoritarianism
(Worthy, Gary and Kahn, 1969); trust as a personality variable (Schlenker,
Helm & Tedeschi, 1973); levels of defensiveness and anxiety (Doster,
1975); self-esteem (Otis, 1975); and need for approval (Schofield, 1975).
Experimental findings such as these suggest that the group therapist can
maximize the probability of consistently beneficial effects for group
members by careful structuring of the situation.
The concept of an externally imposed structure for a group therapy
-1-

-2-
situation is antithetical to the traditional conceptualization of the
manner in which groups function; that is, that an ambiguous, unstructured
situation is most conducive to client self-exploration and the natural
development of the group (cf. Rabin, 1970). The necessity for ambiguity
in therapy groups is derived from theories that stress the importance
of a therapeutic relationship (individual as well as group) being unlike
the client's historical relationships in which he has learned to behave
in maladaptive ways (Singer, 1970). i
But the experimental literature is replete with data on the detri
mental effects of ambiguous situations on behavior, specifically the
relationship between ambiguity and anxiety (Lazarus, 1966). A classic
study by Dibner (1958) compared the amount of anxiety experienced by
patients who were prepared for a psychiatric interview (told how to
behave and what to expect) with that experienced by unprepared patients
(highly ambiguous situation). Behavioral data, subjective reports, and
GSR data from the patients all demonstrated that unprepared patients
experienced a significantly greater degree of anxiety. The destructive
effects of anxiety induced by ambiguity are also documented by studies
of perceptual acuity (Postman & Bruner, 1948) and visual organization
and synthesis of cues (Verville, 1946). When attention is divided, or
when the individual has difficulty knowing which elements of a situation
to attend to, the perception of speech is likely to be confused and
distorted (Broadbent, 1972). And the destructive chain continues with
distorted perceptions leading to increased anxiety (Korchin et al., 1958).
On the basis of their experience alone group therapists can testify
to the evident anxiety, inappropriate behavior, and desperate looking to
the therapist for help that occur in the initial meetings of groups.

-3-
Johnson (1963) and Yalom (1966) point out two additional experiences
familiar to group therapists, i.e., that norms and behavior patterns
established in early sessions tend to persist throughout the life of
the group, and that as many as one-third of the group members drop out
during the early (1st to 12th) group meetings. In the face of accumu
lating evidence of the deleterious effects of situational ambiguity on
the behavior of clients in therapy, and in the light of their own ex
periences, clinicians have begun to make some attempts at preparing
clients for the therapy situation.
Pretherapy Training
Bednar et al. (1974) have reviewed literature pertaining
to the effects of pretherapy training on outcome in both individual and
group psychotherapy. The reviewers describe the group literature as
"consistent and impressive" in substantiating the effectiveness of pre
group training on outcome. The great majority of the studies were
carried out with patient populations and used outcome measures such as
personality inventories and therapist's ratings of clients' therapy
behavior. Experimental manipulations fell into one of two categories:
(1) verbal instructions; or (2) vicarious modeling.
Bednar et al. concluded that pretherapy training is effec
tive when the therapist "deliberately and systematically prepares
clients for group participation at a cognitive and behavioral level."
Apparently the pretraining affects the interaction of the group members
in a therapeutic manner so that the process of group therapy leads to
client improvement. However, the relationship between the type of

-4-
training chosen and the process variable studied, and between process
and outcome is unclear. For example, it is difficult to conceptualize
how listening to excerpts of "good" patient in-therapy behavior can lead
to decreased MMPI scores (Truax & Wargo, 1969), especially when no
definition of "good" behavior is given.
Whalen (1969) conducted a study on pregroup training with a popula
tion of college students. In her study she clearly defined the variables
which she was seeking to manipulate: particular types of verbal inter
action, e.g., personal discussion and feedback. Whalen's study was
limited to the effects of pretraining on group interaction, and no
"outcome" variables were measured.
What seem to be lacking in the research are studies in which the
critical factors in the group situation are defined and some relation
ship is established among pretraining, group process, and outcome.
Obviously the parameters of the group situation seen as crucial to
client improvement depend on one's theoretical orientation; the critical
process and outcome variables and their relationship to factors in the
group situation are likewise dependent on one's conception of how
therapeutic change takes place; and the most effective means of in
fluencing the crucial group parameters depends on the researcher's
conceptions in the areas stated above plus one's evaluation of the
evidence as to how learning occurs.
Critical factors in the group situation. Most modern systems of
psychotherapy are interpersonally based. Stemming from the basic theses
propounded by Harry Stack Sullivan in his interpersonal theory of
psychiatry, current psychotherapies view the origins of psychopathology
as existing in disordered interpersonal relations; thus, the appropriate

-5-
raethod of treatment is also seen as essentially interpersonal. Even
behavior theorists such as Wolpe regard "improved interpersonal
relationships" as among the primary criteria for evaluating therapeutic
change (Wolpe, 1969). Thus it would appear that the major characteris
tic of the group over which the therapist must exert some control is
that of the ways in which the members behave toward each other. Such
a statement seems both simplistic in its obviousness and antagonistic
to the pusposes of group therapy: the idea that the therapist can or
should control the members' behavior. Yet, when no consistent in
fluences are brought to bear on the members' behavior in the situation
one may observe the relative lack of structure and norm-ambiguity which
lead to heightened anxiety and low risk-taking on the part of the group
members.
What particular aspects of interpersonal relations have therapists
considered relevant? One of the most frequently cited is what might be
summarized as "interpersonal openness," or the "readiness to convey
personal information about one's life, one's feelings, and one's im
pressions of other individuals" (Whalen, 1969, p. 510). Rogers (1959)
emphasizes the importance of such behavior for both client and therapist,
as do Jourard (1971), Gendlin (1970), and Mowrer (1964), and the en
counter movement therapists (e.g., Bradford, 1964; Schtz, 1973). Even
analytically oriented therapists such as Franz Alexander (1946) have
recognized the importance of the client's experiencing in the present
and honestly reporting that experience in order to undergo a "corrective
emotional experience."
What are the factors which might contribute to a group situation
in which the members consistently behave in an interpersonally open

-6-
fashion? The question of whether setting of norms can facilitate
specific behaviors has been investigated with uniformly positive results.
Ribner (1974) found that groups receiving an explicit group contract
calling for self-disclosure increased significantly both their frequency
and depth of self-disclosure.
The clearest demonstration of building into a group a norm of
interpersonal openness is Whalen's study, in which the combination of a
film demonstration modeling interpersonal openness and detailed instruc
tions of expected behavior produced significantly greater effects than
either independent variable alone. In explaining her results Whalen
relied on an attention-information format detailed by Thibaut & Kelley
(1959). The norm-sending process involved rule statements (filmed
modeling plus instructions), surveillance-evaluation (presence of other
group members), and application of sanctions (applied by the subject as
well as the other group members). Such a conceptualization stresses
the learning of complex verbal response classes, a type of learning
requiring that the subject attend to the relevant variables in the
situation rather than simply imitate a discrete response.
Interpersonal Trust
How can interpersonal openness contribute to the therapeutic effec
tiveness of the group for its members? One of the means by which group
therapy is considered to bring about changes in an individual's habits
and attitudes is through allowing him to try out new and different ways
of relating to others in the group. Yalom (1970) has emphasized the
importance of group feedback in modifying interpersonal distortions and

-7-
clarifying interpersonal relationships. However, an individual is un
likely either to risk such new behavior or, once having done so, to
accept the feedback he receives except where group members share a
fairly high degree of interpersonal trust.
Deutsch (1962) defined a trusting situation, as differentiated from
a gambling situation, as one in which one "has much to lose or little
to gain," probably the view most new group members hold of the therapy
group. He reasoned that one needs "considerable confidence in a posi
tive outcome" to trust. From his research on cooperation, competition,
and trust, Deutsch defined trust in motivational terms: cooperative or
competitive motives will determine the subject's behavior in the situa
tion; but an individualistic orientation, neither cooperating nor
competing but simply trying to maximize one's gains, resulted in the
subject's behavior being much more influenced by the specific experi
mental conditions. In terms of Deutsch's model, if a group member
exposes himself to possible criticism, ridicule, rejection, or even
potential blackmail by other members, such behavior is indicative of
trust "much to lose." Whether he engages in such behavior will be
strongly influenced by situational variables; in this case the behavior
of the other group members.
Schlenker, Helm, and Tedeschi (1973) offer a definition of inter
personal trust as "a reliance upon information received from another
person about uncertain environmental states and their accompanying
outcomes in a risky situation" (emphasis supplied). In that definition
the importance of situational variables in the operation of interpersonal
trust is even more central. And Rotter (1967, 1971) likewise emphasized
reliance upon communications from other individuals in developing his

-8-
Interpersonal Trust Scale (1967). Interpersonal trust cannot be said
to exist unless the person has some intelligible communication from
another as to how the other will respond to his actions.
In a group therapy situation, how does a member come to trust, rely
on, his fellow group members? Clearly the credibility of those relative
strangers is a primary determinant (Gahagan & Tedeschi, 1968). And
such credibility can only be based on past behavior in the group of the
group members, since in a group therapy situation members typically
know little, at least initially, of each other's behavior outside the
group. In order to maximize the degree of interpersonal trust in a
group, therefore, the members need some demonstration of what behaviors
they can mutually expect under specified conditions. Interpersonal
openness as a norm is the most likely vehicle to produce such a demon
stration. Under such.conditions members of the group may discern over
a relatively brief period of time (1) to what degree each adheres to
group norms, and (2) how each is likely to respond to candid admissions
about one's feelings, experiences, or expectations. Such information
is clearly more likely to produce interpersonal trust among the members
than is the typical initial meeting of a therapy group in which social
chitchat, exchanges of demographic data, and other such "low-risk"
behaviors are the rule:
Hypothesis 1: Groups in which interpersonal openness is the
norm will develop a higher degree of interpersonal trust
than those in which no such clear norm is established.
Disclosure of Painful Secrets
Therapy patients bring with them to the group certain painful

-9-
secrets which they regard as (1) extremely important to their present
problems and (2) too dangerous or embarrassing to reveal to anyone else.
With the possible exception of psychoanalysts, most therapists do not
regard the client's revelation of such secrets as crucial to his
improved functioning. Thus, "self-disclosure" to the extent of re
vealing in detail one's greatest worries, shames, and fears has not been
shown to be an effective or even desirable goal of psychotherapy.
Although Jourard (1971) and other humanistic psychologists have
maintained that "Self-disclosure is a symptom of personality health and
a means of ultimately achieving healthy personality," research findings
have been contradictory. Cozby (1973) pointed out in his review of the
self-disclosure literature that one reason for the difficulty in com
paring results across studies is the lack of consistency in defining
"mental health." And Allen (1973), reviewing the methodological
limitations of self-disclosure research, has listed as primary the fact
that the research is based on relatively nonintimate disclosures. Thus,
the extension of findings from such studies requires substantial con
ceptual leaps. Particularly with regard to groups there is reason to
question the therapeutic benefits to be gained from self-disclosure,
since the majority of findings are based on self-report or dyadic
interactions. Finally, "self-disclosure" has been defined in virtually
all of the research in terms of the Jourard questionnaire or variants
of it (Allen, 1973), and the validity of these measures has yet to be
established.
Despite the lack of evidence establishing the therapeutic relevance
of actual self-disclosure thus defined, an index of the extent to which
we trust each other is our willingness to have them know our "painful

-10-
secrets" as we ourselves define them. We do not give our most powerful
weapons into hands that are likely to turn them against us. One of the
functions of the therapist in a group is to protect the individual
group member who has attained such a degree of trust in the group from
revealing such secret data at a time when the other members may not be
ready to receive it. That is, when one has reached the stage of being
willing to share, the actual sharing may be anticlimactic.
Hypothesis 2: Members of groups in which interpersonal
openness is the norm will show a greater willingness to
risk disclosing painful secrets than will members of groups
in which no such norm has been established.
Establishing a Norm by Pregroup Training: Modeling
The studies by Deutsch and by Schlenker et al., indicate that
subjects in a group situation will be likely to perform in accord with
the norms of the situation. How can such a norm be established, and
how can such complex verbal behavior as "interpersonal openness" be
"built into" group members? Whalen's and Otis's research findings con
cerning the powerful effects of modeling on a subject's later behavior
can be interpreted in terms of Bandura's social learning theory, which
can be summarized as follows:
Behavior acquired through observational learning is coded into
mediating images or words which are stored until appropriate environ
mental cues suggest that performance of the observed behavior is likely
to be rewarded. The effects of modeling on the observer may be
exercised through (1) acquisition of novel responses, (2) inhibition

-11-
or disinhibition of previously learned responses, depending on the
consequences which the model is observed to experimence, or (3) facili
tation of responses through focusing of the observer's attention on
discriminative stimuli in the environment (Bandura, 1969).
Especially in the case of vicarious learning of complex verbal
response patterns, the observer's attention must be focused on the
essential elements of the model's behavior. As Whalen pointed out, what
the subject must learn is a set of rules and regulations, rather than
discrete responses; hence the importance of orienting instructions to
the observer in facilitating the coding of verbal mediators. The effects
of observational learning are maximized by such variables as a moderate
degree of similarity of the model to the observer and the observer's
arousal and motivational levels.
There is experimental evidence in support of Bandura's theory as
it relates to observational learning of responses which deviate from
the patterns of behavior which the observer would ordinarily expect in
the situation in which the model is viewed. Apparently, the disinhibit-
ing effects of the modeling occur through (1) establishing a norm for
such behavior; and (2) providing discriminative stimuli that inform the
observer as to when the norm is in operation (Freed, Chandler, Blake &
Mouton, 1955; Grosser, Polansky & Lippit, 1951; Kimbrell & Blake, 1958;
Walters 6 Parke, 1964).
According to Bandura's theory of observational learning and the
experimental evidence supporting that theory, Hypothesis 3 is stated as
follows:
Hypothesis 3: Groups of individuals exposed to explicit,
detailed instructions and to a videotape of a group in which

-12-
the members model Interpersonal openness will engage in
significantly more interpersonally open behavior than will
groups who do not receive such pretraining.
In summary, it was hypothesized that pretraining via modeling and
detailed instructions would lead to greater interpersonal openness on
the part of group members; that such openness would in turn result in
more interpersonal trust among the members; and that a greater degree
of interpersonal trust would also be reflected in a greater willingness
to risk disclosure of painful secrets to the group.
The Risky Shift
Studies using the Choice Dilemma Questionnaire developed by Kogan
and Wallach (1964) have often found a shift in decision-making in a
"riskier" direction following group discussion. However, the shift is
not uniform for all items (some shifting consistently in a more con
servative direction), and stronger risky shifts are associated with
items having the least serious consequences (Dion, Baron & Miller,
1970). Explanations of the phenomenon have been many and varied, in
cluding the decision-making model of Pruitt (1971) ; a social-comparison
interpretation (Jellison & Riskind, 1970, 1971) ; and the expected-
utility model of Vinokur (1971), recently amplified in the subjective
expected utility model of Kahan (1975) .
As pointed out by Dion et al., however, a strong reservation in
interpreting the results of any risky-shift study using the CDQ is the
extent to which the tasks lack any (much less severe) consequences for
the group members. And Freedman (1969) argues cogently against the

-13-
generalizability of such studies on the basis that people are very poor
predictors of their own behavior.
Those few recent studies using designs having real consequences for
the subjects have obtained mixed results (Blascovich, Ginsberg & Veach,
1975; Blascovich, Veach & Ginsberg, 1973; McCauley, Stitt, Woods &
Lipton, 1973; Yinon, 1975). It seems reasonable at this point that
theoristic explanation should wait upon further research involving con
sequences that are real and personal to the subjects.
The present study has as one of its aims an investigation of shift
in risk-taking as it pertains to the actual behavior of group members
in a situation where they have much to lose in the form of sharing
painful aspects of themselves with relative strangers. A shift in the
"risky" direction would presumably indicate an expected positive outcome
to the group members of sharing their secrets which would outweigh an
expected loss, the possibility of embarrassment or ridicule.

CHAPTER II
METHOD AND PROCEDURES
Sub j ects
The subjects were 72 female undergraduate students enrolled in
introductory psychology courses at the University of Florida, who
volunteered to participate in order to fulfill one of the requirements
of the course. The subjects were told that two experimental sessions,
one week apart, would be held and that participation in both was re
quired. Any subject who indicated she would be unable to attend the
second session was eliminated from the study.
As a pretreatment condition each subject was seen individually in a
private room and given written instructions as follows: "Here are 20
slips of paper and an envelope. You are to write on each slip of paper a
secret about yourself that you would not be willing to share with a group
of female strangers. For most people such secrets center around subjects
such as sex, violence, failures, loneliness, mental health, stealing,
cheating, and so on. However, these are only examples, and you should
not be limited to them. After you have finished, fold the slips of pa
per, place them in the envelope, and seal it. Then write your name across
After you are
finished, I will lock your envelope in this file cabinet. I have the
only key, and iio one will ever read what you have written unless you
decide on your own to show it to them. The envelope will be returned
to you after the study is over. When you have finished, please call
the flap of the envelope like this.
-14-

-15-
me." The subject areas for the secrets were taken from data on intimate
secrets obtained by Norton, Feldman & Tayofa (1974). When a subject
had finished the pretreatment phase she was randomly assigned to one
of three treatment conditions and to one of six groups of four within
that condition. An exception to the randomness was that no subject was
assigned to a group in which she was acquainted with the members.
Procedure
Condition 1. Interpersonal openness. Condition 1 was an attempt
to structure the six groups in that condition in such a way as to maxi
mize interpersonal openness. The videotape model and instructions were
adapted from those used by Whalen and described in detail in her study
as "Film-detailed instructions condition."
The videotape model consisted of a twelve-minute, untitled video
tape recording (VTR) produced for this study. Data from previous
studies (e.g., Bandura, Ross & Ross, 1963; Otis, 1975)
indicate that videotaped models are as potent as live models in elicit
ing imitative behavior. In the VTR a group of four females conversed
after implying that they had just met. Two of the participants spent
a few seconds conveying biographical information, and then a third
member, interrupting, maintained that the group members were not using
the best method of getting to know each other. For the remainder of
the time the participants talked on a more personal level, describing
their anxieties and other feelings, and feeding back their impressions
whether positive, negative, or neutral, of the other group members.
An attempt was made to make the VTR group appear as similar as possible
in composition to the subject groups, e.g., in age, sex, and dress.

-16-
Similarity between subjects and models was maximized because of the
possibility that large subject-model differences might lead the sub
jects to conclude that behavior matching was inappropriate, unnecessary,
or impossible (Goldstein, Heller & Sechrest, 1966; Hollander, 1958).
Detailed instructions to the interpersonal-openness groups were
given in written form and also by audiotape as follows: "This is a
group dynamics laboratory. It is a rare opportunity to improve your
ability to communicate with others by learning how you appear to them;
that is, how you present yourself and what impressions they form about
you. To accomplish this, it is important that you be open and honest
with other members of the group and discuss freely your immediate
feelings, attitudes, and opinions, both about yourself and others in
the group. The information used to form your impressions of the other
group members will come from their appearance, words, and behavior. We
want you to feed back your impressions to the others as soon as the
impression is formed in your mind. This will be difficult, but it is
essential to obtain the maximum benefits from the experience. The amount
you learn about yourself will depend on the extent to which you par
ticipate in the group.
"To show you what we mean by open discussion of feelings and
feedback of impressions, we are going to show you a videotape of a
group similar in composition to yours. The group members in this tape
provide excellent examples of desirable types of group behavior. You
will note that they are open in talking about themselves. In addition,
they freely express their immediate impressions of other members of the
group, and each member makes an honest effort to accept and examine the
truth or falsity of the impression as it applies to her.

-17-
"After the tape you will form your own group, and we would like you
to try and interact in a manner similar to that of the videotape group.
No experimenter will participate in your group, but an observer will be
watching through the one-way mirror in order to record the interaction
variables we are examining.
"After you have seen the tape you may begin. You will have 40
minutes, and you will be notified when your time is up. Remember, it
is important that you openly discuss your feelings and immediate im
pressions of others in the group, as well as invite similar comments
about yourself.
"Next week at this same time you are to return to this room for
your second group session."
Condition 2. Therapy group. Condition 2 was an attempt to have
the six groups in that condition vicariously experience a typical in
itial meeting of a therapy group. Participants in the videotape model
were the same as those on the videotape used in Condition 1. However,
interaction among the group members in the tape was limited to the
following categories: (1) exchange of demographic data; (2) comments
about the physical environment (building, rooms, chairs, etc.);
(3) wondering aloud about how to proceed (to "get to know each other");
(4) extragroup process (Whalen's term for impersonal discussion un
related to the group or the experiment); (5) laughter, silences, grunts,
etc.
Instructions to groups in the therapy-group condition were given
as follows: "This is a group dynamics laboratory. We are interested
in examining some variables important in the development of group
interaction. To help us to study these variables we will ask you to

-18-
interact as a group and to try to get to know each other. In order to
help you gain a clearer understanding of how a group like yours might
function, we will show you a videotape of four people in a situation
similar to the one you will find yourselves in. After the tape is over
you may begin your own interaction. No experimenter will participate
in your group, but an observer will be watching through the one-way
mirror in order to record the interaction variables we are examining.
You will have 40 minutes, and you will be notified when your time is up.
"Next week at this same time you are to return to this room for
your second group session."
Condition 3. Control group. Condition 3 was an attempt to examine
the effects of a group's simply meeting together (after viewing a
videotape not depicting group interaction) on the dependent variables
under investigation. The videotape shown was of the same length as
those shown to groups in Conditions 1 and 2 and with the same partici
pants, but depicting simply a variety of modes of interaction (teaching,
working at a task together, drinking and eating together).
Instructions to the groups in the control-group condition were the
same as those for Condition 2 groups, except that description of the VTR
was altered as follows: "This is a group dynamics laboratory. We are
interested in examining some variables important in the development of
group interaction. To help us study these variables we will ask you to
interact as a group and to try to get to know each other. In order to
help you get in the mood for your group experience, we will show you a
videotape of people interacting in various ways. After the tape is
over you may begin your own interaction. No experimenter will partici
pate in your group, but an observer will be watching through the one-way

-19-
mirror in order to record the interaction variables we are examining.
You will have 40 minutes, and you will be notified when your time is up.
"Next week at this same time you are to return to this room for
your second group session."
Each of the groups in the three conditions saw their VTR and re
ceived their instructions as a group but apart from the other seventeen
groups. Each group met in a room containing four chairs located around
a circular table with each chair partly facing a one-way mirror.
Observing and videotaping was done from the adjoining room. The sub
jects were met by the experimenter when they entered the experimental
room and were asked not to talk to the other group members until the
experiment began. When all four subjects had arrived, the experimenter
passed out the written and played the taped instructions, and then
showed the VTR appropriate to the particular treatment condition.
After the subjects had seen the tape and before the experimenter
left the room she handed out a 5 x 8 index card and a marking pen to
each subject, requesting that she write her first name on it, fold it
in half, and stand it up on the table in front of her so that the
other members could see it. The experimenter then left the room. After
40 minutes had elapsed, the experimenter re-entered the room to end
the interaction, to remind the group that they were due back for the
second session one week later, and to request that they not discuss
the experiment with their friends.
The second session. One week after the first group meetings
all groups returned for a "second session." Before the second
session began, each subject was left alone in a private room to
fill out Rotters Interpersonal Trust Scale. After she had com
pleted the Scale (reproduced in Appendix A),

-20-
she was handed an empty envelope, her original envelope containing her
20 secrets, and a typed instruction sheet which read as follows:
"Please take out of your envelope aj^ random that means, without
looking into the envelope as many, from zero to twenty, of the slips
of paper as you are willing to put in the pot for discussion in the
group tonight. The actual discussion session of your group will not be
recorded tonight. Please put the slips of paper you select into this
second envelope, write your name across the front of it, and drop it in
the bowl on the table. Hold on to your original envelope and call me."
When the four members of a group were assembled the experimenter
instructed them as follows: "Before you begin your group tonight I
would like you to discuss among yourselves the number of secrets which
each member of your group should contribute to discussion. You are not
to discuss the content of the secrets but only the number, and you must
agree as a group on one number." Discussion to consensus was video
taped. This portion of the experimental manipulation was a check on
whether the risky-shift phenomenon was operating in these groups.
After consensus was reached, the experiment was over, and the subjects
were debriefed regarding their expectations and hypotheses about the
experimental manipulations.
Each subject was seen individually at this point and given a slip
of paper on which she was asked to rate on a scale from 1 to 7 how
painful or embarrassing her secrets were (a rating of 1 representing
"not embarrassing or painful at all" and a rating of 7 representing
"extremely embarrassing or painful"). She was then given a typewritten
slip of paper with directions as follows:
"You were told that no one would ever read your secrets

-21-
unless you decided on your own to share them.' That is still
true. However, if you have no objections, I would like to
analyze the content of your secrets. This could be done by
your tearing open both your envelopes and dumping the contents
into a bag with other secrets so that 1 would never know to
whom a secret belonged. However, this is not an integral part
of the experiment and I want you to feel free to refuse if you
have any reservations."
Dependent Variables
The dependent measures consisted of (1) mean score for each group
on the Interpersonal Trust Scale; (2) behavioral data in the form of
mean number of secrets per group selected to share with the group;
(3) ratings of the total verbalizations of each group on a revised
version of Whalen's rating scale; and (4) differences between (a) mean
number of secrets chosen for sharing by group members prior to dis
cussion to consensus and (b) number agreed upon by each group.
Interpersonal openness rating scale. For purposes of this study
the six broad response categories of Whalen's scheme were used. They
are described in detail below. The descriptions following each response
category were taken as criteria upon which the judges based their
ratings.
Response categories:
1. Personal discussion. (a) Personal self-disclosure: Discussion
of feelings, attitudes, and behaviors unrelated to current group inter
action which are non-public, not ordinarily readily volunteered, and

-22-
which may make the individual vulnerable to negative evaluations from
others. Examples of actual comments categorized as personal self
disclosure by Whalen's raters are "I'm a status seeker like a lot of
people," "I might even try homosexuality," or "I just broke down and
started yelling and screaming." (b) Immediate feelings: Discussion of
feelings and attitudes related to current group interaction which are
relatively private, not ordinarily volunteered by most people, and
which may make the individual vulnerable to negative evaluations from
others (e.g., "I feel alienated here" or "Now I'm embarrassed").
(c) Personal questions: Questions relating to an individual's opinions,
feelings, or actions which are usually considered to be private in
formation: "Have you always been heavy?" or "I wondered how you felt
when we were so silent."
2. Feedback. (a) Positive feedback: The conveying of an indi
vidual's positively valenced impressions of or reactions to the comments,
appearance, or actions of another group member. This category includes
compliments, flattery, etc. Examples are "I like the way you said
that," or "You seem like an alert, intelligent person." (b) Negative
feedback: Negative evaluation of the personality, intelligence,
actions, etc. of another group member. This category includes hos
tility, ridicule, criticism and implications that the individual is not
telling the truth. Examples of comments rated as negative feedback
are "You were a fool," "Big deal!" or "Your argument isn't very ration
al." (c) Neutral feedback: This category includes comments about a
group member that cannot be evaluated as either positive or negative,
as well as direct advice from one group member to another. Some
examples are "You seem like the silent type" or "You ought to go up

-23-
and tell her." (d) Acceptance of feedback: Acceptance by a group
member of the positive, negative, or neutral feedback she receives,
for example, "Yeah, I guess you're right." (e) Reject feedback:
Refusal to accept the feedback about oneself provided by other group
members, for example, "No, you see it's really like this ..."
(f) Request feedback: A direct request that the group or particular
members provide the subject with an evaluation of herself, for example,
"What do you think of me?"
(These first two responses classes comprised the primary types
of interaction modeled in the videotape recording for the
interpersonal-openness groups.)
3. Impersonal discussion. (a) Impersonal self-disclosure:
Nonpersonal biographical information about oneself which is either
generally accessible or readily volunteered in appropriate contexts by
most people. Examples: "I come from Chicago" or "I don't know the
outskirts of the city very well." (b) Extragroup process: Nonpersonal
topics unrelated to the experiment or the group interactions "All
universities don't accept correspondence courses" or "I know a girl who
is in the Navy now." (c) Impersonal questions: Questions regarding
public, easily accessible, readily volunteered information about a
group member "Have you ever seen the San Diego campus?" or "How many
hours are you taking?"
4. Group process. Nonpersonal discussion relevant to the experi
ment such as comments regarding the experimental manipulations, the
purposes of the experiment, the behavior of the experimenter or par
ticipants in the VTR, a plan of action, etc. Examples: "Are they
watching us?" or "Let's all talk at the same time that ought to

-24-
goof them up."
5. Descriptive aspects of communicative speech: (a) Agreement:
Any verbal indication of agreement, for example, "Uh huh," and "Yeah,
that's right." Head nods are not scored. (b) Disagreement: Any
verbal indication of disagreement, for example, "NO," and "That's not
true!" Head nods or shakes are not scored. (c) Laughter: Any amount
of audible giggling or laughing by one or more group members. Smiles
and grins are not scored. (d) Silence: This category is scored only
if the duration of the silent period exceeds 3 seconds. (e) Inter
ruption : Scored whenever one group member stops another from talking
or when two or more individuals are speaking simultaneously.
6. Unscoreable utterances. A scoring category for those remarks
which are either inaudible, unintelligible, or too difficult to
categorize immediately.
Rating procedure. The middle twenty minutes of discussion for
each of the eighteen groups was rated independently by two trained
raters using an Esterline-Angus event recorder. The resultant scores
described total group interaction in terms of number of seconds spent
in discussion rated as falling into each of the response categories
(no ratings of individual members).
The raters were advanced graduate students in clinical psychology
who were trained on practice tapes to a criterion of 94% agreement on
a tape. After training, the raters also analyzed the three modeling
tapes in terms of time spent in each of the six response categories.

CHAPTER III
RESULTS
Interrater Reliability
To determine whether the raters were reliable in their judgments
of the total number of seconds spent in each category of discussion,
Spearman's rank correlation coefficient was computed for the 108 pairs
of scores. (Appendix A shows the raters' scores.) The resulting test
statistic of r^ = .986 is highly significant (j) < .001). Individual
correlation coefficients were also computed over the 18 groups for
each of the six categories. The results are shown in Table 1.
Table 1
Interrater Reliability
Response Category
Spearman's r
£
Personal discussion
.999
.01
Feedback
.955
.01
Impersonal discussion
.981
.01
Group process
.944
.01
Descriptive aspects of
communicative speech
.964
.01
Unscoreable utterances
.844
.01
Since interrater reliability was judged to be satisfactory, the
average of the two raters' scores was used in further computations.
-25-

-26-
Analysis of the Modeling Tapes
To determine whether the three 12-minute tapes used in the three
treatment conditions actually depicted the types of verbal interaction
which they were designed to model, the judges rated each of the tapes
separately. Each verbalization was rated as falling into one of the
six response categories.
The modeling tape for the interpersonal-openness condition was
designed to depict primarily the categories of personal discussion and
feedback, since those were the types of verbalizations which the inter-
personal-openness groups were supposed to model. Tapes for the therapy-
group and control-group conditions were designed to model primarily
impersonal discussion. As can be seen from Table 2, the tapes did
model the types of verbal interaction for which they were designed.
(The interpersonal-openness condition tape included some personal dis
cussion as an example of the kind of interaction which was not to be
modeled by the subjects.)
Table 2
Analysis of Modeling Tapes: Time (in seconds) and Percent of
Total Time Spent in Each of the Six Response Categories
Response
Category
Inter
personal
Openness
%
"Therapy"
Group
%
Control
Group
/
Personal Discussion
417
58
0
0
0
0
Feedback
83
12
0
0
0
0
Impersonal
Discussion
87.5
12
680.5
95
575
80
Group Process
78.5
11
0
0
0
0
Descriptive aspects
of Communicative
Speech
54.0
07
39.5
5
145
20
Unscoreable Utterances
0
0
0
0
0
0
720.0
100
720.0
100
720.0
100

-27-
Analysis of the Subjects' Interaction in the Groups
From the forty minutes of interaction had by each group, a video
tape of the middle twenty minutes was analyzed by the raters with each
verbalization judged as falling into one of the six response categories
(personal discussion, feedback, impersonal discussion, group process,
descriptive aspects of communicative speech, and unscoreable utterances).
The Kruskal-Wallis one-way analysis of variance was used to determine
whether significant differences existed among the groups in the three
treatment conditions with regard to the number of seconds spent in each
category. Results of the analysis showed that the groups differed
among themselves significantly in three of the response categories:
personal discussion (p < .01), feedback (p < .05), and impersonal dis
cussion (p < .01). Table 3 shows the results of the analysis.
For each of the response categories in which a significant dif
ference among the groups was found, the Mann-Whitney U Test was performed
to determine where the differences lay. As can be seen from Table A,
groups in the interpersonal-openness condition spent a significantly
greater amount of their time in personal discussion and a significantly
lesser amount of time in impersonal discussion than did groups in the
therapy-group and control-group conditions, which did not differ sig
nificantly from each other. In the category of feedback, the only
significant difference was between the interpersonal-openness and the
therapy-group conditions.

Table 3
Mean Number of Seconds, Percent of Total Time, and Kruskal-Wallis Analysis
of Variance for the Three Treatment Conditions
Condition Means and Percent
Interpersonal
Therapy
Control
Openness
Group
Group
Response
Category
Mean
Percent
Mean
Percent
Mean
Percent
H
Pers Disc
1112.83
45.1
8.00
.3
12.58
.5
12.51**
Feedback
Impers
107.92
4.5
0.00
.0
9.92
.4
6.12*
Disc
1002.42
42.5
2178.08
90.2
2206.42
91.5
9.58**
Group
Process
59.08
2.4
44.83
1.9
28.00
1.2
.07
Descrip
116.63
4.9
165.50
6.9
140.33
5.8
4.16
Unscore
13.75
.6
16.92
.7
13.33
.6
.26
2412.63
100.0
2413.33
100.0
2410.58
100.0
*£ < .05
**£ < .01

-29-
Table 4
Differences between Pairs of Treatment Conditions on the
Three Significantly
Different Response Categories
Response
Category
Treatments (in Increasing
Order of the Means of the Ranks)
Personal
Discussion
T-G C-G
1-0
Feedback
T-G C-G
1-0
Impersonal Discussion 1-0 T-G C-G
Note: Treatments connected by a line are not significantly different
at the .05 level.
Comparison of Groups and Their Models
Figures 1 through 3 represent comparisons of the percent of time
spent by group members and by their models in the six categories of
discussion. Groups in the therapy-group and control-group conditions
most closely matched their models. In the interpersonal-openness con
dition the groups matched their models except in the category of
impersonal discussion, where the group members spent 42.5% of their
time, compared to the models' 12%.
Dependent Measures
To determine whether the groups in the three treatment conditions
differed among themselves with regard to the five separate response
variables, the Kruskal-Wallis one-way analysis of variance was computed
for each of the dependent variables. Reference to Table 5 shows that
the groups did differ significantly on three of the dependent variables:

Percent
-30-
100
Models
90
80
Group
Members
70
60
50
40
30
20
10
0
Pers Feedback
Disc
Impers
Disc
Group Descrip Unscore
Process
Figure 1. Comparison of models and groups in the interpersonal
openness condition on percent of time spent in each of
the six categories of response.

Percent
-31-
Disc Disc
. B
Group Descrip
Process

Models
Group
Members
Unscore
Figure 2. Comparison of models and groups in the therapy-group
condition on percent of time spent in each of the six
categories of response.

-32-
100
| Models
90
80
70
60
c
QJ
U
u
0)
P-.
50
40
30
20
10
0
Pers Feedback Impers
Disc Disc
|jjfl Group
Members
Group Descrip Unscore
Process
Figure 3. Comparison of models and groups in the control-group
condition on percent of time spent in each of the six
categories of response.

-33-
mean number of secrets which they were willing to share (p < .05);
mean number of group members willing to allow the experimenter to
analyze the content of their secrets (p < .01); and mean number of
secrets which the group as a group agreed to reveal after discussion
to consensus (p < .05). The groups did not differ on their individual
ratings of how painful or embarrassing they considered their secrets
to be or on their mean scores on Rotter's Interpersonal Trust Scale.
Table 5
Differences Among the Treatment Conditions on
the Five Dependent Variables
Dependent
Variable
Interpersonal
Openness
Mean Rank
Mean
Therapy
Group
Mean Rank
Mean
Control
Group
Mean Rank
Mean
H
Secrets shared
14.33
7.16
7.00
2.96
7.17
2.78
7.38*
Analysis of
secrets
permitted
15.00
4.00
7.30
2.16
6.17
2.0
10.50**
Painfulness of
secrets
(1-7)
10.50
4.80
10.25
4.68
7. 75
4.53
.98
Trust-Scale
Scores
6.22
66.75
12.25
69.53
6.92
64.66
3.02
Secrets chosen
by consensus
14.00
2.70
8.50
1.00
6.00
.37
7.67*
*p < .05
**p < .01

-34-
For each of the dependent variables on which the groups differed
significantly, the Mann-Whitney U-Test was performed for each pair of
treatment conditions. As Table 6 shows, groups in the interpersonal
openness condition were willing to share a significantly greater number
of their secrets with other members of their group, and more of the
group members were willing to permit the experimenter to analyze their
secrets, than were groups in the other treatment conditions, who did
not differ significantly from each other. As to the mean number of
secrets which the groups agreed to share after discussion to consensus,
the significant difference (p < .05) occurred between groups in the
interpersonal-openness condition and those in the control-group
condition.
Table 6
Differences Between Pairs of Treatment Conditions on
the Three Significantly Different Dependent Variables
Treatments (in decreasing order
Dependent Variable
of the
means of the
ranks)
Secrets shared
1-0
C-G
T-G
Analysis of secrets
permitted
1-0
T-G
C-G
Secrets chosen by
consensus
1-0
T-G
C-G
Note: Treatments connected by a line are not significantly different
at the .05 level.

-35-
Table 7 is a further analysis of the "risky-shift" data. By
reference to that table it can be seen that the groups in the three
treatment conditions differed significantly in the number of secrets
which they were willing to share both before and after discussion to
consensus (p < .05). However, after discussion to consensus the only
significant difference was between the interpersonal-openness and
control-group conditions. The differences between the number of secrets
subjects were willing to share before and after discussion were signi
ficant in each of the treatment conditions and were all in the direction
of a "conservative" rather than a "risky" direction (Sign Test).
Table 7
Comparison of Mean Number of Secrets Subjects Were
Willing to Reveal Before and After Discussion to Consensus
Interpersonal Therapy Control
Openness
Group
Group
£
Before
Discussion
7.16
2.96
2.78
.05
After
Discussion
2. 70
1.00
.37
.05
Shift
5.46
1.96
2.41
£
.02
.03
.02
Note: Treatments connected by a line are not significantly different
at the .05 level.
To discover whether the groups differed among themselves with
reference to the amount of time they spent in their discussion to con
sensus, the Kruskal-Wallis test was performed. The mean amount of time

-36-
in minutes for groups in each treatment condition was as follows:
interpersonal-openness, 6.46; therapy-group, 3.28; control-group, 10.34.
The groups did not differ significantly at the .05 level.

CHAPTER IV
DISCUSSION
The results of the study are discussed in the order in which the
data were obtained and analyzed. The effects of modeling (Hypothesis
3) are discussed first, followed by their influence on group interaction
and the outcome variables (Hypotheses 2 and 1, respectively).
Hypothesis 3: The Differential Effects of Modeling
As Whalen and others have shown, complex verbal response patterns
can be taught to groups by the use of descriptive and exhortative in
structions and the use of videotape models of interpersonally open
verbal interaction. The data from this experiment support this finding
and additionally indicate the need for specific instructions and un
ambiguous models to facilitate a particular verbal behavior.
Apparently, personal discussion as defined herein is a more dif
ficult behavior to inculcate in group members than is impersonal dis
cussion. The data show that, even with the use of models whose
interaction included very minimal impersonal discussion, the subjects
produced more than three times as much impersonal discussion as the
models exhibited. In the groups viewing models of primarily impersonal
discussion the interaction matched the models much more closely. How
ever, the only condition in which personal discussion predominated was
that in which the models exhibited it as the primary means of interaction.
-37-

-38-
Thus, while interpersonally open behavior is difficult to induce, the
interpersonal-openness models and instructions used in this study were
effective in producing a significantly greater amount of that behavior
as compared to the other two conditions, the therapy-group and the
control-group.
An additional essential element in effective modeling other than
the clarity of the behavior to be matched and the explicit instructions
appears to be motivation. Only groups in the interpersonal-openness
condition received as part of their instructions the statement that
"[This] is a rare opportunity to improve your ability to communicate
with others by learning how you appear to them; that is, how you present
yourself and what impressions they form about you." A motivational
factor was also present in the modeling tape in that a member who was
open was reinforced by the other group members with positive feedback.
It seems probable that the treatment effects in the present study were
exerted through Bandura's principle of disinhibition of a previously
learned response due to the positive consequences experienced by the
models. Lakin and Carson (1966) have observed that people most often
avoid engaging in personal discussion because of the fear of being per
ceived as weak and inadequate. To counteract such a fear and disinhibit
the modeled response, the subject must have been motivated by an
expectation of positive consequences.
Thus, Hypothesis 3 is supported with the qualification that suf
ficient motivation in the subjects to model the behavior must also be
induced.

-39-
Hypothesis 2: Willingness to Disclose Painful Secrets
Hypothesis 2 is supported by the finding that members of groups in
the interpersonal-openness treatment condition were willing to disclose
a significantly greater number of secrets to their fellow members than
were subjects in the other two treatment conditions. The fact that the
subjects did not differ across conditions in terms of how painful or
embarrassing they considered their secrets to be indicates that the same
degree of risk obtained for all the group members.
Hypothesis 2 predicted that the readiness to risk disclosing secrets
would be related to the presence or absence of a norm of interpersonal
openness. Analyzing the treatment condition in terms of the norm-sending
process detailed by Thibaut and Kelley (1959), one finds that the three
components of norm-sending were clearly present only in the interpersonal
openness condition. The components and their counterparts in the
present study are (1) rule statements (instructions and videotape);
(2) surveillance-evaluation (presence of other group members and the
experimenter's observation from the next room); and (3) application of
sanctions (applied by the subject and the other group members in the
form of rewards for interpersonal openness; the expected reward promised
in the instructions of "improving your ability to communicate with
others . ."). Additionally, the finding that groups in the inter
personal-openness condition actually spent approximately half of their
time in personal discussion and feedback indicates that a norm of
interpersonal openness was operating.
There is also some support for the idea that a norm of interper
sonal openness generalized outside the group. Such support is found in
the fact that a significantly greater number of members of groups in

-40-
the interpersonal-openness condition were willing to permit the ex
perimenter to analyze the content of their secrets. The readiness to
allow a stranger to read what one had previously identified as secrets
one would not be willing to share, even when promised anonymity,
indicates at least a state variable of being open about one's self.
It is also worth emphasizing that there was no pressure on any of the
subjects to permit analysis of the content of their secrets; indeed,
one-third of the subjects in the therapy-group and in the control-group
conditions declined to allow such analysis.
Hypothesis 1: Interpersonal Trust
Members of groups in the three treatment conditions did not differ
significantly on the variable of interpersonal trust as measured by
Rotter's Scale. The means for all of the groups were well within one
standard deviation of the mean for the standardization group of females
and also for a group of 411 females at Ohio University recently studied
by Wright & Tedeschi (1975), indicating that the population from which
our samples were drawn does not diverge markedly from the norms for
this scale.
Rotter speaks of interpersonal trust as measured by his scale as
a "generalized expectancy that the oral or written statements of other
people can be relied upon." In thus describing this variable he dif
ferentiates it from expectancies specific to the situation in which an
individual finds himself. In other research such a differentiation is
sometimes described in terms of "trait" versus "state" variables. Thus
Rotter would predict that scores on his scale would remain relatively
stable over time, since generalized expectancies are based on long-term

-41-
life experiences with parents, teachers, and peers.
As measured by Rotter's scale, then, interpersonal trust is an
attitude that would not be expected to fluctucate over such a short
period of time as was involved in the present study unless some very
dramatic change occurred in the lives of the group members. Apparently,
the experimental manipulation, while it was effective in altering the
expectancies of the subjects relative to the specific situation of the
experiment (witness the significant differences on other dependent
variables) was not potent enough to effect a change in what might be
thought of as a personality or trait variable of trust. Such an effect
as personality change might be obtained over a longer period of time
with repeated experiences in groups such as those in the present study.
One would expect that a successful therapy-group experience might pro
duce such a shift in attitude.
Alternatively, if interpersonal trust is less narrowly defined,
the data on the subjects' willingness to share secrets and to allow them
to be analyzed by the experimenter could be taken as an indication that
trust in other persons was affected differentially by the treatment
conditions. Certainly, some degree of trust is indicated by a readiness
to disclose painful aspects of one's self to others. However, such a
conclusion is limited in its applicability until further research
demonstrates that the effects endure over time and across other groups
or individuals. Hypothesis 1 is not clearly supported by the data
obtained in the present study.
The Risky Shift
Groups in all three treatment conditions shifted significantly in

-42-
a conservative direction following discussion to consensus on number of
secrets to be shared. However, after discussion the difference between
groups in the interpersonal-openness condition and those in the therapy-
group condition disappeared. Groups in the control-group condition
were not significantly different from those in the therapy-group
condition.
The reasons for such a uniformly conservative shift are difficult
to discover. Each group member decided on her own to reveal a certain
number of secrets. Yet, apparently she was not willing to impose on
other group members a greater degree of risk than she had accepted for
herself. One could also theorize that the less trusting members of the
group prevailed; an element of distrust in the person of one reluctant
member might have dissipated the atmosphere of trust built up in the
group. In sum, being unsure of the reasons for a member or members'
reluctance to disclose her self to any degree, the other members may
have concluded (a) that they should not try to persuade her (or them)
otherwise; or (b) that perhaps that person or persons was not really
so trustworthy as they had thought. Certainly, in therapy groups a
client is often heard to say, after the session is over, that he or she
had planned to talk about such-and-such a topic, but had lost the courage
to do so when other members seemed reluctant to expose themselves.
Unlike studies using the Choice Dilemma Questionnaire, the present
study involved a real and personal risk for the group members. It would
seem that the gain to be expected from taking such a risk was, in this
case, outweighed by other factors which became operative only in the
context of the discussion to consensus. What those factors were is a
matter of speculation. However, further research involving analysis of

-43-
the videotaped discussions might possibly reveal the answers. Such an
analysis is beyond the scope of the present investigation.
Conclusions
The results of the present study suggest some important conclusions
for the theory and practice of group psychotherapy. If the data herein
can be viewed tentatively as applicable to a clinic population, we
might conclude:
1. That members of a group can be taught to behave in thera
peutically relevant ways (e.g., relating in an interpersonally open
manner);
2. That such training can be carried out in a simple and practical
manner through the use of brief instructions and videotape models;
3. That, when members of a group engage in such interpersonally
open behavior, a climate of trust is built up wherein the group members
become willing to share with each other painful and embarrassing aspects
of their lives;
4. That, while the actual disclosure of such secrets is not
necessarily a sine qua non of therapy, the readiness to do so may be;
and
5. Finally, that trust in one's fellow group members does gener
alize, at least to the extent in the present study of trusting the
experimenter.
Necessarily, generalization of the results of this study to a
clinic population must wait upon replication with actual therapy groups.
However, such research seems clearly justified on the basis of the data
obtained herein.

-44-
Whalen's rating scale as modified in the present study has shown
itself to be highly reliable; the raters reported no great difficulty
in distinguishing among the categories. Hence it would seem to be an
instrument of choice for further research in this area.
No conclusions can be drawn regarding the reasons for the "con
servative shift" obtained from our subjects. The present research
fortunately included videotaping of the actual discussion-to-consensus
sessions. Analysis of the group interaction in such sessions would
seem to be a natural outgrowth of this study and perhaps have a bearing
on the interactional aspects of the risky shift phenomenon.

APPENDIX A
Interpersonal Trust Scale (Rotter, 1967)
General Opinion Survey
This is a questionnaire to determine the attitudes and beliefs of
different people on a variety of statements. Please answer the state
ments by giving as true a picture of your own beliefs as possible. Be
sure to read each item carefully and show your beliefs by marking the
appropriate number of your answer sheet.
If you strongly agree with an item, circle number 1. Circle number
2 if you mildly agree with the item. That is, circle number 2 if you
think the item is generally more true than untrue according to your
beliefs. Circle number 3 if you feel the item is about equally true as
untrue. Circle number 4 if you mildly disagree with the item. That is,
circle number 4 if you feel the item is more untrue than true. If you
strongly disagree with an item, circle number 5.
1. Strongly agree
2. Mildly agree
3. Agree and disagree equally
4. Mildly disagree
5. Strongly disagree
1. Most people would rather live in a climate that is mild all year
around than in one in which the winters are cold.
2. Hypocrisy is on the increase in our society.
-45-

-46-
3. In dealing with strangers one is better off to be cautious until
they have provided evidence that they are trustworthy.
4. This country has a dark future unless we can attract better people
into politics.
5. Fear of social disgrace or punishment rather than conscience pre
vents most people from breaking the law.
6. Parents usually can be relied upon to keep their promises.
7. The advice of elders is often poor because the older person doesn't
recognize how times have changed.
8. Using the honor system of not having a teacher present during
exams would probably result in increased cheating.
9. The United Nations will never be an effective force in keeping
world peace.
10. Parents and teachers are likely to say what they believe themselves
and not just what they think is good for the child to hear.
11. Most people can be counted on to do what they say they will do.
12. As evidenced by recent books and movies morality seems on the
downgrade in this country.
13. The judiciary is a place where we can all get unbiased treatment.
14. It is safe to believe that in spite of what people say, most people
are primarily interested in their own welfare.
15. The future seems very promising.
16. Most people would be horrified if they knew how much news the
public hears and sees is distorted.
17. Seeking advice from several people is more likely to confuse than
it is to help one.
18. Most elected public officials are really sincere in their campaign
promises.
19. There is no simple way of deciding who is telling the truth.
20. This country has progressed to the point where we can reduce the
amount of competitiveness encouraged by schools and parents.
21. Even though we have reports in newspapers, radio and television,
it is hard to get objective accounts of public events.
22. It is more important that people achieve happiness than that they
achieve greatness.

-4 7-
23. Most experts can be relied upon to tell the truth about the limits
of their knowledge.
24. Most parents can be relied upon to carry out their threats of
punishment.
25. One should not attack the political beliefs of other people.
26. In these competitive times one has to be alert or someone is likely
to take advantage of you.
27. Children need to be given more guidance by teachers and parents
than they now typically get.
28. Most rumors usually have a strong element of truth.
29. Many major national sport contests are fixed in one way or another.
30. A good leader molds the opinions of the group he is leading rather
than merely following the wishes of the majority.
31. Most idealists are sincere and usually practice what they preach.
32. Most salesmen are honest in describing their products.
33. Education in this country is not really preparing young men and
women to deal with the problems of the future.
34. Most students in school would not cheat even if they were sure of
getting away with it.
35. The hordes of students now going to college are going to find it
more difficult to find good jobs when they graduate than did the
college graduates of the past.
36. Most repairmen will not overcharge even if they think you are
ignorant of their specialty.
37. A large share of accident claims filed against insurance companies
are phony.
38. One should not attack the religious beliefs of other people.
39. Most people answer public opinion polls honestly.
40. If we really knew what was going on in international politics, the
public would have more reason to be frightened than they now seem
to be.

1
4
7
11
14
17
oup
1
4
7
11
14
17
APPENDIX B
Time in .5-sec. Intervals in the Six Response Categories by Rater for
the Six Groups in Each Treatment Condition
1. Personal Discussion
Interpersonal-Openness
Therapy-
Group
Control
Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
948
997
3
29
30
2
19
21
708
934
6
0
0
5
0
0
1071
1073
9
0
0
8
0
0
107
110
12
0
0
10
0
0
1395
1470
15
19
18
13
0
0
1591
1630
18
0
0
16
0
0
5820
6214
48
48
19
21
2.
Feedback
Interpersonal-Openness
Therapy-Group
Control-
-Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
0
0
3
0
0
2
48
52
30
36
6
0
0
5
6
5
493
501
9
0
0
8
0
0
0
0
12
0
0
10
4
0
70
80
15
0
0
13
0
0
46
39
18
0
0
16
0
0
639
656
0
0
58
57
i
-p-
oo
i

1
4
7
11
14
17
OUP
1
4
7
11
14
17
3. Impersonal Discussion
nterpersonal-Openness
Therapy
-Group
Control1
-Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
1483
1428
3
2260
2223
2
2088
2072
1419
1218
6
2291
2316
5
2244
2274
419
446
9
2065
2150
8
2191
2159
2168
2169
12
2220
2257
10
2316
2336
727
698
15
1958
1979
13
2194
2177
360
334
18
2221
2197
16
2216
2211
6576
6293
13,015
13,122
13,249
13,229
4. Group Process
Interpersonal-Openness
Therapy-
Group
Control-
-Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
0
0
3
0
0
2
35
30
54
10
6
0
10
5
25
26
192
160
9
50
44
8
0
0
0
0
12
0
0
10
7
6
148
145
15
216
208
13
0
0
0
0
18
0
0
16
71
71
394
315
266
262
131
127
-49-

1
4
7
11
14
17
oup
1
4
7
11
14
17
5. Descriptive Aspects of
Communicative Speech
Interpersonal-Openness
Therapy-Group
Control-
Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
87
70
3
157
154
2
97
93
113
116
6
116
119
5
145
130
135
131
9
266
233
8
175
174
115
109
12
198
183
10
111
123
91
89
15
151
166
13
175
175
185
158
18
117
126
16
187
194
726
673
1005
981
890
889
i
Ln
O
I
6. Unscoreable Utterances
Interpersonal-Openness
Therapy-
Group
Control-
Group
Rater
Rater
Rater
R
M
Group
R
M
Group
R
M
0
0
3
0
0
2
25
35
46
35
6
9
0
5
0
0
41
16
9
0
0
8
0
0
12
15
12
0
0
10
0
20
0
0
15
26
26
13
0
0
_0
_0
18
63
20
16
_0
_0
99
66
98
96
25
55

REFERENCES
Alexander, F. Psychoanalytic therapy: Principles and applications.
New York: Ronald Press, 1946.
Allen, J.G. Implications of research in self-disclosure for group
psychotherapy. International Journal of Group Psychotherapy,
1973, 23, 306-321.
Bandura, A. Principles of behavior modification. New York: Holt,
Reinhart, 1969.
Bandura, A., Ross, D., & Ross, S. Imitation of film-mediated aggressive
models. Journal of Abnormal and Social Psychology, 1963, 66,
3-11.
Bednar, R.L., Weet, C., Evensen, P., Lanier, D., & Melnick, J. Empirical
guidelines for group therapy: Pretraining, cohesion and modeling.
Journal of Applied Behavioral Science, 1974, 10_ (2), 149-165.
Blascovich, J., Ginsberg, G.P., & Veach, T.L. A pluralistic explanation
of choice shifts on the risk dimension. Journal of Personality
and Social Psychology, 1975, JU, 422-429.
Blascovich, J., Veach, T.L., & Ginsberg, G.P. Blackjack and the risky
shift. Sociometry, 1973, _36, 42-45.
Bradford, L.P. Membership and the learning process. In L.P. Bradford,
J.R. Gibb & K.D. Berne (Eds.), T-group therapy and laboratory
method. New York: Wiley, 1964.
Broadbent, D. Attention and the perception of speech. In Perception:
Mechanisms and models. San Francisco: W.H. Freeman, 1972.
Cozby, P.C. Self-disclosure: A literature review. Psychological
Bulletin, 1973, 79, 73-91.
Deutsch, M. Cooperation and trust: Some theoretical notes. Nebraska
symposium on motivation, 1962.
Dibner, A.S. Ambiguity and anxiety. Journal of Abnormal and Social
Psychology, 1958, 56, 165-174.
Dion, K., Baron, R., & Miller, N. Why do groups make riskier decisions
than individuals? In L. Berkowitz (Ed.), Advances in Experimental
Psychology, vol. 5. New York: Academic Press, 1970.
-51-

-52-
Doster, J.A. Individual differences affecting interviewee expectancies
and perceptions of self-disclosure. Journal of Counseling
Psychology, 1975, 2/2, 192-198.
Freed, A., Chandler, P.J., Blake, R.R., & Mouton, J. Stimulus and back
ground factors in sign violation. Journal of Personality, 1955,
23, 499.
Freedman, J.L. Role playing: Psychology by consensus. Journal of
Personality and Social Psychology, 1969, 13, 107-114.
Gahagan, J.P., & Tedeschi, J.T. Strategy and the credibility of promises
in the Prisoner's Dilemma game. Journal of Conflict Resolution,
1968, 12, 224-234.
Gendlin, E.T. A theory of personality change. In J.T. Hart & T.M.
Tomlinson (Eds.), New directions in client-centered therapy.
Boston: Houghton Mifflin, 1970.
Goldstein, A.P., Heller, K., & Sechrest, L.B. Psychotherapy and the
psychology of behavior change. New York: Wiley, 1966.
Grosser, D., Polansky, N., & Lippit, R. A laboratory study of behavioral
contagion. Human Relations, 1951, 4^, 115-142.
Hollander, E.P. Conformity, status and idiosyncrasy credit.
Psychological Review, 1958, 6J3, 117-127.
Jellison, J.M., & Riskind, J. A social comparison of abilities inter
pretation of risk-taking behavior. Journal of Personality and
Social Psychology, 1970, 1_5, 375-390.
Jellison, J.M., & Riskind, J. Attribution of risk to others as a
function of their ability. Journal of Personality and Social
Psychology, 1971, 20, 413-415.
Johnson, J.A. Group therapy: A practical approach. New York: McGraw-
Hill, 1963.
Jourard, S.M. The transparent self. New York: Van Nostrand, 1971.
Kahan, J.P. A subjective probability interpretation of the risky shift.
Journal of Personality and Social Psychology, 1975, 31 (6),
977-982.
Kimbrell, D., & Blake, R.R. Motivational factors in the violation of
a prohibition. Journal of Abnormal and Social Psychology, 1958,
56, 132-133.

-53-
Kogan, N., & Wallach, M.A. Risk taking: A study in cognition and
personality. New York: Harper, 1964.
Korchin, S.J., Basowitz, H., Grinker, R., Hamburg, D., Persky, H.,
Sabshin, M., Heath, H., & Board, F. Experience of perceptual
distortion as a source of anxiety. American Medical Association
Archives of Neurology and Psychiatry, 1958, ^0, 98-113.
Lakin, M., & Carson, R.C. A therapeutic vehicle in search of a theory
of therapy. Journal of Applied Behavioral Science, 1966, 2^,
27-40.
Lazarus, R. Psychological stress and the coping process. New York:
McGraw-Hill, 1966.
McCauley, C., Stitt, C.L., Woods, K., & Lipton, D. Group shift to
caution at the race track. Journal of Experimental Social
Psychology, 1973, _9, 80-86.
Mowrer, O.H. The new group therapy. Princeton: Van Nostrand, 1964.
Norton, R., Feldman, C., & Tayofa, D. Risk parameters across type of
secrets. Journal of Counseling Psychology, 1974, _21, 450-454.
Otis, M.R. The effects of interpersonal skills training with videotaped
and live models on verbal and nonverbal components of behavior.
Unpublished master's thesis. University of Florida, 1975.
Postman, L., & Bruner, J. Perception under stress. Psychological
Review, 1948, 55, 314-323.
Pruitt, D.G. Conclusions: Toward an understanding of choice shifts in
group discussion. Journal of Personality and Social Psychology,
1971, 20, 404-510.
Rabin, H.M. Preparing patients for group psychotherapy. International
Journal of Group Psychotherapy, 1970, jM), 135-145.
Ribner, N.G. Effects of an explicit group contract on self-disclosure
and group cohesiveness. Journal of Counseling Psychology, 1974,
21 (2), 116-120.
Rogers, C.R. Client-centered therapy. In Silvano Arieti (Ed.),
Ameridan handbook of psychiatry, vol. 3. New York: Basic Books,
1959.
Rotter, J.B. A new scale for the measurement of interpersonal trust.
Journal of Personality, 1967, 35, 651-665.
Rotter, J.B. Generalized expectancies for interpersonal trust.
American Psychologist, 1971, 2jj, 443-452.

-54-
Schlenker, B.R., Helm, B., & Tedeschi, J.T. The effects of personality
and situational variables on behavioral trust. Journal of Per
sonality and Social Psychology, 1973, 15 (3), 419-427.
Schofield, J.W. Effect of norms, public disclosure, and need for
approval on volunteering behavior consistent with attitudes.
Journal of Personality and Social Psychology, 1975, 31 (6),
1126-1133.
Schtz, W.C. Encounter. In Raymond Corsini (Ed.), Current
psychotherapies. Itasca, Ill: F.E. Peacock, Inc., 1973.
Singer, E. Key concepts in psychotherapy. New York: Basic Books,
19 70.
Thibaut, J.W., & Kelley, H.H. The social psychology of groups.
New York: Wiley, 1959.
Truax, C.B., & Wargo, D.B. Effects of therapeutic conditions, alternate
sessions, vicarious therapy pretraining, and patient self-explora
tion on hospitalized mental patients during group therapy.
Journal of Consulting and Clinical Psychology, 1969, 33, 440-447.
Verville, E. The effect of emotional and motivational sets on the
perception of incomplete pictures. Journal of General Psychology,
1946, 69, 133-145.
Vinokur, A. Cognitive and affective processes influencing risk taking
in groups: An expected utility approach. Journal of Personality
and Social Psychology, 1971, 2(), 472-486.
Walters, R.H., & Parke, R.D. Influence of response consequences to a
social model on resistance to deviation. Journal of Experimental
Child Psychology, 1964, 1, 269-280.
Whalen, C. Effects of a model and instructions on group verbal be
haviors. Journal of Consulting and Clinical Psychology, 1969, 33,
509-521.
Wolpe, J. The practice of behavior therapy. New York: Pergamon
Press, 1969.
Worthy, G., Gary, A.L., & Kahn, G. Self-disclosure as an exchange
process. Journal of Personality and Social Psychology, 1969,
13 (1), 59-63.
Wright, T., & Tedeschi, R. Factor analysis of the Interpersonal Trust
Scale. Journal of Consulting and Clinical Psychology, 1975, 43,
470-477.
Yalom, I.D. A study of group dropouts. Archives of General Psychiatry,
1966, 14, 393-414.

-55-
Yalom, I.D. The theory and practice of group psychotherapy. New York
Basic Books, 1970.
Yinon, Y. Risky aggression in individuals and groups. Journal of
Personality and Social Psychology, 1975, 31, 808-815.

BIOGRAPHICAL SKETCH
Sallye B. Peterson was born in Shreveport, Louisiana, on October 31,
1934, and grew up there, graduating from Byrd High School in 1951. She
received her B.A. magna cum laude from Louisiana State University in
1965 and her M.A. in Clinical Psychology in 1967.
Between 1967 and 1973 she taught at colleges in Rhode Island and
Tennessee and worked as a staff psychologist at the Helen R. McNabb
Community Mental Health Center in Knoxville, Tennessee. She is presently
employed as Coordinator of School Consultation at the Child, Youth, and
Family Center in Gainesville, Florida, where she has been working since
April, 1975. She is the mother of three children: Michael Andrew,
aged seven; Moetke Laren, aged six; and Megan Kathleen, aged two.
-56-

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Professor of Clinical Psychology
1 certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
James C. Dixon
Professor of Psychology
* I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
y
/( /7 c><-4y
Nathan W. Perry
Professor of Clinical Psychology
Acting Chairman of the Department of
Clinical Psychology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Barry R. Schlenker
Associate Professor of Psychology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
'Y } 7-
Marvin E. Shaw
Professor of Psychology
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
This dissertation was submitted to the Graduate Faculty of the Department
of Clinical Psychology in the College of Arts and Sciences and to the
Graduate Council, and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
March 1977
Dean, Graduate School



-16-
Similarity between subjects and models was maximized because of the
possibility that large subject-model differences might lead the sub
jects to conclude that behavior matching was inappropriate, unnecessary,
or impossible (Goldstein, Heller & Sechrest, 1966; Hollander, 1958).
Detailed instructions to the interpersonal-openness groups were
given in written form and also by audiotape as follows: "This is a
group dynamics laboratory. It is a rare opportunity to improve your
ability to communicate with others by learning how you appear to them;
that is, how you present yourself and what impressions they form about
you. To accomplish this, it is important that you be open and honest
with other members of the group and discuss freely your immediate
feelings, attitudes, and opinions, both about yourself and others in
the group. The information used to form your impressions of the other
group members will come from their appearance, words, and behavior. We
want you to feed back your impressions to the others as soon as the
impression is formed in your mind. This will be difficult, but it is
essential to obtain the maximum benefits from the experience. The amount
you learn about yourself will depend on the extent to which you par
ticipate in the group.
"To show you what we mean by open discussion of feelings and
feedback of impressions, we are going to show you a videotape of a
group similar in composition to yours. The group members in this tape
provide excellent examples of desirable types of group behavior. You
will note that they are open in talking about themselves. In addition,
they freely express their immediate impressions of other members of the
group, and each member makes an honest effort to accept and examine the
truth or falsity of the impression as it applies to her.


CHAPTER IV
DISCUSSION
The results of the study are discussed in the order in which the
data were obtained and analyzed. The effects of modeling (Hypothesis
3) are discussed first, followed by their influence on group interaction
and the outcome variables (Hypotheses 2 and 1, respectively).
Hypothesis 3: The Differential Effects of Modeling
As Whalen and others have shown, complex verbal response patterns
can be taught to groups by the use of descriptive and exhortative in
structions and the use of videotape models of interpersonally open
verbal interaction. The data from this experiment support this finding
and additionally indicate the need for specific instructions and un
ambiguous models to facilitate a particular verbal behavior.
Apparently, personal discussion as defined herein is a more dif
ficult behavior to inculcate in group members than is impersonal dis
cussion. The data show that, even with the use of models whose
interaction included very minimal impersonal discussion, the subjects
produced more than three times as much impersonal discussion as the
models exhibited. In the groups viewing models of primarily impersonal
discussion the interaction matched the models much more closely. How
ever, the only condition in which personal discussion predominated was
that in which the models exhibited it as the primary means of interaction.
-37-