LIBRARY
Michigan State
University

 

 

 

THES‘S

This is to certify that the

thesis entitled

Participant and Observer Perspectives on the
Interpersonal Climate of Small Groups

presented by
Lori J. Slough

has been accepted towards fulfillment
of the requirements for

Master'schgeeh,Psxchology

     

Major profes 1'

Date 5/15/86

0-7839 MS U 1': a! Waive Action/Equal Opportunity Institution I

 

MSU

LIBRARIES
155—3—-

 

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

.‘ 4'
J
.
. .
’ / 1"

 

 

 

PARTICIPANT AND OBSERVER PERSPECTIVES ON THE
INTERPERSONAL CLIMATE OF SMALL GROUPS

BY

Lori J. Slough

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

1985

ABSTRACT

PARTICIPANT AND OBSERVER PERSPECTIVES ON THE
INTERPERSONAL CLIMATE OF SMALL GROUPS

by

Lori J. Slough

Derived from diverse theories and empirical studies,
hypotheses concerning developmental shifts of interpersonal
climate within small groups were formulated for exploration.
Data were collected through routine end-of—session
administrations of MacKenzie's (1978) 12-item Group Climate
Questionnaire to nonparticipating observers and all 82
members of 11 groups that convened about 20 times for nearly
SO-hours over nine weeks. Interitem correlations were
computed independently for groups' early, middle, and late
sessions, separately for members and observers. Analyses of
these six matrices identified a progressively clearer

bipolar cluster, designated Affiliative (items Cared,

 

Confronted, Participated, Reasoned, & Revealed) versus

Disaffiliative (items Angry, Anxious, Avoided, Distanced, &

 

Rejected), that overshadowed initially hypothesized shifts
on individual items. Despite bottoming near midgroup,

Affiliative total ratings generally strengthened with more

 

group experience. Largely complementary, Disaffiliative

 

ratings peaked at midgroup. Members' ratings were much more
favorable and less dispersed than were observers', although

each source's data supported these patterns.

ACKNOWLEDGMENTS

I would like to extend great thanks to the members of
my committee, John Hurley, Bob Caldwell, and Dozier Thornton
for their assistance with this project and for guiding my
development as a researcher. Particular thanks goes to John
Hurley whose enduring support and direction made this study

possible.

ii

TABLE OF CONTENTS

List of Tables . . . . . . . . . . . . . . . . .
List of Figures . . . . . . . . . . . . . . . .
CHAPTER I - INTRODUCTION . . . . . . . . . . . .
Literature Review . . . . . . . . . . . . .
Theories of Group Development . . . . . . .
Empirical Studies of Group Development . .
Group Climate . . . . . . . . . . . . . . .

Divergent Perspectives of Observers and
Group Members . . . . . . . . . . . . . .

Group Climate and the Principal Dimensions
of Interpersonal Behavior Hypotheses . .

CHAPTER II - METHOD . . . . . . . . . . . . . .
Participants . . . . . . . . . . . . . . .
Measures . . . . . . . . . . . . . . . . .
Procedure . . . . . . . . . . . . . . . . .

CHAPTER III - FINDINGS . . . . . . . . . . . . .
Interperiod Shifts . . . . . . . . . . . .
Correlational and Cluster Analyses . . . .

Linkage to the Central Dimensions of
Interpersonal Behavior . . . . . . . . .

CHAPTER IV - DISCUSSION 0 o o o o o o o o o o 0

REFERENCES 0 O O O O O O O O O O O O O O O O O 0

iii

PAGE

 

vi

ll
13

15

19
22
22
23
26
28
28

32

38
42

55

APPENDICES O O O O O O O O O O O O O O O O O O O O O O

A-l

A-2

Group Climate Questionnaire (GCQ-S,
ObserverS) C O O O O O O O O O O O O O O O O O C

Group Climate Questionnaire (GCQ-S,
Members ) O O O O O O O O O O O O O O O O O O O 0

Members' Mean Ratings of Group Climate Items,
Standard Deviations, and t-Values of Inter-
periOd Shifts O O O O O O O O O O O O O O O O O O

Observers' Mean Ratings of Group Climate Items,
Standard Deviations, and E-Values of Inter-
periOd Shifts O O O O O O O O O O C O O O O O O O

Hotelling's 52 for Multiple Analysis of

Variance of Members' and Observers' Ratings
Across Early, Middle, and Late Sessions, Followed
by Univariate E-test (l, 20) Results . . . . . .

Hotelling's 52's for Multiple Analyses of
Variance of Group Climate Ratings Across Early,
Middle, and Late Sessions by Members and
Observers, Followed by Variate E-tests . . . . .

Early, Middle, and Late (E, M, L) Inter-
correlations Among 12 Climate Items for
Member-Based Ratings with Eleven Groups . . . . .

Early, Middle, and Late (E, M, L) Inter-

correlations Among 12 Climate Items for
Observer-Based Ratings with Eleven Groups . . . .

iv

59

59

6O

61

62

63

64

65

66

Table 1.

Table 2.

LIST OF TABLES

PAGE

Intercorrelations Among 12 Climate Items
for Member-Based Ratings of Eleven Groups
During Late Group Sessions . . . . . . . . . 34

Product-Moment Correlations of Members'

Mean ARS and ARO Ratings of Group Units

After About 20— and 40-Hours by Self and

by Peers with their Rating on the Subclusters
of Affiliative and Nonaffiliative

Items on Three Periods . . . . . . . . . . . 40

 

Figure 1.

Figure 2.

Figure 3.

LIST OF FIGURES

PAGE
Group Climate Shifts of Members and
Observers for Early, Middle, and Late
Time Periods . . . . . . . . . . . . . . . 29

Intercorrelational Linkages of 12 Group
Climate Items at Three Periods . . . . . . 35

Mean Rating of Each Session of 28 Groups
by All Members . . . . . . . . . . . . . . 47

vi

INTRODUCTION

Small groups connect people to society and when
individuals are cut off from significant groups they usually
feel alienated and confused (Durkheim [1894], 1951). Small
groups have been classified into two general types, primary
groups and task groups (Ridgeway, 1983). Cooley (1909) used
the word primary to describe groups having significant emo-
tional attachments, relative permanence, and a nonspecific
purpose. The nonspecific purpose of these groups occurs
because emotional relationships are the focus. Dunphy (1972)
divided primary groups into four types:

1. Families

2. Free-association peer groups of childhood,

adolescence, and adulthood; delinquent gangs
and some small cohesive political elites
(cabals); most close friendship groups
3. Informal groups existing in organizational
settings such as classroom groups, factory
work groups, small military units
4. Resocialization groups such as therapy
groups, rehabilitation groups, and self-
analytic groups (Ridgeway, 1983, p. 16).
Ridgeway (1983) has said ”Primary groups are the one place a
person can go to be responded to as a whole person" (p. 14).
Dunphy believes it is within these groups that people learn
the rules and restraints of society. Task groups exist to
accomplish specific goals that range from making a policy

decision, solving a complex problem, producing a product, to

playing in a symphony, or moving a piano.

1

2

The boundaries between primary and task groups are not
always clear but their predominant functions are the usual
basis of classification. This study focuses on the small
interpersonal skills group, a subvariety of the self-
analytic or resocialization group. More specifically, this
work will explore the development of interpersonal skills
groups over time. According to Schutz (1955) the inter-
personal perspective, central to this study, is relevant to
a wide variety of groups because the efficiency of inter-
personal interaction determines the amount of energy which
will be available for task fulfillment functions.

Although therapy, rehabilitation, and self—analytic
groups have different purposes and orientations, theories
and studies based on one type of resocialization group are
often generalized to the others. Two primary viewpoints are
currently used to understand the development of inter-
personal skill enhancement groups. Most widely supported is
the view that such groups progress through identifiable
stages (Bennis, 1964; Tuckman, 1965; Yalom, 1975;
Lacoursiere, 1980). The substantiating evidence is largely
anecdotal and unverified (Cissna, 1984), however, and empir-
ical studies have not generally supported any consistent
patterns of group development (Lakin & Carson, 1964; Lubin &
Zuckerman, 1967). Gibb (1964) has presented evidence that
these groups move through major developmental phases much
less systematically and maintained that focal issues salient

to all groups recur intermittently. This perspective is rem-

3
iniscent of Bion (1959), whose group culture studies were
influenced by Freud's efforts to integrate ego and group
psychology of large groups or "masses." The present study
proposes to explore group development by systematically
assessing interpersonal climate. Exploring group climate,
this study will attempt to link developmental patterns found
in the present interpersonal groups to existing divergent

theories of group evolution.

LITERATURE REVIEW

Theories of Group Development

The term ”group development“ has been used to describe
aspects of group functioning ranging from specific changes
in relationships (intermember and leader-member) to general
trends across the life of a group, such as shifts in anxiety
and cohesion (Lacoursiere, 1980). In the present work group
development refers to a sequence of behavioral norms that
evolve from the interactions of group members (Mackenzie,
1978). Group development has also been described in terms of
stages, phases, trends, and cycles. Stages and phases com-
monly refer to predictable sequences of events which occur
at particular points of a group's existence (Lubin & Zucker-
.man, 1967). According to the dictionary of psychological
terms (English & English, 1958), as noted by Lacoursiere
(1980), stages are discrete divisions, phases are states in
a series of changes, and trends are dynamic tendencies.
Schutz (1964) defined group cycle as a recurring process by
which change occurs.

Attention to the prepotent features of group inter-
action is necessary to any comprehensive understanding of
group functioning. Lacoursiere (1980) contended that social-
emotional behaviors and the level of task fulfillment are
the most salient features of any group. Moos (1974) has pre-

sented evidence that "environments” can be adequately char-

4

5

acterized by the three general factors: relationship, per-
sonal development, and system maintenance. Relationship fac-
tors refer to involvement, support, and expressiveness. Per-
sonal development factors concern the specific tasks
addressed by a group, while system maintenance consists of
organization, clarity of expectation, and control within the
environment. Group climate, as defined by Moos, appears to
be a meaningful conceptual approach to the development of
interpersonal groups because it addresses both the task and
social-emotional aspects of such groups, described as
important by Lacoursiere (1980). MacKenzie (1978) developed
a measure of group climate which assesses the relationship,
personal, and system maintenance aspects of an interpersonal
group. Because this measure was sensitive to the predominant
aSpects of interpersonal group interaction it appeared to be
a reasonable measure of group development.

As with many areas of clinical psychology, Freud was a
pioneer examiner of the dynamics of group interaction. In

Group Psychology and Analysis of the Ego, Freud (1922) com-

 

mented that "each member is bound by libidinal ties on one
hand to the leader . . . and on the other hand to the other
members of the group" (p. 45). Freud believed that the pro-
jection of ego ideals onto the leader created unity and a
reduction of ego and superego functioning among group mem—
bers (Kernberg, 1984). Although Freud's ideas about group
processes originated in concerns for large organizations,

such as the church and the army, the issues of authority,

6
intimacy, and introjected self-concept addressed by Freud
have consistently been identified with the interpersonal
dynamics of small groups. While Bennis and Sheppard (1956)
and Bion (1959) seemed most directly influenced by Freud's
group theories, the primary issues that concerned Freud were
reflected in most subsequent conceptualizations of group
development.

Providing another early theoretical approach to group
dynamics, Bion's (1959) central idea was that two abstract
processes undergird every functional group: the "work group"
and the ”basic assumption" group. According to Bion, the
work group functions emerge from the group's task and paral-
lel the ego functions of an individual. "Basic assumptions"
refer to behaviors and thoughts manifested in groups that
are usually beyond the awareness of individual members. The
three basic assumptions hypothesized by Bion are: depen-
dency, pairing, fight-flight. Dependency refers to the group
members' reliance on the leader for psychological nurturance
and/or material rewards. Pairing is the mental state active
in the group's aspirations to produce a Messiah to solve all
possible problems. The third basic assumption, fight-flight,
is that the group exists either to fight or to flee some-
thing. Bion held that all assumptive behavior was instinc-
tual and that neither the work group nor the basic assump-
tion group exists as a pure culture. ”What one sees in real-
ity is a work group which is suffused by, intruded into, and

supported by the assumption groups" (Rioch, 1970, p. 62).

7

Further building on the ideas of Bion and Freud, Bennis
and Sheppard (1956) formed a theory of group development
emphasizing the concepts of dependence (authority relations)
and interdependence (personal relations). They hypothesized
two primary phases, each composed of three subphases. Phase
one, Dependence, is composed of dependence-flight,
counterdependence-flight, and resolution-catharsis. During
dependence—flight group members were thought to turn towards
group leaders for direction and, if allowed, would discuss
issues existing outside of the group. Counterdependence was
characterized by "fighting" among group members, distrust
and ambivalence towards leaders, and division into allying
subgroups. Intense involvement in the group task was thought
to occur during resolution. There would be discussion about
the trainer's role, the group would unify, and develop an
internal authority system. Phase two, Interdependence,
included enchantment-flight, disenchantment-flight, and
consensual validation. During enchantment the group ”becomes
a respected icon beyond further analysis” (Tavistock
Institute, 1956). Spirits were thought to be high and
laughter to abound. Disenchantment followed and during this
period members questioned the goals of the group and their
own commitment to those goals. During the phase of
consensual validation acceptance and understanding were
hypothesized to be prevalent and assessment of participation
to occur. Qualifying this, Bennis wrote:

. . . in this attempt to generalize into a sys-
tematic theory the sequential relations of group

8

life, there has been a tendency to force into

categories behaviors and actions which are more

indeterminate and overlapping than the theory

implies (p. 269).

Of six groups Bennis tested this theory on, only one fol-
lowed all of the predicted stages.

Separate from the psychodynamic perspective of Bennis
and Sheppard, Tuckman (1967) suggested that most theories of
group deve10pment fit one model. He reviewed fifty articles
related to stages of group development. These articles were
based on studies of therapy groups, T-groups, natural and
laboratory groups. In his discussion of group development
Tuckman considered task groups and interpersonal groups
separately.

The model of group development derived from this anal—
ysis of interpersonal groups contained the stages: Testing-
Dependence, Conflict, Cohesion, and Functional Role Related—
ness. The term Testing-Dependence was based on the attempt
of group members to understand what behaviors are appro-
priate in the group based on the reactions of the trainer or
therapist. During the phase labeled Conflict group members
were thought to become hostile toward each other and the
trainer in an effort to resist conformity to group struc-
ture. The third phase, Cohesion, was characterized by accep-
tance of the group and the idiosyncracies of other group
members. During Functional Role-Relatedness, Tuckman's final

phase, group member roles were thought to be well estab—

lished and the group turns to task related issues.

9

The stages Tuckman labeled Testing-Dependence,
Conflict, and Cohesion were identified in the majority of
therapy and T—groups. The only noteable deviation from these
groups was the lesser tendency of Conflict to appear in
therapy groups (13 of 26). In reference to the final stage
of T-groups, Tuckman did not report how many identified what
he termed Functional Role-Relatedness. He stated:

There is some tendency for T-groupers, as there

was for therapy groupers, to emphasize the task

aspects of the final stage, namely the emergence

of insight into the interpersonal process (Tuck—

man, 1956, p. 393).
He proposed that a group's increased ability to focus on the
interpersonal task was due to support and opportunity for
experimentation provided by the group. Tuckman aptly dubbed
the four stages of group development which he identified by
his review as forming, storming, norming, and performing.

Analyzing therapy groups, training groups, classes,
families, teams, and committees, Gibb (1964) also articu-
lated what he considered the four basic concerns of personal
growth in groups. According to Gibb, as groups move toward
actualization, they struggle with the focal concerns of
acceptance, data flow, goal-formation, and control. Gibb
believed that acceptance of self and others, and consequent
reduction of fear of self and others, was vital to increased
confidence and trust. Data flow referred to the behavioral
norms influencing the degree of spontaneity versus caution

acceptable in the group. Goal—formation was the reevaluation

of personal and group motives, and the conversion of this

10
knowledge into action. Concerns of Control referred to the
desire of group members to manipulate their behavior and the
behavior of others. Gibb claimed Control was strongly
related to Acceptance and Data Flow. He also reported that
defensive personal needs (punishment, distance, persuasion,
and control) declined in constructive groups while growth
needs (love, intimacy, realization, freedom) increased.
Although he regarded acceptance to be the catalyst for move-
ment on the three remaining factors, he stated that "what
seems most likely is that group growth is gradualistic and
global, in which themes and subthemes may intertwine but in
which the dramatic quality is wholeness, or the Gestalt"
(Gibb, 1964, p. 289).

Yalom (1975) articulated three general phases of group
development characterized by (a) orientation, hesitance, and
search for meaning; (b) conflict, dominance, and rebellion;
and (c) cohesiveness. Additionally, Yalom noted that group
interaction was not rigidly tied to these stages. Making no
explicit reference to phases, Yalom, like Gibb, contended
that issues intermittently arose, receded, and resurfaced to
be dealt with more thoroughly. Yalom has cited D. A. Hamberg
(personal communication, 1968) as referring to the
reemergence of these common issues as ”cyclotherapy."

Following a thorough review of empirical and theoret-
ical documents concerning group development, Lacoursiere
(1980) recently introduced a theory centered on stages of
Orientation, Dissatisfaction, Resolution, Production, and

Termination. Lacoursiere's theory is similar to Tuckman's

11
except for his attention to termination. The present author
found no other theory of group development that included the
period of loss and mourning which group members often exper-
ience at groups' end. According to Lacoursiere, although
these primary characteristics are particularly visible with-
in their respective stages, considerable overlap and ”hints
of most stages can often be found during each of the other
stages" (p. 28). Lacoursiere's hypotheses compromised
elements of stage and simultaneous-process theories.

Bennis (1964), Tuckman (1965), Yalom (1975), and
Lacoursiere (1980) seem to agree that interpersonal groups
are characterized by an initial orientation or dependence
phase, followed by a period of increased conflict, which in
turn leads to cohesion or resolution. Bennis's conceptual-
ization differs from the others in that it has two periods
of both conflict and resolution. These authors seem to agree
that the final phase of group interactions contains high
levels of acceptance and cohesion but Tuckman and
Lacoursiere additionally claim that this is also a period of
increased productivity. Bennis, Yalom, and Lacoursiere all
noted that the theories they put forward represented gen-
eral, but not rigid, behavioral trends. The theories of Bion
(1959) and Gibb (1964) were largely divergent from the

previously mentioned authors.

Empirical Studies of Group Development

 

12

Although Philip and Dunphy (1959) and Bales and
Strodtbeck (1951) were able to empirically derive a pattern
of development common to problem-solving groups, studies of
interpersonal group development have rarely yielded equally
clear patterns. Two works appear representative of the dis-
parity between empirical investigation and theory in this
area. Studying four interpersonal training groups that met
for 12 sessions during a single week, Lubin and Zuckerman
(1967) collected data using the Multiple Affect Adjective
Check List (MAACL) and five perceptual-cognitive task
aSpects of these groups. The MAACL yielded scores of
anxiety, depression, and hostility, and the other five
scales addressed worth of session, activity level, sharing
of feelings, amount of conflict, and cogency of group con-
tent.

Analysis of variance indicated significant session

meeting differences on all eight variables. Some

similarity of group trends over sessions, there—

fore, is present. However, group by session inter-

actions were significant for six of the eight var-

iables, which implies some degree of dissimilarity

of the group trends over sessions (for the differ-

ent groups). Thus the hypothesis of consistency of

trends from one group to another is not supported

(Lubin & Zuckerman, 1967, p. 365).
In spite of these findings Lubin and Zuckerman (1967) noted
that all of the variables measured peaked during session six
of eight sessions and that as feelings were shared more
openly, anxiety, hostility, and depression declined. While

the present study does not address differences between

groups, the work of Lubin and Zuckerman (1967) is important

13
because it is one of the few systematic studies of inter-
personal group development.

Similar findings emerged from Lakin and Carson's (1964)
investigations of the extent to which participants of inter—
personal training groups perceived developmental changes.
They collected data from four groups attending a 16-session,
two—week training program at a state mental health agency.
Participants' ratings of each session on 11 variables dis-
played considerable intergroup variability.

Because there is currently no singly accepted theory of
group development, and because the purposes and duration of
such groups vary greatly, little empirical work has been
conducted on the length of group stages. Bales and
Strodtbeck (1951) reported that common stages were evident
in single sessions, but Psathas (1960) found no evidence of
sequence conformity. While considerable anecdotal material
supports the existence of uniform developmental patterns,
regardless of group duration, empirical evidence has not led

to a single model of interpersonal group development.

Group Climate

 

The usefulness of group climate as a measure of inter-
personal group development has previously been discussed.
Following a thorough survey of the relevant literature,
MacKenzie (1978) found no reliable measure of group climate.
MacKenzie (1979) subsequently developed the Group Climate

Questionnaire (GCQ) with concern for the importance of

l4

describing ”not what should, might, or could occur in . . .
a group, but what actually does happen” (p. 473).

MacKenzie's (1978) original GCQ had 32 items. Factor
analysis of these items revealed the following clusters:
engagement, support, practicality, disclosure, cognition,
challenge, conflict, and control. A shortened version of the
Group Climate Questionnaire (GCQ-S), contains 12 items
derived from those clusters. Based on their predominant
emphasis these 12 items will be referred to as: Revealed,
Participated, Reasoned, Confronted, Cared, Normative, Angry,
Depended, Avoided, Anxious, Distanced, and Rejected. GCQ—S
clusters are represented by either one or two items. This
accounts for the discrepancy between the original number of
clusters and the number of items appearing on the shortened
Group Climate Questionnaire. A subsequent GCQ-S factor anal-
ysis yielded (MacKenzie, 1983) scales labeled Engaged,

Avoiding, and Conflict. In the 1983 American Group Psycho-

 

therapy Monograph Series, MacKenzie reported that the GCQ

 

was a clinically useful measure of group progression. He
asserted that the dominance of particular group behaviors
surface with the GCQ because raters consider the behavior of
all group participants for an entire session. He viewed
GCQ-S as particularly useful for understanding blocked
interactions noting that the ratings of "stuck” groups often
reflected

difficulty identifying the relationship between

positive bonds and interpersonal work; they do not

conceptualize the idea of avoidance of personal

initiative; and friction is seen only in relation-
ship to meeting group expectations (p. 168).

15

Although MacKenzie (1983) administered the GCQ-S to members
of 12 psychotherapy groups that met for approximately 35
sessions, he only analyzed data from the first 14 sessions.
MacKenzie (1983) identified sessions from 1 through 4 as
featuring Engagement, sessions 5 through 6 as featuring
Avoidance, sessions 7 through 10 as featuring rising Con—
flict and decreasing Avoidance, and sessions 8 through 14

with increasing Engagement.

Divergent Perspectives of Observers and Group Members
Related to empirical considerations of group develop-
ment, it appears important to consider the source of group
development ratings. Studies of group development often rely
on data collected from observers (Bales & Strodtbeck, 1951;
Lieberman, Yalom, & Miles, 1973; Psathas, 1960; Stock &
Thelen, 1958) in addition to participants. Notable differ-
ences in how events are perceived by active participants
(actors) versus others (observers) have been documented by
Nisbett and Jones (1971). Actors consistently attribute
their own behavior to situational factors, while observers
tend to attribute actors' behavior to actors' stable dis-
positional factors. Additionally, Cunningham, Starr, and
Kanouse (1979) reported that passive observers attributed
negative interactions to actors more often than the actors
attributed negative dispositions to themselves. In their
classic study of encounter groups, Lieberman, Yalom, and

Miles (1973) also found that detached observers rated the

16
behavior of leaders less favorably than did participating
group members. Due to the frequent use of Observers' data
and the well-documented divergent perspectives of partic-
ipants versus observers, it is important that data be
assembled from each source. While not a direct study of
attribution, significantly lower ratings of group climate by
nonparticipant observers versus group members would be gen-
erally supportive of the work of Nisbett and Jones (1971)
and specifically consistent with the findings of Cunningham

et a1. (1979).

Group Climate and the Principal Dimensions of Interpersonal
Behavior

As mentioned earlier, in addition to the differing per-
spectives held by group members versus observers, large dis-
crepancies have been reported (Lakin & Carson, 1964; Lubin &
Zuckerman, 1967) in the perceptions of groups participating
in programatically similar experiences. It is conceivable
that such differences have also been a function of unique
personality constellations that formed the culture of these
groups (Schutz, 1955).

Wholly independent of this information about group
climate and person perception, agreement has apparently been
reached (Hurley, 1980; Kiesler, 1983; Wiggins, 1982) within
the broader domain of interpersonal literature that only two
principal bipolar dimensions, often labelled affiliation and

dominance (Wiggins, 1982), undergird a wide variety of

 

l7
empirical findings. Hurley (1976, 1980) articulated the
primacy of these two dimensions in diverse theories of per-
sonality and psychopathology including those of Benjamin
(1974; Berne (1966); Foa (1961); Leary (1957); Lorr, Bishop
& McNair (1965); and Symonds (1939). Many other authorities
have also recognized the centrality of these two dimensions
to various theories of personality (Adams, 1964; Carson,
1969; Freedman, Leary, Ossorio, & Coffey, 1951; Kiesler,
1983; Schaefer, 1961). Hurley's (1976, 1984) group-oriented
measures of Acceptance versus Rejection of Others (ARC) and
Acceptance versus Rejection of Self (ARS) appear to assess
affiliation (ARC) and dominance (ARS) with reasonable
adequacy.

These ratings have also been positively linked to group
outcome (Hurley, 1978), as the mean ARS and ABC ratings of
individuals by all other group members appeared to correlate
positively with the interpersonal gains of clients within a
SO-hour psychotherapy group. ARS and ARO ratings also cor-
related highly (median ; = .71) with the interpersonal
effectiveness of paraprofessional child workers as judged
separately by self, patients, co-workers, and professional
mental health supervisors (Small & Hurley, 1978). Addition-
ally, among mental health professionals attending annual
institutes of the American Group Psychotherapy Association,
ratings of leaders' ARS and ARO behaviors by members of

interpersonal groups have been found to link consistently

18
and positively with group members' mean ARS and ARO gains
(Hurley & Rosenthal, 1978a, 1978b; Hurley, 1984).

Separate from these findings, Lieberman, Yalom, and
Miles (1973) linked the climate of interpersonal groups to
the benefits realized by group members. Their study,
reported in Encounter Groups: First Facts, provided the
single most comprehensive exploration of interpersonally-
oriented groups. They found that elements of group climate
correlated positively with their multifaceted and elaborate
measure of “group yield” or outcome. Their group observers
completed a questionnaire consisting of 12 semantic differ-
ential items (tense/relaxed, fast/slow, angry/harmonious,
etc.) after each group session. Factor analyses of these
ratings identified dimensions labeled "involvement inten-
sity" and ”harmony/anger" as primary. Late in the exper-
ience of their 18 30-hour long groups, "group yield“ was
found to correlate positively with both "involvement
intensity” (.56) and ”harmony” (.52). The studies of Hurley
(1978, 1984) and Hurley and Rosenthal (1978a 8 1978b)
indicated that the ARS and ARO measures were useful in
predicting group outcome. Lieberman, Yalom, and Miles (1973)
have linked group outcome to group climate. In light of
these relationships it appears reasonable to expect that
groups' mean ratings on the ARS and ARO measures will link

directly to features of their emotional climate.

l9
Hypotheses

The literature reflects considerable divergence between
theoretical and empirical works supporting the ideas that
groups evolve through stages that are discrete, overlapping,
or cyclic. Based on the works of Bennis (1964), Tuckman
(1965), Yalom (1975), and Lacoursiere (1980) the hypotheses
reflect the expectation that the initial stages of group
interaction will be characterized by higher levels of Depen-
dence, Normative behavior, Anxiety, and Reasoning, while the
middle sessions will contain higher levels of conflict or
Anger, Confrontation, and Rejection. Based on Gibbs' (1964)
assertions that acceptance, data flow, goal formation, and
control would fluctuate throughout group sessions it was
hypothesized that levels of Avoidance, Care, Participation,
Distance, and Disclosure would not vary significantly from
early, to middle, to late group sessions of group climate.
Well-known differences in how situations were perceived by
participants versus observers suggested that including both
perspectives might illuminate these issues. Due to the
availability of data related to the two principal dimensions
of interpersonal behavior an exploration of the possible
relationships of these data to group climate was conducted.
Hypotheses relating Acceptance versus Rejection of Self and
Acceptance versus Rejection of Others to group climate were
originally made because the ARS and ARO measures appeared to
assess aspects of behavior that would likely be reflected in

particular components of group climate. Because it later

20
appeared important to look at the relationship of GCQ-S
items and item clusters, the ARS and ARO measures were cor—
related with existing clusters instead of particular group
climate items specified in the following hypotheses.

Hypothesis 1: The mean ratings of the GCQ-S's Depended,

 

Normative, Anxious, and Reasoned, as separate—
ly perceived by members and observers, will be
significantly higher during early sessions
than for late sessions.

Hypothesis 2: Members' and Observers' mean GCQ-S ratings for

 

Confronted, Angry, and Rejected will be sig-
nificantly higher during middle sessions than
for either early or late sessions.

Hypothesis 3: Members' and Observers' mean GCQ-S ratings for

 

Avoided, Cared, Participated, Distant, and
Revealed elements will not vary significantly
from early to middle to late sessions.

Hypothesis 4: Groups' mean ARS ratings, representing inter-

 

personal behavior's dominance dimension, will

 

correlate positively with members' GCQ-S rat-
ings of Confronted, Disclosed, and Partic-
ipated, but negatively with Anxious, Depended,
Normative, and Avoided.

Hypothesis 5: Groups' mean ARO ratings, representing inter-

 

personal behavior's affiliative dimension,

 

will correlate positively with members' GCQ-S

Hypothesis 6:

 

21
ratings of caring, but negatively with Anger,
Rejected, and Distanced.
When matched for sessions and groups, observ-
ers' mean ratings of group climate will be

significantly less favorable than members'.

METHOD

Participants

The participants in this study were junior and senior
level undergraduates enrolled in a regular course in psy-
chology at Michigan State University (PSY 400: Small Exper-
iential Groups for Interpersonal Learning or SEGIL). In
this course students interact informally in small groups
with the aim of developing interpersonal and intrapersonal
skills through feedback, disclosure, empathy, and confronta-
tion. These groups typically contain five to seven members
and are 5010- or co-led by undergraduates with prior SEGIL
experience and who also subsequently prepared for their
leadership role by observing these groups for a term. SEGIL
groups meet for a total of about 50 hours over a 10-week
period. Two 90-minute sessions are held weekly along with
12-hour marathon sessions occurring near the term's third
and seventh weekends.

Students may subsequently enroll in PSY 400 as observ-
ers to enhance their understanding of interpersonal groups
and as possible preparation for later leading SEGIL groups.
Their responsibilities typically include directly observing
two groups per week, giving immediate postsession feedback
to group leaders, keeping a journal of their observations,
participating in their own experiential group, attending

weekly didactic and supervisory meetings, completing GCQ's,

22

23
and writing a term paper. Space and other considerations
make it undesirable to have more than two observers present
for any single SEGIL session. Selected largely because they
had received more GCQ-S reports from observers than other
groups throughout each term, a total of 11 groups (69% of
the total available groups) from the Fall, Winter, and
Spring terms of 1983 and 1984 were studied. While these
groups had accumulated a greater number of observations than
non-selected groups, no other notable observational biases
were expected. Observer attendance at group meetings was
primarily determined by the academic schedules of observers.
Selecting groups that had received greater cumulative
observer attendance also diminished the considerable im-
balance between the number of GCQ-S reports available from

members versus observers.

Measures

Group Climate Questionnaire-~8hort Form

 

As previously mentioned, MacKenzie's Group Climate
Questionnaire (GCQ) has long (32 items) and short (12 items)
forms. Because it was to be administered repeatedly at the
end of each group session, the longer version appeared
unduly burdensome in the present circumstances. The GCQ-S
is a Likert—type scale with seven response alternatives for
each item ranging from "not at all” to ”extremely.” In that
it provided a space for additional comments and one less

item ("Everything considered, I gained something of value

 

24
from today's session"), the Observers' GCQ-S form differed
slightly from that used by group members. Copies of observ-
ers' and members' GCQ-S are given in Appendices A-1 and A-2.
The GCQ-S takes less than five minutes (only about two
minutes after several prior administrations) to complete and
contained seven single or dual—item elements: ”engagement,
disclosure, support, conflict, challenge, cognition, and
control." As previously noted the elements were represented
by 12 items that are presently referred to as: Revealed,
Participated, Reasoned, Confronted, Cared, Normative, Angry,

Depended, Avoided, Anxious, Distanced, and Rejected.

Acceptance versus Rejection of Self and Acceptance versus

 

Rejection of Others

 

Hurley's ARS and ARO measures, representing the two

principal interpersonal dimensions of affiliation and dom-

 

inance (Wiggins, 1982), are each composed of four bipolar

subscales. A Liked-Disliked scale was positioned before all

 

others in an effort to diminish confounding perceptions

(Smith, 1979). The ARS subscales are: Shows Feelings-Hides

 

Feelings, Expressive-Guarded, Active-Passive, and Dominant-

 

Submissive. The ARO subscales include: Warm-Cold, Helps

 

 

Others-Harms Others, Gentle-Harsh, and Accepts Others-

 

 

 

Rejects Others.

 

GCQ data were divided into early, middle, and late

segments for analysis, with the early time segment consist-

25

ing of sessions 2 through 6, the middle segment, sessions 7
through 12, and the late segment, sessions 14 through 18.

For each set of five groups data was consistently
received from 82 members. This produced 410 reports of
group climate for each time period. There were respectively
48, 32, and 29 observations made in early, middle, and late
time periods. The observations produced 240, 160, and 145
group climate reports. For each time period the means of
members' and Observers' ratings of each group were used to
compare group climate information.

Data supporting the ability of the ARS and ARO measures
to consistently reflect interpersonal behavior style has
previously been cited (Hurley, 1978; Hurley, 1984; Small &
Hurley, 1978). The construct validity of the ARS and ARO
scales has been supported by their substantial and differen-
tial correlations with prototypical measures (Wiggins, 1982)

of affiliation and dominance. Thus, Gerstenhaber (1974)

 

 

found that LaForge and Suczek's (1955) LOV factor (affilia-
tion) correlated .55 (p < .001) with ARO, but .00 with ARS,
while ARS correlated .70 (p < .001) with their DOM factor

(dominance), which correlated minimally (.18) with ARO.

 

Additionally, the ARS and ARO measures have been found
(Hurley, 1983) to correlate significantly with relevant
features of Lorr and McNair's (1965) Interpersonal Behavior
Inventory (IBI). Administration of the IBI near groups' end
to 47 undergraduate members of six small experiential groups

who had earlier made ARS and ARO ratings after both 22— and

26
45-hours of group interaction yielded peers' mean ARS
ratings which correlated positively (.41 & .63) with the
181's five-scale Dominance factor, but negatively (-.39 &
-.44) with its Intropunitive factor. As also expected, the
ARO scale correlated strongly (.73 & .74) with the 181's
six-scale Affiliation factor, but nonsignificantly with the

181's Dominance and Intropunitive factors.

Procedure

GCQ-S, ARS, and ARO ratings are routinely collected
from all members of Psychology 400 SEGIL groups. SEGIL
observers also regularly complete and return the GCQ-S with-
in 12 hours after observing each group session. Group mem-
bers complete the GCQ-S immediately after each regular 90-
minute session. An appointed group member routinely assem-
bles and collates these data.

Including two lZ-hour marathon sessions, SEGIL groups
average about 20 meetings (range = 18 to 21) per term. To
control for the likely inflation of GCQ ratings at the close
of 12-hour sessions, marathon GCQ-S ratings are, instead,
collected from members at the start of the first post-
marathon meeting, usually after totals of about 20- and 40-
hours of group interaction. Due to their 12-hour length,
marathon sessions are not attended by observers, nor do they
normally attend either the first or final meeting of SEGIL
groups. Consequently, all GCQ data from these atypical ses-

sions were excluded from all comparative analyses. ARS and

27
ARO ratings are collected from members following the close
of each group's first postmarathon session to reduce undue
attention to behaviors occurring within the atypical mar-

athon sessions, and also as a precaution against inflated

rating.

FINDINGS

Interperiod Shifts

 

An overview of members' and Observers' separate mean
ratings of each group climate item for the early, middle,
and late periods, plus all interperiod shifts (as assessed
by E-test), is given in Figure 1 and more fully reported in
Appendices B-1 and B-2. Confidence in these member-based
ratings exceeds that of the observer-based data because the
period means of members typically represented 410 (82
persons 3 5 sessions) ratings versus an average of 182 ([240
+ 160 + 1451/3) ratings that undergirded the parallel
observer means. Members and observers generally agreed on
the relative intensity (”not at all” to 'extremely") of
these ratings on each occasion. Thus, both sources rated
behaviors associated with five items (Cared, Reasoned,
Participated, Confronted, and Revealed) as much more common
than those reflected by the seven trailing items (Normative,
Anxious, Angry, Avoided, Distanced, Depended, and Rejected).
The product-moment correlation between members' and
Observers' all-period mean ratings of these 12 items was
high (g = .85, p < .01). Beyond this broad intersource
agreement, however, members differentiated between behaviors
associated with these two subsets of items much more sharply

(towards either "extremely” or ”not at all”) than did

28

29
Figure 1

Group Climate Hean Shifts of Members and Observers
for EarlyL Middle, and Late Time Periods

 

 

 

 

 

 

 

 

SOURCE OF RATINGS P E R l 0 0 S
MEMBERS OBSERVERS Early Middle Late
16.0: EXTREMELY
\r
495.0. A GREAT DEAL
Reasoned T
K
Careda'c a c f
Participated ’ --_i/.o: QUITE A BIT
Confronted 9
Revealed Participated .r~ ' a t
Confronted ‘ ‘ ~ ‘ . I ’ j
Cared My ‘ ~ .. , ' , - .. ~65
Reasoned ‘~“:~- ‘ ' ,2 .I' ' ,‘ )-
“\’ ' ‘ ~ ‘. \‘\‘ I ’ I I i
' ‘ s‘ ‘ I I
Normative ' \s s I . .‘ ‘ I ’ .
a {“""." I I T~
Depended s I ‘ v ’ I \ 4.3.01 HODERATELY
Anxious : ' ‘~-.3 I _ _ "\ \ :
Revealed ‘ ‘ , ‘ s ‘ ~ ’ \ ‘
s ‘ ‘ l‘ ‘ ‘ \
I § ‘ > ‘ ‘ ~
I \ s ‘ ‘ 3
z ’z' ' I " ' ' R
Avoided "
Normativea'b
Anxious
Angry SOHEHHAT
b Distanced
Depended.’
Angry b
Avoided
Distanced Rejected
Re'ectedc P
' «~10: A LITTLE BIT
D"

 

1.0: NOT AT ALL

aShifted significantly from early to late.

I"

C

early to middle.
” middle in late.

30
observers. Observers' mean ratings were typically (in 31 of
36 instances) nearer this seven-point scale's midpoint
("moderately”), between the upper quintet and lower septet
of members' ratings. Another notable difference was that
members' interperiod shifts attained statistical signif-
icance (p < .05, 2-tailed) much more often (11 vs. 1) than
did Observers'.

An overall MANOVA contrasted members' and Observers'
ratings of the early, middle, and late phases. The depen-
dent variables were each group's cumulative mean rating of
each item for each period. Fully reported in Appendix C,
MANOVA revealed an expected main effect indicating that
members generally rated the climate of these groups more
favorably (p < .04) than did observers. No significant
interactions between members' and Observers' ratings were
found. All univariate contrasts between these ratings of
members and observers were statistically significant, except
for Participated. Due to item phrasing, a more favorable
view was indicated by higher numeric scores on all five

items of the upper quintet (labeled Affiliative), but by

 

lower scores on the seven (Nonaffiliative) items. Examina-

 

tion of Figure l and Appendices B-1 and B-2 reveals that

members rated each item more favorably than did observers.
Three separate MANOVA's were used to assess the hypoth-

eses concerning interperiod shifts (upward, downward, and

unchanged) in group climate items. The results of these

31
MANOVA's appear in Appendix D. No significant interactions
between period (early, middle, late) and source of rating
(member vs. observer) were found. As predicted, however, the
MANOVA for items Depended, Normative, Anxious, and Reasoned
was significant from early to late sessions (F = 6.32, p <
.003). An examination of the mean group climate ratings
(Figure 1) illustrates that scores on these items decreased
over time. A large amount of the variance in this four-item
multivariate series was attributable to Depended (F = 26.95,
p < .000). Contrary to hypotheses, Confrontation, Anger, and
Rejection were not significantly higher during middle than
in early and late sessions. As predicted, mean ratings on
Avoidance, Cared, Participated, Distanced, and Revealed did
not shift significantly from early to middle to late
sessions.

Overshadowing these MANOVA outcomes were the marked
discrepancies between members' and observers' ratings of
group climate, and the smaller observer sample size, both
discussed earlier. Additionally the greater variability of
Observers' ratings must be noted when considering MANOVA
results. The standard deviations of Observers' mean ratings
of individual items (median SD = 1.00) was often twice or
more that of members (median SD = .445). This greater var-
iance probably reflects Observers' less consistent atten-
dance at sessions and their lesser emotional involvement in

group interaction.

32

T-tests, summarized in Figure l, and Appendices B-1 and
B-2, were used to clarify these MANOVA's. From the perspec-
tive of members, Depended (on leader) and Normative Behavior
declined significantly from both early to middle (p's <.01)
and early to late (p's<.001) while Anxious declined from
early to late (p<.04). Reasoned did not change significant-
ly. While Anger, Confronted, and Rejected were not signif-
icantly higher during middle sessions, Rejected declined
significantly from middle to late sessions (p<.03). Of the
five items expected to remain stable, only Revealed and
Distanced did, while Avoided declined significantly from
middle to late sessions (p<.02) but Cared and Participated
increased from both early to middle (p's<.03 and .01) and
from middle to late (p's<.05) sessions.

The only significant interperiod shift in Observers'
ratings was Depended's early to late decline. The sparcity
of significant changes in Observers' data seems generally

attributable to their notably greater variance.

Correlational and Cluster Analyses

The changes in group climate reflected by the p-tests
were further clarified through correlations of all group
climate items. Separately for the data of members' and
observers', mean ratings of the 12 group climate items were
correlated for each time period and item clusters were iden-

tified. Based upon McQuitty's (1961) elementary factor

33
analysis, an illustrative cluster analysis of later member-
based mean group climate ratings is shown in Table 1. Two
clusters of positively intracorrelated, but negatively
intercorrelated, items are evident. These consist of a

quintet, labeled Affiliative to denote their common theme,

 

and a septet-—labeled Nonaffiliative--to recognize the

 

remaining seven items' content communality as well as their

bipolarity to the Affiliative quintet. Also shown in Table l

 

is each item's contribution to the total covariance (£2) of
each cluster and also to the total matrix. All statistically
significant linkages among these items are diagrammed in
Figure 2's lower—right corner with solid lines denoting
positive correlations and broken lines denoting negative
correlations. The breadth of these lines shows each connec-
tion's relative strength (£2).

Figure 2's five remaining diagrams ilustrate the out-
come of parallel intercorrelational matrices (fully given in
Appendices E-l and E-2) for item clusters of members' mean
ratings of group climate at early and midgroup and also the
Observers' three comparable data sets. The pattern of link-
ages was similar for each source, although members' data
generated nearly twice the total number of significant link—
ages (95 vs. 53) as did Observers' data. OflIMSintercluster
linkages provided by members' ratings, merely 3 were
positive and statistically significant versus a total of 28

negative and significant linkages. The Observers' inter-

34

Hv.¢

wm.~

co>o~m mo mmcwumm oomomluonsoz qu mEouH mumeHU NH OCOE¢ unawumHouuoououcH

mm.m

mm.H

MH.~

mm.

v¢.m

on.H

mm.H

hh.m

os.a

.A non ooH\H so

mm.H

mv.H

no.H
mar
NH

mm!
mVI

mmm
as

vo.~

mm.

mm.m

mm.H
mvl
owl
le
hhi
Hm!
NNI

mN.H

 

haafluHSE «omppfleo mHmEfluwo Had .wuoz
.ummu emanmuum any an mo. v m
.ummu emanauum man an as. v mo
.ummu cannabum and an Hoe. v mm
mm.m mm.m mHmLOB sauna:
HG.H ms.H Nu m>nnmAHnmmmcoz
use: new: embomflmm .a
omen ohm: eoocmumna .e
0mm- ommn msoAxca .e
Hm: mm: cmono>c .H
Has om- emocmoma .m
0mm: 0mm- moose .o
mm- ma: m>aumsuoz .m
m~.~ s~.~ mm.H mu m>anmaanma<
amm non omnmo .m
ma mm emuconmcoo .0
amp oom easemmmm .o
. mmm emnmoaonnumm .m
amm emamm>mm .a
m a

moonmom msouw mung mcfluso mmsouw

.H wanna

.owuuowox "A w mococmuwfl: “x
“mzcflxc< no accoMC>< "H «poocooo: u: “>um:< "O xo>HumEucz "m “couco.um “toucazcnwcoc "c
“oozcwcoz no “ounceflufiuacm "a “poaco>oz n< .Sumzcunm c>flucaon mcacupocfl :uoflz

ocfla mm:cflucacuaco o>wucacz >wfiucwofi mocfla :oxeun “mote: ¢>fiefimcm 3:;m wccfifl c_Hcm

.| ill'l"

“WZEMZfi;

 
 
 

. II...
-FnfinflufllllllllllIL""-..

   

l l Illllz ll ‘Wflunnuhlflhfll’

\ u

    

 

nmzm>zzmc

 

 

wooduom o;L:E um MEouH comflwao macho NH mo wommxcflq HmcoflunamMLOULoucH

m wusmflm

36

cluster bonds were similar, yielding a total of five (of 53)
significant positive intercluster linkages, all for early
data, but 23 significant negative associations. Members'
late ratings, when familiarity with the GCQ-s instrument and
the situation were greatest and comfort was likely highest,
provided the clearest perspective on the structure of these
Figure 2 intercorrelations. Especially clear for members'
ratings were the significant positive connections among the

five Affiliative items (A, B, C, D, a E) and also among the

 

seven Nonaffiliative items (F, G, H, I, J, K, & L), while

 

all significant intercluster bonds were negative during
middle and late sessions in data from both members and

observers. Earlier, however, Affiliative's Confronted (D)

 

linked positively to Nonaffiliative's Anger (G) as rated by

 

both members and observers in addition to Anger's signif-
icant early bonds to Revealed (A) and Participated (B) for
members and to Reasoned (C) for observers.

As shown by their overall contributions to the total
covariance of members' ratings in Appendix E-l, items
Revealed and Participated were most central to the cluster,
closely followed by Reasoned, Confronted, and Cared. Sim-

ilarly, the Nonaffiliative cluster was anchored by Rejected,

 

Distanced, and Anxious, trailed respectively, by Avoided,
Depended, Angry, and Normative. For members, Cared's 21
statistically significant negative linkages to the Nonaffil—

iative septet's items (see Appendix E—l) best bridged these

37

subclusters. Rejected, the core Nonaffiliative item, was a

 

strong secondary bridge with its seven significant negative

correlations to members of the Affiliative quintet.

 

The differentiation of these 12 group climate items

into Affiliative and Nonaffiliative clusters fully paral-

 

 

1eled the relative intensity with which these items had been
endorsed to characterize the present groups in Figure 1. As

shown there, Affiliative behaviors predominated over Nonaf-

 

filiative behaviors within these groups at each period and

 

Affiliative behaviors also generally increased from early to

late, while Nonaffiliative behaviors generally declined.

 

Distinct from the statistical significance of group
climate item shifts, a noteworthy feature of these findings,
clearer in Observers' than in members' ratings, was the
general "V" or EARLY-middle-LATE (Egg) pattern of inter-
period means that characterized ratings of items of the

Affiliative quintet in Figure 1. Holding for all five items

 

in Observers' ratings and for three (Cared, Reasoned, and
Participated) items in members' ratings, it contrasted
sharply with the obverse ”A" or early-MIDDLE-late (eMl) pat-

tern that prevailed for mean ratings of the Nonaffiliative

 

septet. Also clearest in Observers' data (Rejected, Dis-
tanced, Avoided, and Normative), the eMl pattern was sus-
tained by members' ratings on Rejected, Distanced, Avoided,

and Angry. The Nonaffiliative Depended and Anxious ratings

 

departed from eMl by steadily declining according to both

38

sources. Contrarily, members' Affiliative Confronted and

 

Revealed ratings showed a modest but sustained rise, while
members' Normative ratings dropped consistently.
The negative correlations that characterized the link-

ages of items across the Affiliative and Nonaffiliative

 

 

clusters appear fully complementary to these opposite EmL
and eMl patterns. Overall, all items moved towards increased

Affiliativeness and toward decreased Nonaffiliativeness,

 

 

aside from the observed-based minor rises on Avoided and

Angry and trivial decline on Confronted. Nonaffiliativeness

 

apparently peaked in the middle phase.

Linkage to the Central Dimensions of Interpersonal Behavior
Because data for these groups were available on
measures related to interpersonal behavior's prepotent

Affiliation and Dominance dimensions, represented by self

 

 

and peers' ratings on Acceptance versus Rejection of Others
(ARO) and Self (ARS), respectively, an exploration was also
made of group climate's possible linkages to these measures
of broader systematic importance. Although it had been
hypothesized that the ARS and ARO measures would link both
positively and negatively to particular group climate items,
those hypotheses appeared inappropriate in light of sub-
sequent clustering among the GCQ-S items. In view of the

bipolar nature of the Affiliative and Nonaffiliative sub-

 

 

clusters, it seemed reasonable to expect them to link

39
oppositely to members' mean ratings of these groups on the

Affiliation (ARO) interpersonal measure, positively with the

 

Affiliative cluster and negatively to the Nonaffiliative

 

 

one. Given ARO and ARS's theoretical independence, there
seemed no compelling grounds for anticipating that ARS would
associate substantially with either cluster. A comprehen-
sive overview of these correlations for both self- and
peer-based interpersonal ratings after both about 20- and
40-hours of group interaction is given in Table 2. Logic
suggested, however, that the later interpersonal and group
climate data would be more stable and valid due to its much
firmer experiential base.

ARO consistently correlated positively with the Affil-
iative subcluster, as anticipated, although merely one of
these 12 correlations reached statistical significance. Late
ARO always (6 of 6 instances) correlated more positively

with the Affiliative than with the Nonaffiliative sub-

 

 

cluster. Substantial covariance (21%) was represented by
the median .455 correlation between late ARO and the Affil-
iative subcluster, although the very few degrees (9) of
freedom required these correlations be large to achieve
statistical significance. Because they linked data sets

separated by several weeks, these ARO-Affiliative bonds

 

appear robust, as group climate ratings were made at near

each group session's close, while the ARO and ARS ratings

were based upon extended periods of group interaction.

4O

 

_ m~u
m :N
MN- N
om m:
mu mm
+om m:
mused u_om
monoxio:

 

 

.am__mu-oza .o_.v.a+

6212-9: Rove...

 

.t cot oo_\_ >5 >3:35 lamaa_eo m.ms_umn __< .aaoz

 

 

o. .m
m. m.
an ._a
mm mm
o. .8
m. o.
mtooa u.om
mcaozuom

: m
m. m_-
mm: :N
.3 3
NM om
.sm .mm
mcooa u_om

mtaozuo:

 

KN mN
~_u N_u
w: a:
m .1
m_ KN
~_s ON:
mcood u_om
mcaoxno~

 

 

om<

m¢<

can;
o_eo_:
>_tmm

downtowmq

one;

o_nu_z

>_tmu

 

maaaeeawhwq

 

moobcoa match on many. mowerokmm ucm

momucmeHNw mo undone—unam one :0 mc_umm t_o;h zu.3 mecca we can m_om xn

 

mtaoxuo: new row unon< toue< mu_c: quota mo mmc_un¢ oz< new m¢< emu: .mconeoz mo mco_um_oct0u acoeozuuuaoOLm

. ~o_nmh

41

Of Table 2's total of eight negative correlations, six

concerned the Nonaffiliative cluster and four of these

 

involved ARS, although npne approached statistical signif-
icance. Thus, ARS or self-acceptance was less clearly asso-
ciated with group climate in these data, although late ARS
ratings by both self and peers closely approached sig-

nificance with Affiliative climate in several instances.

 

The overall pattern of these Table 2 associations suggests
that acceptance of both self and others tends to correlate

with a favorable (Affiliative minus Nonaffiliative) group

 

 

climate.

DISCUSSION

This study explored developmental patterns of small
groups as separately perceived by group members and non-
participant observers. It had been hypothesized that some
features of group climate (Anxious, Defended, Reasoned, and
Normative) would peak in groups' early sessions, while
others (Confronted, Anger, and Rejection) would be highest
during their middle phase, and still others (Cared, Avoided,
Participated, Distanced, and Revealed) would remain stable.
Another expectation was that members' ratings of group
climate would generally be more favorable than those of
observers, although broadly similar perceptions of group
climate were expected from each source. Lastly, although
specific group climate items had been hypothesized to cor-
relate with the two central dimensions of interpersonal
behavior, as measured by ARS (Acceptance versus Rejection of
Self) and ARO (Acceptance versus Rejection of Others), these
hypotheses were rendered dysfunctional by complex inter-
linkages among the 12 group climate items employed here.
These items interlinked in a bipolar structure featuring

clusters that seemed appropriately labeled Affiliative

 

(Cared, Reasoned, Participated, Confronted, and Revealed)

and Nonaffiliative (Normative, Rejected, Depended, Anxious,

 

Angry, Avoided, and Distanced). These Affiliative and Non-

 

42

43

affiliative clusters were subsequently correlated with ARS

 

and ARO.

The Affiliative and Nonaffiliative clusters became

 

 

increasingly distinct from groups' early to later phases.
Positive intracluster bonds became stronger near groups' end
than during earlier sessions and the negative intercluster
bonds also increased in number and strength over time.

Additionally, Affiliative item ratings generally increased

 

over time, while Nonaffiliative items tended to decline.

 

Members' mean ratings shifted significantly (p < .05) on
seven items (Cared, Participated, Normative, Anxious,
Depended, Avoided, and Rejected), always toward increased

Affiliativeness or decreased Nonaffiliativeness, while

 

 

Observers' ratings shifted significantly only on Depended
(downward).

The MANOVA's yielded a mixture of support and discon-
firmation of the hypothesized shifts in group climate. Cumu-
latively, items Depended, Normative Behavior, Anxiety, and
Reasoned declined as expected while items Rejected, Confron-
ted, and Anger did not show the hypothesized midgroup rise.
The set of Participated, Avoided, Distanced, Revealed, and
Cared items did not shift significantly over time. Regarding
only members' ratings for individual group climate items,
Depended, Normative Behavior, and Anxious declined while
Reasoned remained stable. Although Anger and Rejected failed
to peak as hypothesized at midgroup, Rejected declined sig-

nificantly from middle to late sessions. Of five items

44

expected to remain stable, only Revealed and Distanced did,
while Avoided declined significantly from early to late ses—
sions, but Cared and Participated generally climbed from
early to late sessions. Anger, Confronted, Revealed, and
Distanced all remained essentially stable over time.

The meaning of individual items' shifts was over—
shadowed by the broader trends reflected in the clusters'

complimentary “V” (Affiliative) and ”A" (Nonaffiliative)

 

 

patterns. These highlighted a general move toward increased
constructiveness that included a temporarily more turbulent

middle phase, evidenced by increased Nonaffiliativeness and

 

decreased Affiliativeness, within a broader movement toward

 

either increased Affiliation or decreased Nonaffiliativeness

 

 

from early to late sessions reported by members on all 12
items, and by observers on 9 of these.

The midgroup peak in Nonaffiliativeness may have

 

occurred because members felt more comfortable experimenting
with potentially threatening expressive behaviors following
their safe passage through the anxiety-laden beginning

phase and before they were confronted with the desire of
ending on a pleasanter note. An independent set of data
supported these broad patterns. MacKenzie (1983) admin-
istered the GCQ-S to 75 outpatient members of 12 psycho-
therapy groups, yielding factors that he labeled Engagement,
Avoiding, and Conflict. His Engaged factor items are iden—

tical to the present Affiliative quintet. The present Non—

 

affiliative septet included MacKenzie's four Avoiding

 

45
(Distanced, Normative, Avoided, and Depended) items plus his
pair of Conflict (Anger and Rejected) items. The Anxious
item did not load differentially on MacKenzie's factors.
Although MacKenzie's groups met for 35 sessions, for
unexplained reasons his analysis was based on only the
initial 14 sessions and 19% of his participants supplied no
usable data. According to his group factors, Engagement
dropped sharply during sessions 7 and 8, then increased
steadily through session 11, and was generally higher during
later (numbers 12-14) sessions. Complementing this, Avoiding
increased during sessions 7 and 8, decreased through session
11, and was slightly lower in sessions 12, 13, and 14. These
two patterns closely paralleled the present study's Affil-

iative and Nonaffiliative clusters. Also congruent was his

 

Conflict duo's midgroup peak at session 10.

The subclustering of group climate items found in the
present study and in MacKenzie's work also appears related
to Alden's (1978) analysis of Moos' Ward Atmosphere Scale.
Alden found that over half of the variance among Moos's
subscales was attributable to a single evaluative (good vs.
bad) factor representing general attitude toward the ward.

This broadly resembles the major axis of the Affiliative

 

versus Nonaffiliative clusters, suggesting that a relatively

 

simple factorial structure may hold among diverse measures
of atmosphere and/or climate.
In addition to the GCQ-S, members (but not observers)

of the present groups responded to a 13th item, ”Everything

46
considered, I gained something of value from today's ses-
sion“ on their postsession reports. For a series of 28 SEGIL
groups that included eight of the present series of 11,
group members' mean aggregated all-session ratings of Item
13 correlated strongly (p = .73, p < .01) with their mean

aggregated composite Affiliative-Nonaffiliative ratings of

 

group climate. Depended and Normative were omitted from this
revised composite because of their generally weaker rela-

tionships to other Nonaffiliative items. Composite group

 

climate ratings and Item 13 ratings were also plotted over
all group sessions. These two curves were nearly parallel,
as shown in Figure 3. Each peaked near groups' fifth, four-
teenth, and final sessions, while reaching low points for
groups' initial and middle sessions. This parallelism con-
firmed the existence of identifiable shifts in group atmos-
phere that were marked by midterm conflicts and later reso-
lution. Although not shown here exclusion of the unusually
high ratings associated with SEGIL marathon and final ses-
sions did not notably alter the parallelism of Figure 3's
curves.

This general rise of Affiliation, accompanied by its

 

present midphase turbulance and temporary decline, broadly
support the group development theories of Bennis (1964),
Tuckman (1965), Yalom (1975), and Lacoursiere (1980). These
authors characterized the early phases of group life as
featuring dependent and orienting behaviors and later seg-

ments with rising consensual validation, cohesiveness, reso-

47

Figure 3. Mean Rating of Each Session of 28 Groups by All Members (N's of 1&2
to lSl except for sessions l9 [ll3], 20 [8i], and 2i [34]).

 

Extremely:

“Everything considered, I gained something of value ’
I

from today's session.” _§ ,

A great deal.

I

 

quite a bit:

K\\‘-Composite of ten group climate items [5 Affiliative

minus 5 Disaffiliative].

I

Mbdérately:

Somewhat:

A little bit:

V

 

Not at all:

 

 

48
lution, and productivity. The midgroup decline appears to
reflect phenomena designated as resolution—catharsis, con-
flict, and dissatisfaction, although this phase is usually
completed by midgroup in the theories of Tuckman and
Lacoursiere. The general shifts of the present group climate
clusters also broadly support stage-oriented theories of
group development. The relatively stable ratings of the
Distanced, Anger, Reasoned, Confronted, and Revealed items,
according to both members and observers, also seems consis-
tent with Gibb's concepts (1964) about the continual fluctu-
ations of levels of trust and of openness.

Similar to the rising Affiliativeness evidenced by mean

 

ratings shifts, a trend toward increased interpersonal con-
structiveness emerged in the intercorrelations of group
climate items. The primary example of this was Anger's

shifting linkages to components of Affiliative and Nonaffil-

 

iative aspects of group climate. Early, members positively
linked Anger with Participation, Confronted, and Revealed,
while observers similarly linked Anger only to Confronted
and Revealed. During the middle sessions, however, both
members and observers linked Anger inversely to Cared.
Members also positively linked Anger to four other Nonaffil-
iative items, whereas observers linked Anger positively only
to Anxiety. Later, the negative correlation between Anger
and Cared was sustained by observers and supplemented by
Anger's strong bond to Depended, while members negatively

correlated Anger with Revealed and Participated (contrary to

49
their early linkages) and also positively associated Anger
with Anxiety, Rejected, and Distanced. These shifting
patterns suggest that group members and observers learned
that confrontation and disclosure need not necessarily be
associated with defensiveness.

While members and observers evidenced generally similar
correlational patterns and group climate shifts, as shown by
the EmL and eMl patterns, members' ratings of group climate
were much more favorable than were those of observers.

Members' more sharp differentiation between Affiliative and

 

Nonaffiliative ratings and the Observers' significantly less

 

favorable ratings may be attributed to several factors.
Jones and Nisbett (1971) have said that actors, or individ-
uals involved in interactions or tasks, more often attribute
events to circumstancial versus personal characteristics.
Contrarily, nonparticipant observers tend to attribute
events more to actor's personal traits than to circum—
stances.

Jones and Nisbett (1971) partially attributed actor-
observer differences to the Observers' relatively lesser of
historical awareness of contextual events, the greater ego
investment of actors (akin to present group members), and to
the differential visual perspectives by actors and
observers. Having little or no information about the
personal history of actors in groups' prior sessions,
observers seem more likely to attribute behaviors to

individuals' predispositions, or character traits, rather

50
than to circumstances. The present divergence in member-
observer ratings may also be partly due to the Observers'
much more fragmented contact with these groups.

On the other hand, differences may occur merely because
observers primarily attend to actors, while actors focus on
visual stimuli outside themselves. This phenomenon might be
clarified by a study that controlled for historical aware-
ness by systematically varying the number of sessions that
observers attended and comparing the ratings of observers
who had greater and lesser degrees of exposure to particular
groups. To further delineate observer biases, groups of
experienced (previous group members) and inexperienced
observers could be compared. The present author would expect
historical awareness to account for some differences in
perception.

The Observers' less favorable ratings indirectly sup—
ported the findings of Cunningham et al. (1979) that passive
observers tend to attribute negative interactions more to
actors (group members) than do the actors themselves. Group
members may similarly inflate the ratings of their group to
enhance their own self—esteem. Observers' less favorable
ratings may reflect a lesser need to perceive interactions
positively, associated with their relative detachment.
Ratings indicating less complimentary perceptions may be
ego—threatening to members. For example, the increased
Anger, decreased Confrontation, and relatively unchanged

Distance and Avoidance ratings of observers evident in these

51
groups' later periods may have emerged because they were
removed from the more vulnerable feelings associated with
groups' endings. The Observers' generally less favorable
present ratings were congruent with the report of Lieberman
et a1.'s (1973) that nonparticipant observers rated group
leaders less favorably than did members.

While members rated the present groups much more
positively than did observers, both sources described them
as distinctly more constructive than neutral and as also
becoming increasingly constructive over time. These trends
are reflected in group climate cluster relationships to the
central interpersonal dimensions represented by the ARS and
ARO scales. Each of the latter scales generally linked

positively to the Affiliative cluster of group climate

 

items, but less consistently and positively and more in-

consistently with the Nonaffiliative cluster. Different

 

interpersonal postures were apparently represented by these

Affiliative and Nonaffiliative poles of group climate. The

 

 

Affiliative quintet was composed of items that blended

 

capacities for both intimacy (Cared, Revealed, and Partic-
ipated) and differentiation (Reasoned, Confronted). The cor-

relation of Confrontation with other Affiliative elements

 

indicates that intimacy can be effectively integrated with

individuation. The Nonaffiliative septet represents an

 

interpersonal orientation characterized by insecurity,
detachment, and the evasion of interdependence. Because

Cared and Rejected contributed the largest number of neg-

52
ative intercluster correlations, it is reasonable to assume
that the associated behaviors are reasonably fundamental
features of group culture.
Studying data from 107 human societies, Rohner (1975)
concluded that acceptance (versus rejection) of children by
parents was crucial to personality development. Similarly,

it is understandable that the Affiliative cluster would

 

generally correlate positively, and the Nonaffiliative

 

cluster less positively, with ARS and ARO because self-
acceptance is a likely introject from nurturant inter-
personal environments, while the acceptance of others seems
the likely projection of such a positive self-evaluation.

In the context of marked differences in the administra-
tion times, and feedback provided by the group climate and
behavior ratings, the robustness of their linkages seems
notable. Although each instrument provided feedback to group
members, this was individualized and discussed in depth for
the behavior ratings, but group-oriented and only briefly
viewed for the GCQ-S. Members may also have responded to
defend themselves or their group from unfavorable and ego-
alien perceptions on either measure. Because they reported
how each person was rated by self and all others, the ARS
and ARO measures are likely more anxiety-provoking than the
GCQ-S. Additionally, the behavior ratings were gathered and
presented to the group only twice, while the group climate
ratings offered general group feedback on each session. More

statistically significant linkages between group climate and

53
behavior ratings occurred late than early. This was a likely
consequence of the much broader experiential base (43 versus
20 hours of group interaction) that undergirded this second
series of behavior ratings.

The tendency of these interpersonal groups to be exper-
ienced positively may not be readily generalizable to all
other small groups. Their constructiveness may have been
facilitated by an undergirding theme that emphasized respect
for the individual. Group development was augmented by the
direction of trained leaders, readings, didactic instruc-
tions, and personal reflections recorded in a journal.
Apparently these combined procedures contributed to the
positive perceptions of most participants and some com-
bination of them may be integral to constructive group
development.

In summary, several factors indicated that members of
these groups generally moved toward a greater sense of
security and freedom. The increasingly strong endorsements

of Affiliative behavior, their overall evolution toward

 

Affiliation and away from Nonaffiliativeness, and the

 

 

progressive differentiation of the related item clusters,
all supported the interpersonal constructiveness of these
groups. Similar shifts were independently reported by
MacKenzie's account of changes during psychotherapy groups,
the positive correlation between group climate and Item 13
("Everything considered, I gained something of value from

today's session"), and the general positive correlation of

54
members' ratings of their groups on ARS and ARO with the
composite Affiliativeness climate measure. The shift toward
increased Affiliativeness, and temporary midgroup elevation
of Nonaffiliativeness, accompanied by the relative stability
of group climate elements (Anger, Revealed, Distanced, and
Confronted) associated with the two poles of group atmos-
phere, supports an integration of stage and cyclic theories
of group development. The Observers' consistently less
favorable perceptions of group climate may be attributable
to their more fragmentary awareness of the course of each
group, as well as by group members' greater investment in
self—esteem enhancing ratings, and the reception of differ—
ential visual stimulation. A study that controlled for
historical awareness among observers might further clarify

these issues.

REFERENCES

Adams, H. B. (1964). “Mental Illness” or interpersonal
behavior? American Psychologist, 11, 191-197.

Alden, L. (1978). Factor analysis of the ward atmosphere
scale. Journal of Consultingiand Clinical Psychology, 4g,
175-176.

Bales, R. F., & Stodtbeck, F. L. (1951). Phases in group
problem solving. Journal of Abnormal Social Psychology,
fig, 485-495.

Benjamin, L. S. (1974). Structural analysis of social
behavior. Psychological Review, 81, 392-425.

Bennis, W. G., & Sheppard, H. A. (1956). A theory of group
development. Human Relations, 9, 415-438.

Bennis, W. R. (1964). Patterns and vicissitudes in T-group
deve10pment. In L. P. Bradford, J. R. Gibbs, & K. D. Benne
(Eds.), T-group theory a laboratory method (pp. 248-278).
New York: Wiley 8 Sons.

Berne, E. (1966). Principles of group treatment. New York:
Grove Press.

Bion, W. R. (1959). Experience in groups. New York: Random
House.

Carson, R. C. (1969). Interaction concepts ofypersonality.
Chicago: Aldine.

Cissna, K. N. (1984). Phases in group development: the neg-
ative evidence. Small Group Behavior, 15, 3-32.

Cooley, C. H. (1909). Social organization: a study of the
larger mind. New York: Scribner's.

 

Dunphy, D. C. (1972). The primary group: a handbook for
analysis and field research. New York: Appleton-Century-
Crofts.

Durkheim, E. (1894). Suicide. Glencoe, IL: Free Press, 1951.
English, H. B., & English, A. C. (1958). A Comprehensive

Dictionary of Psychological and Psychoanalytic Terms. New
York: Longmans Green.

55

56

Foa, U. G. (1961). Convergences in the analysis of the
structure of interpersonal behavior. Psychological Review,
§§. 341-353.

Freedman, M. B., Leary, T. F., Ossorio, A. G., & Coffey, H.
S. (1951). Interpersonal dimensions of personality.
Journal of Personality, 2Q, 143-161.

Freud, S. (1919). Group psychology and the analysis of the
eg . London: Hogarth.

Gerstenhaber, L. M. (1975). Acceptance versus rejection of
others and self in personality self-reports (Doctoral dis-
sertation, Michigan State University, 1974). Dissertation
Abstracts International, 36, 44lB-442B.

 

Gibb, J. R. (1964). Climate of trust formation. In L. P.
Bradford, J. R. Gibb, & K. D. Benne (Eds.), T-group theory
a laboratory method (pp. 279-309). New York: Wiley & Sons.

Hurley, J. R. (1976). Two prepotent interpersonal dimensions
and the effects of trainers on T-groups. Small Group
Behavior, 1, 77-98.

Hurley, J. R. (1980). Two interpersonal dimensions relevant
to group and family therapy. In L. R. Wolberg & M. L.
Aronson (Eds.), Group and family therapy 1980. New York:
Brunner-Mazel.

Hurley, J. R. (1984). Leaders' interpersonal behavior and
group members' changes. Manuscript submitted for publica-
tion.

Hurley, J. R. (1983). [Correlations among measures of the
principal dimensions of interpersonal behavior.) Unpub-
lished raw data.

Hurley, J. R., a Rosenthal, M. (1978a). Perceptions within
AGPA's annual institute groups. GROUP, 2, 220-238.

Hurley, J. R., & Rosenthal, M. (1978b). Interpersonal rating
shifts during and after AGPA's institute groups. Inter-
national Journal of Group Psychotherapy, 28, 115-121.

 

Kernberg, O. F. (1984). The couch at sea: psychoanalytic
studies of group and organizational leadership. Inter-
national Journal of Group Psychotherapy, 34, 5-23.

Kiesler, D. L. (1983). The 1982 interpersonal circle: A tax-
onomy for complementarity in human transaction. Psycholog-
ical Review, 90, 185-214.

 

Lacoursiere, R. B. (1980). The life cycle of groups. New
York: Human Sciences Press.

57

LaForge, R., & Suczek, R. F. (1955). The interpersonal
checklist. Journal of Personality, 24, 94-112.

 

Lakin, M., & Carson, R. C. (1964). Participant perception of
group process in sensitivity training. International
Journal of Group Psychotherapy, 24, 116-122.

 

Leary, T. (1957). Interpersonal diagnosis of personality: A
functional theopy and methodology forypersonality evalua-
tion. New York: Ronald Press, 1957.

Lieberman, M. A., Yalom, I. D., & Miles, J. B. (1973).
Encounter groups: First facts. New York: Basic Books, Inc.

Lorr, M., & McNair, D. M. (1965). Expansion of the inter-
personal behavior circle. J. Personal. Social Psychol., 2,

 

Lubin, B., & Zuckerman, M. (1967). Affective and perceptual
cognitive patterns in sensitivity training groups.
Psychological Reports, 22, 365-376.

 

MacKenzie, M. R. (1978). Measurements of group climate
preliminary experiment with new instrument. Paper
presented at the Annual Conference of the Society for
Psychotherapy Research, Toronto, Canada.

MacKenzie, M. R. (1979). Group norms: Importance and measure-
ment. International Journal of Group Psychotherapy, 22,
471-480.

 

MacKenzie, M. R. (1983). Clinical applications of group
psychotherapy. In R. R. Dies & M. R. MacKenzie (Eds.),
Advances in group psychotherapy (pp. 159-170). New York:
International University Press.

 

McQuitty, L. (1961). Elementary factor analysis. Psycholog-
ical Reports, 2, 71-78.

 

 

Moos, R. H. (1984). Evaluating treatment environments. New
York: Wiley & Sons.

 

Nisbett, R. E., & Jones, E. E. (1971). The actor and the
observer: Divergentyperceptions of the causes of behavior.
New York: General Learning Press.

 

 

Philip, H., & Dunphy, D. (1959). Developmental trends in
small groups. Sociometry, 22, 162-174.

 

Psathas, G. (1960). Phase movement and equilibrium tenden-
cies in interaction process in psychotherapy groups.
Sociometry, 22, 177-194.

 

Ridgeway, C. (1983). The dynamics of small groups. New York:
St. Martin Press.

 

58

Rioch, M. J. (1970). The work of Wilfred Bion on groups.
Psychiatry, 22, 56-66.

Rohner, R. P. (1975). They love meL love me not. New York:
Hraft Press.

Small, D. H., & Hurley, J. R. (1978). Work effectiveness and
personal attributes of mental health paraprofessionals. In
K. M. Nash, N. Lifton, & S. E. Smith (Eds.), The parapro-
fessionals: Selected readings (pp. 286-301). New Haven:
Advocate Press.

Schaefer, E. S. (1961). Converging conceptual models for
maternal behavior and child behavior. In J. C. Glidewell
(Ed.), Parental attitudes and child behavior (pp. 124-
146). New York: C. Thomas.

Schutz, W. C. (1955). What makes groups productive? Human
Relations, 4, 429-465.

Smith, J. A. (1979). Use of the Liked-Disliked scale in
interpersonal ratings. Unpublished Master's Thesis, Mich-
igan State University.

Stock, 0., & Thelen, H. A. (1958). Emotional dynamics of
group culture. New York: National Training Laboratories.

Symonds, P. N. (1939). Theypsychology of parent-child rela-
tions. New York: Appleton Century.

Tavistock Institute (1956). Table describing group develop-
ment. Human Relations, 9.

Tuckman, B. (1965). Developmental sequences in small groups.
Psychological Bulletin, 4;, 384-399.

Wiggins, J. S. (1982). Circumplex models of interpersonal
behavior in clinical psychology. In P. C. Kendal & J. N.
Butcher (Eds.)., Handbook of research methods in clinical
psychology (pp. 183-221). New York: Wiley.

Yalom, I. D. (1975). The theory andgpractice of group psy-
chotherapy (2nd ed.). New York: Basic Books.

59

APPENDIX A'I

GROUP CLIHAT U STIONNAIRE 6C '5

Observer:
Group:
Day: Date:

Instructions: Read each statement carefully and try to
think of the group as a whole. Using the
Rating Scale as a guide. circle the num-
ber of each statement which best describes
the group during today's session. PLEASE
NARK ONLY ONE ANSWER fOR EACH STATEMENT.

I. The members LIKED and CAREO about each other

2. The members tried to UNDERSTAND why they do the things they
do. tried to REASON it out

3. The members AVOIDED looking at important issues going on
between themselves

A. The members felt what was happening was IHPDRTANT and there
was a sense of PARTICIPATION

. The members DEPENDED upon the group Ieader(s) for direction

5
6. There was FRICTION and ANGER between the members
7. The members were DISTANT and WITHDRAWN from each other
8

. The members CHALLENGED and CONFRONTED each other in their
efforts to sort things out

9. The members appeared to do things the way they thought
would be ACCEPTABLE to the group

lO.The members DISTRUSTED and REJECTED each other

ll.The members REVEALED sensitive personal information or
feelings

l2.The members appeared TENSE AND ANXIOUS

Rating Scale

0 not at all

i a little bit
2 somewhat

3 moderately

b quite a bit

5 a great deal
6 extremely

0 I 2 3 b 5 6
O l 2 3 b 5 6
O l 2 3 b 5 6

l23b56
123b56
3356
l23A56

0000
.e
N

O I 2 3 b 5 6

O l 2 3 b 5 6
O I 2 3 b 5 6

O l 2 3 b 5 6
O l 2 3 b 5 6

Please describe briefly the event that was most PERSONALLY important to you
during today's session. This might be something that involved you directly.
or something that happended between other members. but which made you think
about yourself. Explain what it was about the event that made it important

for you PERSONALLY.

EVENT IT'S MEANING TO YOU

-continue on back if necessary-
Copyright - I980 Dr. Roy MacKenzie

60

APPENDIX A-2

Group:

Initials:

Date:

‘0

2.

The members LIKED and CARED about each other

The members tried to UNDERSTAND why they do the things they
do. tried to REASON it out

. The members AVOIDED looking at important issues going on

between themselves

. The members felt what was happening was IMPORTANT and there

was a sense of PARTICIPATION
The members DEPENDED upon the group leader(s) for direction

There was FRICTION and ANGER between the members

. The members were DISTANT and WITHDRAWN from each other

. The members CHALLENGED and CONFRONTED each other in their

efforts to sort things out

. The members appeared to do things the way they thought

would be ACCEPTABLE to the group

lO.The members DISTRUSTED and REJECTED each other

ll.

12.

I3.

The members REVEALED sensitive personal information or
feelings

The members appeared TENSE AND ANXIOUS

Everything considered, I gained something of value from
today's session

UUDUW

'55

1‘5

‘l5

“5
‘05
‘05
‘05

A5

“5
‘05

‘'5
“5

“5

O‘C‘O‘G‘

0‘0‘

0‘0‘

..mma aa..m.-~ .mo._v a

.smo. ea..m.-~ ..o. v a
w .3.

.amma aa..ma-~ ..oo. v a
it...

61

 

 

 

mo:_m>-c ovum—ox pcm muc_cm po_comcouc_ m.cmoz

 

 

mmco_um_>up ocmocmumv new menu:

... u m_z ..< .maoz
m~.. NN. ..m.~- ~m.- as.. o..- ..s.e am. .~s.e m~.. .am.e ao.. nasuuqoe
.s. so. oh..- o~.- .m.- s..- .~:.V as.. .ms.v oe.. ..s.v 0a.. noucmam.a
me. D. 4.3.? mm.. 3.7 2.- 3.1V 3.. 8m; m5. 3m; 3.. .3293
mN. No. co..- a~.- aa.- a... ..~.e am.. .me.v ~m.. Ass.e m~.. scmc<

..mm.m- N..- as..- m..- ...om.s- om.- .Na.e am. .km.e N... .ma.e sw.. smacaaaa
ma..-.oM.- km.- A..- .os.~- ks.- ..m.e mm.. .mm.v oo.~ ..s.e om.~ mao.xc<
..m..m- N..- _m.- mo.- ...o.k- .m.- .m... m».. .s... ~m.. .ss.. sm.~ a>..me.oz
.3. mo. mm; 3. R.— .m. RR; mmé as; .5». Sm; 36 3.3.6..
.k. mo. ~m.. m.. km.. «N. .ss.e No.s .ks.v km.m .mm.e mk.m uuacotccou
mm..- m..- *.~c.m .s. .m~.~ NN. .s:.v -.s ..m.e .m.m .m~.e oo.s uoama.u.a.aa
N..- ~o.- .oa.~ mm. .o~.~ .m. .ms.e ~s.: .ms.e mo.s .m~.e ...s notmu
mm..- -.- mm.. MN. so. .o. .-.e -.s .om.e so.s .-.v a~.s uncommom
.m “a .m on m. .m
.x..ma-u.ue.z u.ee.z-uam. x.tmm..u.m. .mwmm u.ue.z .xwumm .mmmm

 

mue_cm oo_coocuuc_ to mo:_m>uu tam .mco_um_>oa ocmpcmum .msou_ oumE__u moccu mo mac—box cmoz .mcooEoz

.-a x.acoaa<

62

.zmaz am._ma-~ .mo._v a
*

.__ u m.2 __< .muoz

 

 

 

 

 

 

m:.. mm. mm..- ~m.- ~a.- s~.- .sm.v ~m.. .GN... m~.~ .mw.. .a.. amauomox
3.. ..m. R..- mm.- 8.- .o.- .3; mm; 87: m...~ .24 ..m.. 86:32..
om.. ms. .o.~- m~.- mo. .o. .o...v m:.~ .am... o~.m .No... N¢.~ sou.o><
wk.. km. oo. oo. m~.. km. .w.... oo.~ .so..e oa.~ .ma.. mo.~ ..mc<
mm.- s:.. we.- on.. ..m.~- :~.- .mo..v e~.~ .om... am.~ .mk.e oo.m naucaaao
k~.- mo.- mo..- s~.- sm.- mm.- .o...e ~o.~ .mo... am.~ ..m.. mm.~ mao.xc<
3. MN. 3..- .:f 8.7 Sf .34 $.N 3...: Tim $7: .~.m 3.3502
$.- omf 3.. R. 3.. A... 37: and” ST: mm.~ 35v mad 8.3.6..
3.7 mm.- mm; 3. mm.- ...- Rm; .3...” 8a.: 34. 3mg mm.m umacotcou
.m.- m~.- ~N.. .3. mo.. m.. .oe.. em.m .mm.v ms.m .me.e ma.m nuama.u...ma
mm.- m..- No. ON. mo. .o. .oo... om.m .s.... om.m .em.v ms.m enema
a..- mm.- A... .a. Na. NN. .5... ma.m .mm... No.m ..m.. Gs.m encommua
w m m. .m .m .m
m_.¢a-o_es.z m.na.:-u.m. .x.tmm-m.m. .mwmm o.ue.z .maumm .mmmfl
mo:_c>-. eosc.o¢ 6cm m.._;m no.comcouc. m.caoz .mco_um.>oo utmocmumv can name:

 

m.c.zm cc_coacuuc. mo mo:.m>-n ocm .mco_um_>oc ocmocmum .mEuu_ meme..u Quota mo mmc.um¢ cmoz .mcu>cumao

~-. x..c.a.<

63

ooo.

Neo.

avo.

“Ho.

coo.

moo.

Noo.

Hoo.

Hoo.

ooo.

ooo.

ooo.
coo.

 

Mflca
mucaomcvca_m

oe.m

mm.¢

-.~

no.5

NN.oH

em.-

em.m~

we.mH

No.m~

oo.-

mm.wH

oo.m~
-.m

om.
om
~o.~
om
on.
on
mm.H
oN
mm.
o~
mm.
on
em.
o~
mN.~
ow
fio.
oN
oo.
om
me."
ow
om.
ow

 

«m2 coccu
\mo coccu

O Ch :-I 1.0 L0 N
m In as c-n 0‘ H
O O O O O 0
FIG Hm MO u-ta'l o-am e-EIN c-om Fla) v-n-i s-ILD piq- P-‘g—d
I". e-O

Q
N

O
N

N
0-.

 

em: mwmwzuoohr
\uo mpmmguooxz

m¢.mo

Nu

uaaaa.u.acaa
wooemumwo
concocmeou
umpmm>mm
xcmc<
cacao
anaconda
omuwo><
m:o_xe<
uncomaom
anaconda
m>wuascoz

mcm>cmmno
mzmcm> acumen:

mumucueou
uoceopa

 

mupammm Aom ..v ammo-m.mao_cu>_:: an ooze—pom .meopmmom mono new .mpvu.z .x—cou

 

mmoco< muue__u aaocu to mmc.uom mcm>cwmno use meanemz co oucmwca> mo «Fax—ac< opa.u_:z co»

m» Fuzz—mac:
.o xwecmga<

64

«mm.
ooo.

Hoo.
mvo.
Hmo.
«ma.

cam.
oNN.
ooo.

ohm.

owe.
moo.
moo.
ooo.
moo.

 

M to
oocmommficmwm

oH.H
No.m
oh.~
Nh.m
oh.v
oh.a

mm.~
Ho.H
vh.N
mH.H

mm.
o~.m
om.v

mo.o~

~m.o

M

ma
ma

ma
ma
ma
HA

ma
ma
ma

ma

om
oN
om
om
ha

oH

Q'o-Ir-i

 

 

.mz Locum
\mo coupe

am: monocuommm
\ho mmmocuomam

ooumu
ooamopom

coucmumfio

ooofio><

moumm«0wuwmm
Ho.mm puma
momuo> manna:
monum> wanna

wooed

concoumcoo

oouoomom
ov.a mung
mommu> oaoowz
mbmwu> >mem

uncommom

unamxcm

wofi>mzom o>fiumEuoz

uncommon

mn.o~ out: nonwo> wanna

 

H nunmwucOU

m ooccoam

.mummul M oumwum> an oosoaaom .muo>womoo ocm opened: an encummom comm ocm .oaooaz

.xaumm mmouom mmcaumm oumEAHo ooouw mo oucmfium> mo momaamcm oaowuasz new m.

N+ m .mcaaaouom

 

a xHozmmmo

.t .o. .o. .2 rat...E ..o...eo m.25.... ..< ...ge

2.... 2....-. .2 .a 2....
..... ......- ... .. ......
...o. ......-. ... .a .......

 

 

 

 

an... no... a... 3.. a... no... :3 3.: an... 3.: a... a... . .35 2...!
_
2.. 2.. 2... 2.. .... 2.. 2.. 2.. 2.. 2.. 2.. 2.. . .u...+1..hs..2
. .2 .2 .2 .2 .2 .2 2 .2 .2 .2 2 2 .2 .2 2 .2 2 2 .2. .2. . 2- 2. 2 :- ...- 2- .2. .2. 2 a..- .2- 2 2...... ..
.2 .2 .2 . . . .2 .2 .2 .2 .2 .2 .2 .2 .. .2 .2 2 .2 .2 .2 .2- .2. 2. 2. 2. 2. .2. 2. u... .2- 2- 2. .2- 2. . .82.... ..
.2 .2 .2 .2 .2 .2 . . . . .2 .2 2 .2 .2 .2 .2 .2 2 2 .2 ...- .2- 2. .. 2 . 2.. 2- .2. .2. . ..- .2. 2.. . 8...... ..
"MW .. ... ... ... ... .2. . ... . .. . . . ... ... ... .- ... .. o. .. ... ..- .... ... ...- ..- ..- .... o..- o..- ... .. ..- ..- ..- .. ..e.... ..
.2 2 2 .2 .2 2 2 .2 .2 .2 .2 .2 . . . 2 2 . .2 2 .. 2. 2. .2- 2- 2 2. 2- 2. 2- :- . 2- 2- 2 2. 2...... ..
.2 .2 2 .2 .2 2 .2 .2 .2 .- .2 .. 2 2 . . . .. 2 2 2 .2. .2. 2 .. 2 .2 2. 2. .. .2- ..- .2 .2. 2. .2 2.... ..
.2 .. 2 2 .2 .2 2 2 .2 2 2 .2 .2 2 .. 2 2 2 . . . 2. 2. . 2 2. 2- .- 2 .2- 2- .. 2- 2. . 2. .22.... ..
2.. 2... 82 in m... t... e... 2.. 2.. . 2.. 2.. 2.. . ..F 3.2.2...
.2- .2- 2 .222- ... 1...... 2. 2. 2. 2. 2. 2. 2. 2. 2. 2 2. 2. . . . . .2 2 .2 .2 .2 .2 .2 .2 .2 .2 .2 .2 22...
2- ... 2 2. 2. 2. .- 2 . .2. 2. 2- 2. 2 2. 2 2 2 2 2. 2. .2 2 .2 . . . .2 .2 .2 2 .2 .2 2 .2 .2 222......
2. ...- 2- .2- 2. .2- 2. 2. .2. .2. .2. .2. 2. 2. 2- 2. 2. .- .. 2 .2. .2 .2 .2 .2 .2 .2 . . . .2 .2 2 .2 .2 2 2...... .u
.2- .2. 2 .2. 2. 2- .2- . .- 2.. .. 2. 2. . 2- .2- 2. .2 2.. .. 2. .2 .2 .2 2 .2 .2 .2 .2 2 . . . .2 .2 .2 2.22.22 ..
.2- .2- 2 .2. 2. . .2- 2. . 2. 2. .- 2- 2 2- .2- 2. .2 2- . 2. .2 .2 .2 2 .2 .2 .2 .2 .. .2 .2 .2 . . . 2...... .2
- .— - u a - - .-
. u m “ llPfiLr HI+ILF HILTIP.al|¥lhl . u u u a u u m u -w u .H a m .
nonmaNo damnemeKk-eroz Muuesz mowoemNmkka

 

fl..z.mv mum. ocm .o:o.mo.z .>.cmw um m0:0co co>o.w co meo._ o.mE..o Quota ~— eo mmc_um¢ coo: mcoe< mco.um.occ0u oommm-conEuz

a-.. data...

 

 

 

.2.. .... .... .2.. .... .... ....
.... .... .... .... .... .... ....
. ..- ... ... 2. .. ... 2. ... .. .- .. .- .. 2. 2. .. .. ...
2.- ... .. . . . ... .. .. ... .. .. .. . .. uw. .. . ..- .. .2
.. 2. ... ... .. .. . . . ..- .. .. 2. .. .- ... ... .2 .- . 2
2. ... .. ... .. .2 ... .. .2 . . .2 2. ... .. ... ... .- 2. .2
.- .. .. 2. . .. .. .. .- 2. .. ... . . . ... 2 .. ..- ... ...
2. .2 ... ... .. . ... ... .2 o. 2.. ..- ... . .. . . . ... ... ...
o. .. ..- ... .. .2 .- a 2 .. .. .. ... ... ... ... ... ... . . .
,nv
[nu
.... .... .... .... .... .... .2..
2.. .... 22. .... .... ... .... ... .2 .... ... .- ... ... ... .2.. .... .. .. .2. ..
.2. ... ... .. ... .. ... .. .. .2.. .... ... 2.. . .. .2. .. ... . ... ..-
.. ... ... ... 2.. . .. . .2 .... .... ... 2.. ... ... ... .- ... . .2. ..
2. .... . . 2.. ... ... ... ... .. ... .. ... .. ... .. ... .. ... .. ..-
2. .... .. .. ... ... ... .. .. .... .... ...- ..- ..- ...- .. 2 .. .2 ... ..-
m n m m m u a. m u a m .m- u w u u m n u m u
k. . q .
on mw ”w Mu w. N “V mw a. m2 n2 Nu up .N .w. n“. ~v .2 2w _h~

0...
a...
nun Can-
‘0! 1....
00¢. :-
0’.0 .00
0.. on.
0‘ .a o
a a . a
on. on
0—. b:
no. 0A
000 b:

can
an
on
on-
n...
R...

....
..- ... ...
..- 2.. ..
...- 2. .2
...- .... ..-
... . .-
.2. .2 ...
. ... ..-

....
... .. ..-
... .22 ...
... .. .
... ... .2

..ITF

....
....
..- .... ...
... ... .
..- . ..
...- ...- ...
..- ... ..-
... .- ...
. ... 2.
....
... ... .
... ... ...
... .. ..
... ... .

....
....
.- .... .
. ... ..-
..- ..- ...
.- ..- ..
..- .. ...
.- ... ..
... .. ...
....
... .. ..
... .. .
... .. ..

... .. ..

..... ......-. ... .. ....u.

..... ......-2 ... .. ....

..... ........ ... .. ......u
a... . .2.... .2....

:4 28““ '93....mccol

.. .... 2 2...... ..
.. ... ..- ......... ..
..- .. .2 ....... ..
.... ...- .... ....... ..
..- ..- ...- ........ ..
.. . .. ..... ..
.2 ..- ... ......... ..
.... u.uN 222:...
... ... .2 ...-u ..
... ... .2 .........u ..
... ... o cocoooou .u
... .. .. ............ ..
. . . ........ ..

..ulml...

 

a m u m 3 ~ 9

24.x.mv mum. new .aaogmu_x .>_.mw Lo. monogc cm>m_m .0 we... mumE__u aaogu N. :0 mmc_umx cmoz mcoE< mco_um_m..0u nmmmm-Lo>.mmno

..-. ...?me