a comparison of techniques for building cohesiveness in small ...

A COMPARISON OF TECHNIQUES FOR BUILDING

COHESIVENESS IN SMALL GROUPS

by

JEANNE CLARISSA MILLER, B.A.

A THESIS

IN

PSYCHOLOGY

Submitted to the Graduate Faculty of Texas Tech University in

Partial Fulfillment of the Requirements for

the Degree of

MASTER OF ARTS

Approved

Accepted

May, 1975

.V ^ ^ TABLE OF CONTENTS

LIST OF TABLtS

I. INTRODUCTION .

Group Cohesiveness: Definition Effects of Group Cohesiveness Techniques for Facilitating Group Cohesiveness Research Hypotheses

II. METHOD

Experimental Design Statistical Analysis Subjects Instruments Group Leaders Procedure

III. RESULTS

Hypothesis la Hypothesis lb Hypothesis 2a Hypothesis 2b

IV. DISCUSSION AND CONCLUSIONS

V. SUMMARY . . .

REFERENCES

APPENDIX

Group Cohesiveness (Questionnaire Sociometr ic (Questions C r i t i c a l Inc ident (Question Se l f -Eva lua t ion Quest ionnaire

111

1 3 6 9

10

10 10 10 11 12 12

14

14 14 16 16

23

30

31

33

33 34 35 36

11

TEXAS TECH LIBRARY

LIST OF TABLES

Table Page 1. Cohesiveness Questionnaire Scores: Summary of Group

Means and Standard Deviations - . 15

2. Cohesiveness Questionnaire: Analysis of Variance

Summary Table 15

3. Responses to. Critical Incident Question 17

4. Anxiety Questionnaire Scores: Summary of Group Means and Standard Deviations 22

5. Anxiety Questionnaire Scores: Analysis of Variance Summary Table 22

11

I

CHAPTER I

INTRODUCTION

The purpose of this investigation was to evaluate the effectiveness

of cohesiveness-buiIding techniques used in encounter and therapy groups.

Group Cohesiveness: Definition

Research into small group processes has been conducted with

increasing frequency since the 1920's. Among the early investigations

was South's study of group productivity versus individual efforts

(Kelley & Thibaut, 1954). In the 1930's, this trend was continued by

Watson, Thorndike, and Timmons, among others (Kelley & Thibaut, 1954).

These investigations were usually confined to the variables of group j

size, type of task, and other objectively observed qualities. In the

1940's and 1950's the research emphasis shifted toward emotion and the

interaction of individuals within groups. Group process was investigated

and subjective data derived from interviews and questionnaires was

used. This emphasis became more pronounced following Lippitt and

White's (1952) studies of leadership styles. Research since then has

focused on group characteristics such as cohesiveness, status relation­

ships, and uniformity pressures. The growth of the National Training

Laboratory for Group Development was followed by the boom in the 1960's

of encounter groups of all kinds and the increased popularity of group

psychotherapy.

Of all group characteristics, cohesiveness is one of the most

frequently studied, and yet the large body of research has not yielded

a single, universally-accepted definition of cohesiveness. Lott (1965)

defines cohesiveness as a property of a group that is inferred from the

number and strength of mutual positive attitudes held by members of a

group. This definition rests on the assumption that interpersonal

attraction and attitude similarity are central to cohesiveness and that

other factors may or may not be considered relevant. Although this

definition is easily operationalized, its dependence upon the theory of

attitude similarity as the main source of attraction makes it unaccept­

able to those unwilling to accept the basic premise.

The definition espoused by Kaplan and Sadock (1970 is almost

mystical in its vagueness. They define group cohesion as "effect of

the mutual bonds between members of a group as a result of their concert­

ed effort for a common interest and purpose. Until cohesiveness is

achieved, the group cannot concentrate its full energy on a common task"

(1971, p. 842). This definition would not easily be operationalized,

to say the least. It can be inferred, however, that Kaplan and Sadock

consider group cohesiveness as a necessary condition for success of

a group.

The most widely used definition of cohesiveness is that "it

represents the resultant of all forces acting on members to remain in

the group" (Cartwright & Zander; I968, p. 90- The forces include the

attractiveness of the group and the attractiveness of alternative

memberships. This definition is easily applied to all types of small

groups because it does not attempt to define the sources of the attract­

iveness of the group or of the alternative memberships. Although its

direct application as an operational definition would be difficult and

certainly not practical (Collins £• Raven, I969), it does have implications

3

for measurement as will be discussed later. The "resultant forces"

definition of cohesiveness is the one used in this study.

Effects of Group Cohesiveness

Although much of the research on cohesiveness has not used

encounter and therapy groups, the results are relevant to groups of all

kinds. Those findings as appear most directly relevant to encounter and

therapy groups will be reported here.

Several studies have shown that, as cohesiveness increases, there is

more frequent communication among members and a greater degree of partici­

pation in group activities. In studying natural student groups, Lott

and Lott (1965) obtained a significant positive correlation between

cohesiveness and median frequency of communication among group members

on a discussion topic. Cervin (1956) manipulated solidarity in three-

person groups and found that the average degree of participation in

discussion was greater in the solidarity condition. In studying the

spread of rumors, Festinger, et al (1948) asserted that, in a group of

close friends, it is expected that the channels of communication will

be more numerous and open. Yalom (1970) states that anecdotal evidence

derived from therapy groups supports the notion that participation

increases with cohesiveness, and that this facilitates successful

outcome.

Some of the earliest efforts focused on the effects of cohesiveness

on group productivity. Several studies showed that cohesiveness

increased productivity, while others found conflicting results (Collins

& Raven, 1969). The most likely explanation is that highly cohesive

groups are more responsive to group norms; the norm is followed whether

it calls for increased or decreased productivity. The conflicting

results of the productivity studies have been useful in stimulating

research on the influence of groups on the members. This research has

consistently demonstrated that highly cohesive groups exert more

influence on group members than do less cohesive ones^ An early study

by Festinger, et al (1952), in which attraction to the group was

experimentally manipulated, showed that groups in the high attraction

condition exerted greater pressure for uniformity of opinion than did

low attraction groups. In 1954, Thrasher (Lott £• Lott, 1965) used a

perceptual judgement task and found that pairs of friends were signifi­

cantly more influenced by each other's judgements than were pairs who

were neutral toward each other. Another side of this issue was studied

by Schachter (1951)- Cohesiveness was experimentally manipulated in

groups and then stooges expressed opinions at variance with group

opinion. High cohesiveness groups showed less tolerance for the deviate

than did low cohesiveness groups. Zander, et al (I960), found that

members who were most attracted to their group had personal goals in

relation to a group task that were highly congruent with group goals.

After a review of the literature, Cartwright and Zander (1968) conclude

that, "There can be little doubt that members of a more cohesive group

more readily exert influence on one another and are more readily

influenced by one another" (1968, p. 104).

Another aspect of communication is the expression of hostility

within groups. Yalom (1970) says "Unless hostility is openly expressed,

persistent and impenetrable hostile attitudes may develop which

increasingly hamper effective interpersonal learning" (1970, p. 53).

There is evidence that highly cohesive groups are more willing to express

hostility directly than are less cohesive groups./ Pepitone and Reichling

(1955) produced high and low cohesiveness in experimental groups and

then systematically insulted the group members. The following

discussions showed that the highly cohesive groups were significantly

more able to express their hostility with intensity and directness than

were low cohesive groups. French (194l) compared the reactions of a y

group of strangers and a group of club members when put in a highly

frustrating problem-solving situation. He found that the club members

expressed more interpersonal aggression without disrupting group

functioning.

Insofar as cohesiveness is a function of attraction to a group, it

would be expected that members of cohesive groups would be likely to

remain members of the groups. Evidence to support this assumption

comes from Yalom (1970), who found that the least cohesive members of

four therapy groups terminated within the first twelve sessions. This I

fact has implications for the remaining members as well as for those l'

who dropped out. In the same study, Yalom found that groups with high |

turnover rates are less therapeutic for the remaining members than are \ \

stable groups. Highly cohesive groups also have a lower rate of absences i;

among continuing members, which contributes to a feeling of stability

(Cartwright S- Zander, I968). The correlation between absence rates and

cohesiveness is taken for granted by Berne and other transactional

analysts who use attendance rates as their main measure of group

cohesiveness (Kaplan S- Sadock, 1971).

A large body of evidence has thus been produced which supports the

general notion that group cohesiveness is desirable. Some of the

benefits that are thought to accrue from group cohesiveness are low

rate of absences, low dropout rates, and more frequent communication

among members. In addition, group members are better able to express

hostility without disrupting group process and exert greater influence

upon each other. All of these are thought to facilitate positive group

outcome. y

Techniques for Facilitating Group Cohesiveness

In most laboratory studies of group cohesiveness, the groups meet

only one time and the cohesiyeness is produced by expectations of

attraction. That is, members of groups are told before the meeting that

they have been placed in a group with people who have values and interests

very similar to their own. This technique has been found to produce

significantly higher levels of cohesiveness than that found in groups

given negative expectations. This is useful for laboratory studies,

but has an obvious limitation for other kinds of groups. It can be

used only once, before the group meets. Leaders of ongoing therapy,

encounter, or training groups want techniques that will allow them to

facilitate cohesiveness throughout the life of the group. This is

very important because the effects of cohesiveness are critical to

achieving the goals of such groups.

The use of structured exercises has become popular among groups of

all orientations except psychoanalytic. It is assumed that these

exercises have many beneficial effects, including increased group

cohesiveness, but only one study has systematically investigated

structured excercises. Lieberman, Yalom, and Miles (1973) made an

extensive study of encounter groups using 210 subjects divided into

18 groups and lead by leaders with ten different orientations.

Structured exercises were defined as leader interventions that include a

set of specific orders for behavior. The orders limit the participants'

behavioral alternatives. The results showed that high exercise groups

were significantly more cohesive both in early sessions and later in the

life of the group than were low exercise groups. The authors conclude

that, "it is clear that the use of exercises...have a significant effect

on one important group variable, cohesiveness" (1973, p. 4l6). Although

use of structured exercises did not live up to the expectation that they

would create intense emotional experiences, the results favored the use

of exercises for facilitating cohesiveness.

The study by Lieberman, Yalom, and Miles (1973) compared high and

low exercise groups, but there were no comparisons between specific

kinds of exercises. The authors concluded that the effects of structured

exercises are not known, aside from the correlation between stuctured

exercises and cohesiveness. Since that correlation has been found, it

would be useful to know if certain kinds of structured exercises are •

I more effective in facilitating cohesiveness than other kinds. |i

This study compared two kinds of structured exercises, verbal and j;

nonverbal. The main purposes of both kinds of exercises is to help ji

1; people become more open with each other. Lewis and Steitfeld, who i<

consider themselves part of the Human Potential Movement, suggest

that, "While no one knows exactly why groups work, certain catalysts

are almost always found in successful group reactions. Possibly the

most significant of these is the willingness to share yourself with

others... By opening up, you can help a group - and you - thrive" (1972, p. 205).

The interview exercise used in the present study was chosen because

it is a structured verbal exercise that Lewis and Streitfeld (1972)

recommend as an introductory activity for a new group. This activity

is designed to "open up" the participants and encourage them to reveal

8

important feelings and facts about themselves. This is supposed to

produce greater intimacy and cohesiveness than less structured methods

of introduction.

Structured nonverbal exercises are expected to produce the same

results as mentioned above, but in a different way. According to

Schutz (1967), nonverbal communication provides a valuable alternative

for those who have difficulty in expressing themselves openly. Schutz

also believes that many people are adept at talking without revealing

anything about themselves. Nonverbal exercises do not allow that

escape. An exercise for making contact used by Schutz is the "feeling

space" activity used in this study. The "breaking in" exercise that

was also included in this study is recommended by Schutz for getting

in touch with feelings of isolation. The act of physically entering

the group after exclusion will elicit a feeling of belonging to the

group. The last nonverbal exercise used in the present study comes

from the Human Potential Movement. Lewis and Streitfeld view this

activity as one that increases sensitivity to at least one individual in

a group. This is thought to be an experience that will unify the group.

Another method of facilitating cohesiveness is more indirect.

Causing the group to experience anxiety together is thought to bring

the group closer together and increase attraction within the group.

Silver (I967) asserts that, the less structured a group is, the more

capable it is of inducing anxiety. Lieberman, Yalom, and Miles (1973)

agree that members of heavily structured groups "were spared both the

anxiety stemming from the initial situational ambiguity and the frustration

of facing a leader who would not...lead" (1973, p. 303). In a classic

experiment, Schachter (1959) induced anxiety in subjects and found that

ambiguous situations lead to a desire to be with others, preferably

others who are experiencing similar feelings. Cartwright and Zander

(1968) conclude that, when members of a group encounter ambiguous

situations, their attraction to the group will increase.

For leaders of on-going groups, the use of structured exercises is

a convenient method of facilitating cohesiveness. Refusing the

traditional leadership role and not structuring the group experience

can also be expected to facilitate cohesiveness in an on-going group,

These are the methods investigated by this study.

Research Hypotheses

Group leaders employ certain techniques with the assumption that

they build cohesiveness. The leaders also assume that some techniques

are more effective than others. The major concern of this study was to I r

compare the effectiveness of each technique in building group cohesive- "'

5: ness. It was predicted that: I

r; (la) Each experimental group wi 11 demonstrate higher iji

cohesiveness than the control group.

(lb) Groups will differ from each other, although specific rank order is not predictable.

This study also sought to show that the use of different techniques

would elicit different levels of anxiety in the group participants.

It was predicted that:

(2a) The unstructured experimental group will have higher anxiety levels than the other experimental groups and the control groups.

(2b) The structured experimental groups and the control groups will differ from each other, although specific rank order is not predictable.

CHAPTER I I

METHOD

The subjects for this research consisted of four groups of

college undergraduates. Three groups participated in sessions that

included the use of techniques that are thought to facilitate group

cohesiveness. The fourth group served as a control. Cohesiveness and

state anxiety were measured in all groups at the end of the session.

Experimental Design

The research design employed was a posttest only control group

design (Campbell & Stanley, I966). Subjects were randomly assigned

to one of the four groups. Each group met in an initial session that

was lead by two co-leaders. In three of the groups, cohesiveness-

bui Iding techniques were used. The control group also met in an

initial session, but no formal attempt was made to build cohesiveness.

At the end of the initial sessions, all group participants completed

measures of cohesiveness and state anxiety.

Statistical Analysis

A one-way analysis of variance was used to determine the significance

of differences between cohesiveness and anxiety levels of the experi­

mental and control groups. Specific score comparisons on cohesiveness

and anxiety measures between groups were made using the Duncan New

Multiple Range test (Kirk, I968; pp. 93 £-94).

Subjects

Subjects for this study were 29 undergraduates in a Mental

10

11

Hygiene class at Texas Tech University, Lubbock, Texas. These subjects

were members of four groups, such membership being a requirement of the

course. The primary purpose of the groups was to give students an

intense interpersonal experience. Subjects were randomly assigned to

groups. Subjects ranged in age from 19 to 24, with a mean age of 21;

22 subjects were female and seven were male. Only two of the subjects

were psychology majors; none of the subjects had any group experience.

Instruments

Three measures of group cohesiveness were used. Multiple measures

were used because of problems concerning the methods specificity of

results in much personality and social psychology research.

1. Group Cohesiveness Questionnaire: This questionnaire was taken from

Libo (1953). The five questions measure the strength and direction of

feeling toward the group (see Appendix A).

2. Sociometric Questions: These three questions were designed to

measure friendship choices from among individual group members. The

criteria for sociometric choices were social criteria and rejection

criteria, as used by Mordock, et al (1969). (see Appendix B).

3. The Critical Incident Question: As used by Lieberman, Yalom, and

Miles (1973) this was used to obtain specific information concerning the

effects of group activity (see Appendix C).

4. STAI Questionnai re: The statements measuring state anxiety were

administered to all group members (see Appendix D).

12

Group Leaders

Six graduate students in the counseling psychology program at Texas

Tech University served as group leaders. All of the students had minimal

experience in this role. They were given detailed instructions on the

techniques and the rationale behind the techniques.

Procedure

Each of four groups met for the first time three weeks into the

fall semester. Each group met for approximately one and a half hours.

At the end of the sessions, all group members completed the measures of

cohesiveness and anxiety.

1. Group A (structured verbal exercise): Nine subjects participated in

this group. The structured verbal technique used was an interview tech-i

nique. The group split up into pairs; the persons in the dyads interviewed j <

each other for 20 minutes. The group then reconvened and each person J s

introduced his partner to the group. Group discussion occupied the r

ii remainder of the session. 1;

2. Group B (structured nonverbal exercises): Five subjects participated

in this group. After the group completed each nonverbal exercise, there

was discussion of the experience. At the beginning of the session, group

members sat in a tight circle on the floor with eyes closed. For five

minutes, each person felt the space around himself; this space over­

lapped with the space of other people, and physical contact could be made

or avoided. In the second exercise, the group stood in a circle, holding

hands. One by one, each member was excluded from the circle and had to

find a way in. In the last exercise, the group members split up into

pairs and sat on the floor facing their partners, eyes closed. One person

13

in each dyad explored the face of his partner with one hand while

touching the same features of his own face with the other hand. After

a few minutes, the other partner took the active role.

3. Group C (unstructured): Nine subjects were in this group. After

introductions were made, the group leaders explained that an encounter

group is run by the whole group, not by the leaders. They then declined

to lead the group and didn't talk unless questioned by group members.

4. Group D (control): Six subjects were In this group. The leaders

functioned as leaders but did not use any kind of structured exercises.

The group discussed several topics including the purpose of the group,

the procedure for getting into graduate school, and surgery under

hypnosis. No formal effort was made to build cohesiveness.

5

CHAPTER I I I

RESULTS

This chapter will report research findings. The results are first

discussed with regard to Hypothesis 1 and 2 and statistical tables will

be presented.

Hypothesis la

Each experimental group will demonstrate higher cohesiveness than the control group.

Hypothesis lb

Groups will differ from each other, although specific rank order is not predictable.

The cohesiveness level was measured by the Group Cohesiveness

Questionnaire completed by each participant. Because of multiple com­

parisons involved for testing each group against control for Hypothesis

la and against each other for Hypothesis lb, the Duncan New Multiple

Range Test was used. A mean score was determined for each group; a com­

parison of the group means using the Duncan New Multiple Range test

yielded no significant differences among the groups. The group means

and standard deviations are recorded in Table 1. A comparison of scores

using the one-way analysis of variance also yielded no significant

differences among groups. The statistics for this evaluation are

presented in Table 2.

Responses to the sociometric questions were almost identical

for all group members; thus, no statistical computation was made.

14

I'

15

TABLE 1

COHESIVENESS QUESTIONNAIRE SCORES: SUMMARY OF

GROUP MEANS AND STANDARD DEVIATIONS

Factor Group A Group B Group C Group D

Number of Subjects 9 5 9 6

Mean 23 23.20 21 23

Standard Deviation 3-590 1.72 3-3 2.44

TABLE 2

COHESIVENESS QUESTIONNAIRE: ANALYSIS OF

VARIANCE SUMMARY TABLE

Source SS df MS

Between Groups

Within Group

Total

26.23

264.80

291.03

3

25

28

8.74

10.59

83 N/S

16

With only a few exceptions, the group members answered "all" to the first

question: "If this group broke up and you were starting another group of

similar size, which members of this group would you want to include?";

"none" to the second question: "Which members of this group would you

want to exclude from the new group?", and; "all" to the third question:

"If you were having a party, which group members, if any, would you

Invite?"

Although not quantifiable, the responses to the critical incident

question provide interesting anecdotal information. These responses

are presented in Table 3.

Contrary to predictions, the measured levels of cohesiveness in the

experimental groups were not significantly higher than in the control

group. None of the techniques was more effective in building cohesiveness

than the procedure used in the control group. Hypothesis la is statisti­

cal ly invalid.

The levels of cohesiveness among the experimental groups were not

significantly different from each other. None of the techniques was

more effective in building cohesiveness than any other technique.

Therefore Hypothesis lb is statistically invalid.

Hypothesis 2a

The unstructured experimental group will have higher anxiety levels than the other experimental groups and the control groups.

Hypothesis 2b

The structured experimental groups and the control groups will differ from each other, although specific rank order is not predictable.

The anxiety level of group members was measured by the STAI State

Anxiety Questionnaire. A comparison of group means using the Duncan New

17

TABLE 3

RESPONSES TO CRITICAL INCIDENT QUESTION

Group A - structured verbal

1. Finding out about the personal lives of group members.

2. Felt it was a good start in trying to know and understand the other

members in the group. Feel it is a diverse enough group and that it

will open new aspects and viewpoints on relations with other people.

3. The most meaningful part of the group was actually meeting the

persons - that seems to be a part that I have found lacking in

myself lately.

4. Because of the reaction and what my partner said about me, 1 felt

more comfortable. I really feel close to her and enjoyed exploring

feelings together.

5. I liked the way of getting to know everyone. It seemed we learned

more than superficial stuff.

6. I liked getting to know everyone in the group because it made me

more at ease.

7. I thought it was nice to get to know everyone.

8. In crowds or groups, I am not a very outgoing person due to shyness,

insecurity, inferiority, etc. But with almost any one person alone,

I for some reason find it easy to trust and become open. Once that

ice is broken, everything is fine. This encounter session did this.

18

TABLE 3

I feel I can relate to the others because I know something of them

and they of me. I enjoyed it.

9. It gave me an opportunity help explore myself toward reassurance.

Group B - structured nonverbal

1. We all seemed comfortable with each other.

2. Feel closer to the group!

3. It is especially meaningful for me to experience discussing things

with other kids because I usually don't talk very much.

4. No response

5. The way everyone seems to be open for discussion. For not knowing

each other I felt everyone was pretty much at ease to discuss.

Group C - unstructured

1. It was an unstimulating experience in which I felt somewhat nervous,

I was uncomfortable sitting in a group without conversation.

2. We need to break the ice.

3. I feel that this group should be more direction oriented.

4. No response

5. I enjoyed the discussions concerning hypnosis as well as learning a

little about everyone else.

WKT-ifffS^^C'

19

TABLE 3

6. This group seems like everyone will eventually be freer and more

open. At this time I feel comfortable and not apprehensive at all.

I hope the group proves to be good.

7. It was important to me that we didn't do much and that it made me

expect much more next time.

8. Surgery under hypnosis.

9. We need to get to know each other better.

Group D - control

1. Right now there is nothing that especially strikes me one way or

the other. I am still mainly just curious about the whole thing.

I feel that I will enjoy the group.

2. The fact that it's a positive group session with little pressure.

I like the fact that we're not supposed to experience being

"uncomfortable" with each other.

3. Discussed PhD plan with grad students.

4. The emphasis on "positive" was reassuring and I'm looking forward to

participating. I don't expect miracles or easy answers, but

opportunities to gain insight into myself and others is always

appreciated.

5. I feel that in later meetings my group members should be easy to

20

TABLE 3

get to know, but I am very interested to see how they will react to

me and others.

6. I just enjoyed talking. 1 was worried about something, so it gave

me time to just sit and talk. Really this had been the longest I had

sat for about four days.

ii ii

21

Multiple Range test yielded no significant differences among groups.

Group means and standard deviations are recorded In Table 4. The anxiety

scores were also compared by means of a one-way analysis of variance and

no significant differences among groups were found. Results of this

comparison are summarized In Table 5.

Contrary to predictions, the state anxiety level in the unstructured

group was not significantly different from levels In the other groups.

Hypothesis 2a is statistically invalid.

The anxiety levels among the structured groups and the control

group were not significantly different from each other. Hypothesis 2b

is statistically Invalid.

22

TABLE 4

ANXIETY QUESTIONNAIRE SCORES: SUMMARY OF GROUP

MEANS AND STANDARD DEVIATIONS

Factor Group A Group B Group C Group D

Number of Subjects 9 5 9 6

Mean 47.78 54.20 46.11 51.33

Standard Deviation 13.028 6.046 10.769 7.95I

TABLE 5 if

ANXIETY QUESTIONNAIRE SCORES: ANALYSIS OF ||

VARIANCE SUMMARY TABLE

Source SS df MS

Between Groups

Within Group

Total

265.06

3.132.63

3.385.69

3

25

28

85.35

125.31

.68 N/S

w

CHAPTER IV

DISCUSSION AND CONCLUSIONS

In this chapter, conclusions will be drawn from the results of this

study. These results will be discussed with reference to previous

research and comments will be made regarding methodological and theoreti­

cal considerations.

The purpose of this research was to compare the effectiveness of

frequently used group techniques in building cohesiveness. A secondary

purpose was to compare the levels of anxiety elicited in group members

by the various techniques.

Results failed to support the contentions that (1) the group using

structured exercises would have greater feelings of cohesiveness than

groups not using structured exercises and (2) that various exercises

produced differential levels of cohesiveness. Results also failed to

show that groups in which leaders refuse to structure the group experience

generated higher anxiety levels than groups in which such structure was

provided.

Several researchers concerned with group cohesiveness have based

their conclusions on cohesiveness measured after one group meeting

(Cervin, 1956; Festinger, et al 1952; Schachter, 1951; Zander, et al,

I960; Pepitone & Reichling, 1955). The present study also used measure­

ments obtained after a single meeting. Lieberman, Yalom, and Miles (1973)

suggest that cohesiveness early and late in a group's life may be based

on different factors. Early cohesiveness in a structured group may

reflect comfort, lack of conflict, and gratification of dependency needs.

23

24

Later cohesiveness is more likely a reflection of feelings of being

valued by the group and feelings that the group is facilitating fulfill­

ment of personal goals. This suggests that the effects of structure on

group cohesiveness will be more marked early in group life. The early

measures of cohesiveness in the Lieberman, Yalom, and Miles study were,

however, taken after nine hours of group meetings. The effects of

structure as opposed to lack of structure in the present study might

have been more pronounced if measures had been taken after several hours

of group meetings.

Responses to the critical incident question may be interpreted as

support for the suggestion that greater differences would have been

found later in group life. All of the members of the structured verbal

and structured nonverbal groups expressed positive feelings about the

initial group sessions. In the unstructured group, reactions were mixed.

One response (Table 3, A-1) was completely negative; four responses

(A 5-8) were positive or at least hopeful; three responses (A-2, 3, 9) were

suggestions for improvement. The improvements that were suggested were

in the direction of more structure or increased group activity. If

the expectations of improvement were not confirmed after more group

sessions, it is expected that dissatisfaction would increase and would

be reflected in lowered cohesiveness scores for the unstructured group.

Presumably scores in the structured groups would remain on the same level

or increase.

Aside from the methodological considerations, it was most surprising

that there were no differences in effectiveness among the techniques

used. The unstructured group and the structured verbal and nonverbal

exercises were very different from each other, and yet they had no

25

differential effects upon the degree of cohesiveness or the levels of

anxiety. It was hoped that the results of this study would provide

evidence that one kind of technique was the most effective In building

cohesiveness. No such evidence was found. The lack of significance of

results In comparison with the control group suggest that the experimental

techniques did nothing to increase cohesiveness. It would be useful,

therefore, to look at each technique individually^

Information from several sources indicated that an unstructured

group situation would increase anxiety among group members and result In

high cohesiveness (Schachter, 1959; Lieberman, Yalom, S- Miles, 1973; Silver,

1967; Cartwright £• Zander, I968). The expectation of high cohesiveness

was not confirmed by this study. One apparent factor is that the lack

of structure did not produce significantly greater anxiety than was found

in the structured and control groups.

There is no ready explanation for the lack of discomfort in what

was expected to be a relatively uncomfortable situation. One factor

may have been the expectations of group members. None of the subjects

had any group experience; only two of 29 subjects were psychology majors.

Their lack of information, or possible misinformation, about groups may

have led them to have strong negative expectations. Thus the actual

group meeting, in that it involved little activity of any sort, may

have been an improvement over the expectations.

Again, the critical incident responses provide information that was

not supplied by the anxiety questionnaire. None of the 20 members of

the control and structured groups reported feeling uncomfortable; five

of the 20 specifically reported feeling comfortable. The only response

reporting discomfort came from a member of the unstructured group.

26

Group member C-1 said, "It was an unstlmulating experience In which I felt

somewhat nervous. It was uncomfortable sitting in a group without

conversation" (Table 3). This was the only report of discomfort, yet

this participant's anxiety score was not the' highest in his group; two

other members of that group had scores higher than his. The member with

the highest anxiety score (C-2) said only, "We need to break the ice"

(Table 3). More than one conclusion may be drawn from this. One is that

the anxiety questionnaire was not a sensitive instrument for measuring the

discomfort that was elicited by the unstructured nature of the group.

Another is that other members may have felt the same discomfort as the

one who expressed it on the critical incident question, but the discomfort

was not labelled as such or was not the most critical incident of the

session. It is also possible that the anxiety being experienced by those

with high anxiety scores was not related to the group experience, but to

outside factors; thus anxiety was not reported as part of the critical

incident.

It is also of interest that at least one member of the unstructured

group was quite comfortable. Participant C-6 said, "This group seems

like everyone will eventually be freer and more open. At this time I

feel comfortable and not apprehensive at all. I hope the group proves

to be good" (Table 3). The limitations of the critical incident question

notwithstanding, this response suggests that the demand characteristic of

the unstructured group may not be the same for everyone. That is, it

may be incorrect to assume that everyone will be anxious in a group in

which there is no formal leader. For some people. It might be much more

anxiety provoking to be in a group in which there was a leader who would

call upon members of the group to respond. In order to get a better

27

understanding of the nature of anxiety, future research might expand upon

the critical incident question and, since the present state anxiety

measures appear Inadequate, develop a questionnaire that would inquire

about anxiety as related to the group experience.

Although sociometric questions have been useful In the past (Mordock,

et al, 1969), they did not prove so In this study. Perhaps these questions

would have elicited more discriminating responses later In the life of the

group when group members were better acquainted.

Despite the positive correlation between the use of structured

exercises and cohesiveness obtained by Lieberman, Yalom, and Miles (1973),

the results regarding structured exercises were not all positive. Other

data from their study suggested that groups with heavy use of structured

exercises produced a lower proportion of positive changes in members than

did groups using few structured exercises. In comparing yield scores, all

low exercise groups produced some gain, while high exercise groups had

more variable gain scores. The authors suggest that some leaders are

less able than others to use exercises effectively, Theyconclude that,

"A...conservative interpretation is that exercises are irrelevant to

producing positive change (there are other leader strategies that seem

more productive) and that exercises used in concert with some leader

styles may be counterproductive" (1973, p. 412).

Timing of structured exercises is considered to be important in

insuring that the exercise will have the desired effects. This factor

was controlled for in the present study by using structured exercises

that are generally considered to be warm-up exercises. That is, they

are supposed to break the ice in a new group and help the participants

to get acquainted quickly. According to the responses to the critical

28

incident question, the interview technique used in the structured verbal

group was the most successful in this respect. Eight out of nine members

of Group A mentioned getting acquainted with other group members in the

session. Since the cohesiveness ratings were no higher for this group

than for any other groups, it may be that getting information about other

group members Is not strongly related to feelings of attraction to

the group as a whole. Another possibility is that such "getting

acquainted" answers indicate a first step in the development of cohesive­

ness. Again, the relationship of the critical incident question and the

other questionnaires must be considered.

In addition to the fact that the relationship between structured

exercises and group outcome is not clear (Lieberman, Yalom, and Miles,

1973), the immediate effects of structured exercises are not known. Two

out of four critical incident responses from the structured nonverbal

group mentioned enjoying the discussions. The other two said nothing

specific about the session. No one said anything about the nonverbal

exercises themselves. This could be interpreted as a reaction against

the anxiety-provoking nature of nonverbal exercises; If the exercises

were very threatening, the discussions may have been a great relief. The

anxiety questionnaire does not provide information on this point since it

measured anxiety at the end of the session. It is possible that the

responses are in keeping with the nature of nonverbal exercises. In

giving instructions for the "feeling space" exercise used in this study,

Schutz (1967) comments that, "Discussion following this activity is

usually very valuable in opening up the whole area of feelings about

loneliness and contact" (1967, p. 138). Perhaps the discussion of the

experience is at least as important as the nonverbal experience itself.

29

The issues discussed above point up the fact that little is known

about the effectiveness of cohesiveness-buiIding techniques or about the

source of any effectiveness they may have. As has been noted earlier, the

use of these techniques^by some leaders may even be counterproductive. If

this is the case, it would seem to stem from ignorance about the techniques

and their interactions with leadership styles and other group characteris­

tics. Despite the lack of data, the use of these techniques is widespread.

Although they did not compare the specific techniques, the model used by

Lieberman, Yalom, and Miles (1973) provides an excellent model for future

research with groups. Measuring cohesiveness at different times in the

life of the group, using several kinds of cohesiveness and outcome measures,

comparing groups on several dimensions, and employing different leader­

ship styles are all critical factors in research. It may also be necessary

to develop instruments that will assess more exactly the impact of

specific techniques on individual particpants and on the group as a whole.

CHAPTER V ^

SUMMARY

Prior research on small groups has established the desirability of

cohesiveness for facilitating positive group outcomes. Several techniques

are used by group leaders to build group cohesiveness, but there Is

little formal data to support the belief that any of them do. In fact.

Induce such cohesiveness; neither has there been any research that has

compared specific techniques as to their effectiveness in this regard.

This study Involved four groups of college undergraduates, each group

participating in one session. In three of the groups, cohesiveness-

bui Iding techniques were used. These included structured verbal and

structured nonverbal exercises and an unstructured group experience.

The fourth group served as a control. Cohesiveness and anxiety levels

were measured at the end of the initial group meetings.

Results did not support the use of the techniques investigated. There

were no significant differences in cohesiveness among any of the groups,

nor was there significant difference in anxiety level among the groups.

The results were discussed with reference to previous research, and

with regard to methodological and theoretical considerations.

30

REFERENCES

Campbell, D. T. , and Stanley, J. C., Experimental and Quasi-Experimental Desiqns for Research. Chicago: Rand McNally College Publishinq Co., 1966. J ^ b ,

Cartwright, D., and Zander, A., Group Dynamics: Research and Theory^. (3rd ed.) New York: Harper £• Row, I968.

Cervin, V., Individual behavior in social situations": Its relation to anxiety, neuroticism, and group solidarity. Journal of Experimental Psychology. I956, 51, I6I-I68.

Collins, B. E., and Raven, B. H., Group structure: attraction, coalitions, communication and power. In G. Lindzey and E. Aronson (eds.). Handbook of Social Psychology. Vol. IV. Reading, Mass.: Addison-Wesley Publishing Company, 1969-

Festinger, L., Cartwright, D., et al, A study of rumor: Its origin and spread. Human Relations, 1948, 1, 464-468.

Festinger, L., et al. The influence process in the presence of extreme deviates. Human Relations, 1952, 5, 327-346.

French, J. R. P., Jr., The disruption and cohesion of groups. Journal of Abnormal and Social Psychology, 1941, 36, 361-377.

Kaplan, H. I., and Sadock, B. J., Comprehensive Group Psychotherapy. Baltimore: The Williams S-Wilkins Company, 1971.

Kelley, H. H., and Thibaut, J. W., Experimental studies of group problem solving and process. In G. Lindzey (ed.) Handbook of Social Psychology, Vol. 2. Reading, Mass: Addison-Wesley, 1954.

Kirk, R. E., Experimental Design: Procedures for the Behavioral Sciences. Belmont, California: Brooks-Cole Publishing Company, I968.

Lewis, H. R., and Streitfeld, H. S., Growth Games. New York: Bantam Books, 1972.

Libo, L. M., Measuring Group Cohesiveness. Ann Arbor, Mich.: University of Michigan Press, 1953.

Lieberman, M. A., Yalom, I. D., and Miles, M. B., Encounter Groups: First Facts. New York: Basic Books, 1973.

Lippitt, R., and White, R. K., An experimental study of leadership and group life. In G. E. Swanson, T. M. Newcomb, and E. L. Hartley (eds.) Readings In Social Psychology. New York: Holt, 1952. pp. 340-355.

31

32

Lott, A. J., and Lott, B. E., Group cohesiveness as interpersonal attraction: a review of relationships with antecedent and consequent variables. Psvcholoalcal Bulletin. I965, 64(4), 259-309.

Mordock, J. B., Ellis, M. H., and Greenstone, J. L., The effects of group and Individual therapy on sociometric choices of disturbed. Institu­tionalized adolescents. International Journal of Group Psychotherapy. 1969, 19(4), 510-517.

Pepitone, A., and Reichling, G., Group cohesiveness and the expression of hostility. Human Relations. 1955, 8, 327-337.

Scachter, S., Deviation, rejection, and communication. Journal of Abnormal and Social Psychology. I951, 46, 190-207.

Schachter, S., The Psychology of Affiliation. Stanford: Stanford Uni­versity Press, 1959.

Schutz, W. C , Jov: Expanding Human Awareness. New York: Grove Press, 1967.

Sliver, A. W., Interrelating group dynamic, therapeutic, and psycho-dynamic concepts. International Journal of Group Psychotherapy, 1967, 17 (2), 139-150.

White, R., and Lippitt, R., Leader behavior and member reaction in three "social climates". In D, Cartwright and A. Zander (eds.) Group Dynamics: Research and Theory (2nd ed.), Evanston, 111.: Row, Peterson, i960.

Yalom, I. D., The Theory and Practice of Group Psychotherapy. New York: Basic Books, 1970.

Zander, A., et al. Personal goals and group's goals for the members. Human Relations, I96O, 13, 333-344.

APPENDIX A

GROUP COHESIVENESS QUESTIONNAIRE

Do you want to remain a member of this group?

yes_ no undecided

If you checked yes or no, how stongly do you feel about your preference?

very strong strong moderate slight

very siIght

How often would you like to come to meetings of this group?

more than twice a week twice a week once a week

once a month less than once a month not at all

If this group broke up for a considerable length of time and some people

were trying to get it started again, would you want to rejoin?

yes_ no undecided

*; ._ i^J^

If you checked yes or no, how strongly do you feel about your preference?

very strong strong moderate slight

very slight

33

34

APPENDIX B

SOCIOMETRIC QUESTIONS

If this group broke up and you were starting another group of similar

size, which members of this group would you want to include?

Which members of this group would you want to exclude from the new group?

If you were having a party, which group members, if any, would you invite?

35

APPENDIX C

CRITICAL INCIDENT QUESTION

On the back of this page, briefly describe anything in this group meeting

that was especially meaningful or important to you.

i>^'*

mi

36

APPENDIX D

SELF-EVALUATION QUESTIONNAIRE

1.

2.

3.

4.

5.

6.

7.

8:

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

feel calm

feel secure

am tense

am regretful

feel at ease

feel upset

am presently worrying over possible misfortunes

feel rested

feel anxious

feel self-confident

feel comfortable

feel nervous

am j ittery

feel "high strung"

am relaxed

feel content

am worried

feel over-excited and "rattled"

z o rt

> rt

> —• —'

c/> o 3 (D

s 3 " Q) rt-

o Q . (D -J 01 r* O ^

«< t/) O

< (D ~\

•<

2 C o zr CO O

feel joyful

feel pleasant

-1 ' .Ml