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
41
Embed
a comparison of techniques for building cohesiveness in small ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
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.