Group Title: effects of member orientation upon the development of group structure
Title: The effects of member orientation upon the development of group structure
Full Citation
Permanent Link:
 Material Information
Title: The effects of member orientation upon the development of group structure
Physical Description: viii, 95 leaves : illus. ; 28 cm.
Language: English
Creator: Penrod, William Tony, 1929-
Publication Date: 1962
Copyright Date: 1962
Subject: Small groups   ( lcsh )
Motivation (Psychology)   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis - University of Florida.
Bibliography: Bibliography: leaves 71-73.
Additional Physical Form: Also available on World Wide Web
General Note: Manuscript copy.
General Note: Vita.
 Record Information
Bibliographic ID: UF00097972
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000565645
oclc - 13569812
notis - ACZ2064


This item has the following downloads:

effectsofmembero00penrrich ( PDF )

Full Text







June, 1962

ri I Ei ll F FL RID 21

3 1262 08552 2158

To my parents


I wish to express my gratitude to the chairman of my

supervisory committee, Dr. Marvin E. Shaw, for his infinite

patience and continual encouragement during the preparation

of this dissertation.

An expression of appreciation is extended to the

members of my supervisory committee, Dr. W. B. Webb,

Dr. R. J. Anderson, Dr. R. W. Waters, and Dr. A. W. Combs

for their cooperation and helpful suggestions.

I am indebted to the Department of Psychology for its

cooperation in obtaining needed equipment for this research.

I am also indebted to Dr. B. M. Bass of Louisiana State

University for permission to use his unpublished SIT




AC0KOWLEDGMENTS . . . . . . . . . iii

LIST OF TABLES . . . . . . . . ... vi

LIST OF FIGURES . . . . . . . . . viii

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

A. Statement of the Problem 1
B. Theoretical Background 3
1. Orientation of Group Members 3
2. Origin and Nature of Group Structure 11
C. Basic Assumptions 23
D. Hypotheses 26

II. METHOD . . . . . . . . . . 29

A. Subjects and Groups 29
B. Task 30
C. Dependent Variable Measurements 32

III. RESULTS . .. . . .. .... . 36

A. Interaction Trends 36
B. Role Differentiation 43
C. Status Consensus 51
D. Role and Decision Satisfaction 52

IV. DISCUSSION . . . . . . . . . 53

A. Interaction Trends 53
B. Role Differentiation 58


REFERENCES .* . . ... . . . . . . 71

APPENDIX A .... . . . . . . . . 75

APPENDIX B . . . ... . . . . . 80

APPENDIX C . . ........ .. . . 83


APPENDIX D .. . . . . .......... 85

BIOGRAPHICAL SKETCH . . . ... ... .. 95


Table Page
1 Bales' Interaction Process Analysis
Categories . . . . . . . . . 33

2 Total Acts Per Session for Combined
Categories . . . . . . . . . 37

3 F Values for Combined Categories . . . . 37

4 Total Number of Acts Per Session for
Selected Separate Categories . . . . 39

5 F Values for Separate Categories . . . . 39

6 Total Number of Negative Emotional Acts
in Those Groups Having Different Orders
of Presentation of Information . . . . 41

7 Total Amount of Activity for Each Session . 43

8 The Number of Cases Out of a Total of 48
Sessions in Which an Individual Held Top
Rank on One and Only One of Four
Characteristics . . . . . . . 44

9 Number of Groups Out of Eight in Which the
Same Individual Held Top Rank on Both the
Idea and Liking Characteristic Per
Session . . . . . . . . . 46

10 Number of Groups Out of Eight in Which the
Same Individual Held Top Rank on Both the
Guidance and the Liking Characteristic
Per Session . . . . . . . . 46

11 Mean Bank-Order Correlations Among Rankings
on Talking, Ideas, Guidance and Liking . . 47

12 Mean W for Guidance and Idea Rankings Per
Session . . . .. . . . . .. 51

13 Mean Ratings for Role and Decision
Satisfaction . . . . . . . . 52

14 Analysis of Variance for Combined Category
A, Positive Affect . . . . . . . 75


LIST OF TABLES (Continued)

Analysis of Variance for Combined Category
B, Giving Answers . . . . . .

Analysis of Variance for Combined Category
C, Asking Questions . . .

Analysis of Variance for Combined Category
D, Negative Affect . . . . . .

Analysis of Variance for Tension Release

Analysis of Variance for Agreement . .

Analysis of Variance for Giving Opinion .

Analysis of Variance for Giving
Orientation . . . . . . .

Analysis of Variance for Disagreement . .


. 76

. 76

. 77

. 77

. 78

. 78

. 79

. 79











Figure Page
1 Relation of sociometric choice to
rank on Talking ........... 49

2 Relation of sociometric choice to
rank on Idea . . . . . . 50

3 Relation of sociometric choice to
rank on Guidance . . . .. .. 50




A. Statement of the Problem

The present investigation was designed to study the

effects of the motivation of group members upon the subse-

quent development of group structure and functioning in

small problem-solving groups. Most of the research om small

groups has been performed with collections of individuals

who are previously unacquainted, who meet for less than an

hour and are then disbanded. In only a few instances has

attention been given to changes in group processes that

occur over a series of several meetings. While theorizing

has been comparatively prolific, there has been until

recently little experimentation directed toward discovering

whether systematic changes in group structure can be detected

and measured in any way. The study of group structure is

vital to a rigourous development of group therapeutic

approaches, as well as to the planning of groups constituted

for special purposes.

Most group psychologists admit the importance of

motivation as a factor in determining the nature of the

processes which go on in the group. Yet little work has

been done to make it possible to study the operation of



motivational processes in the group situation, especially

with respect to changes that occur over time. There has

been increasing recognition that the outcomes of a group's

behavior are not merely determined by the intellectual

processes, i.e., the ability to solve problems. Clearly,

the individual possesses personal needs and characteristics

that partly determine his behavior in a group situation.

Each person, because of his unique needs and interests, can

be expected to be more sensitive to certain aspects of a

group's operation than to others.

Since there is an almost infinitely large number of

attributes available for characterizing an individual, it is

difficult a prior to decide just what classification of

personality characteristics is most useful in studying small

groups. However, when individuals are observed performing

in groups, one tends to notice what might be referred to as
"group-relevant" characteristics, such as approach or

commitment to the task, ability to get along with others,

leadership ability, etc. Individuals seem to be oriented

more toward some of these aspects of group functioning than

to others. Thus, the concept of "orientation" seems to

offer a fruitful conceptual approach to group member

motivation and is the one applied in the present study. By

"orientation" is meant each individual's unique set of

predispositions for perceiving and responding in character-

istic ways and for expressing certain affect in group

situations. These predispositions lead him to exhibit a

reasonably stable behavioral pattern that involves

interacting with others in developing and maintaining a

particular pattern of interpersonal behavior.

The problem under investigation may be restated as the

study of the relationship between certain Norientations"

persons have in group situations and the patterning of the

interpersonal behavior that arises out of their interaction

with other group members. Thus, if the members of two

groups perceive the group differently, it is expected that

distinctive structures should appear. The concept of group

structure may be further differentiated in terms of the

patterning and kinds of interaction that takes place and the

style of informal leadership characteristic of the group.

The style of informal leadership is defined in terms of role

structure. By structure is meant the patterning of the

interpersonal behavior in the group. For purposes of the

point of view being developed, role structure refers to the

expectations relative to the performance of each group


In the next section an attempt will be made to define

more fully the terms orientation and role structure as used

in the present study. Relevant research will be cited to

support the concepts used.

B. Theoretical Background

1. Orientation of Group Members

There is evidence in the literature that various per-

ceptions of social situations result in different group

structures. From these studies there may be seen at least


three sets of conditions that contribute to these differing

perceptions of social situations: the physical and social

environment, instructions and experimental procedures, and

personality characteristics.

Physical and social environment

The effect of the physical environment on group

structure may be seen in a study by Lipset, Trow, and

Coleman (1956). These authors report that night workers

are more deeply involved in the printers' community than day

workers. They demonstrate that the greater the proportion

of a printer's career spent on night work, the higher his

score on a "social relations index." Day workers were

subject to pulls from mass entertainment, neighborhood

organizations, and nonprinter friends. When the day workers

went home they were expected to conform to the pattern of

activities of their families. Night workers' schedules were

so unlike that of their families' that they were not

expected to conform. Compared to the day shift, the pace

of the night work was more relaxed, and supervision less

strict. This greater freedom of the night workers tended

to facilitate socializing on the job.

In his study of a gypsum plant, Gouldner (1954) found

that miners, as contrasted with surface men, enjoyed greater

social cohesion and more extensive social relationships.

To the men on the surface the conveyer dictated the pace and

position of the workers. Miners, however, worked in larger

groups and in much closer association. Gouldner observed no

instances of workers helping one another on the surface, but,

if a miner were injured, others would come to his aid and

take up collections for him, even from miners on other

shifts. Unprompted group action, such as workers walking

off their jobs if a miner were hurt by falling rock, was not

uncommon. Gouldner concludes: "In the main, two factors

were closely connected with the greater cohesion among

miners: (1) The peculiar work and spatial arrangements in

the mine and factory; (2) Ihe more hazardous working

conditions of the mine" (p. 136).

The research on human relations by the Survey Research

Center of the University of Michigan, reported by Kahn and

Katz (1960), indicates the influence of the social environ-

ment on the behavior of employees in industry. Kahn and

Katz abstracted four classes of variables which are

consistently related to the productivity of the group:

supervisor ability to play a differentiated role, closeness

of supervision, the quality of supportiveness or employee-

centered orientation, and the amount of group cohesiveness.

They suggest that the effect of supervisory behavior on

motivation may be basic to understanding productivity

differences between groups.

Following Kahn and Katz, Marcus (1960) studied the

effect of employee-orientation of the supervisor on the

behavior of social workers in a large public welfare agency.

Not only did he expect the groups to differ in the type of

interaction displayed, but he also predicted that different

types of informal leaders would emerge in groups with

different kinds of supervision. If the supervisor supplies


the social-emotional qualities or is "employee-centered,"

then the group should respond with higher orientation toward

the task; the group member who best approximates this goal

will become the leader. On the other hand, if the super-

visor is production- or task-oriented, then the group will

become more oriented toward the social-emotional aspects of

group life. The leader who emerges in this case will be the

one who is the most supportive in this area. Both predic-

tions were confirmed.

All of these studies tend to support the general

argument that environment becomes a crucial determinant of

the structure of group relations.

Instructions and experimental procedures

That differential instructions or experimental

procedures may induce different orientations or sets toward

the group is shown by a second group of studies. Lewis

(1944) used the twofold classification of task-orientation

and ego-orientation in her research on the recall of

interrupted tasks. Ego-orientation is defined as being

motivated primarily for self-enhancement or self-reward, to

the exclusion of all other demands either from the objective

environment or from other persons. A person who is task-

oriented responds to the objective situation "for its own

sake," i.e., completing the task is the source of satis-

faction. It was found that a greater recall of interrupted

tasks depends on the existence of task-orientation. When a

person is ego-oriented, recall favors the completed tasks.

In their study of social conformity, Thibaut and

Strickland (1956) distinguish between two psychological sets

that are taken by persons in evaluating the judgments,

perceptions or attitudes that are communicated to them by

other persons. One of these, called aroun set, describe the

individual who is concerned with achieving or maintaining

membership with other individuals whose attitudes are being

communicated to him. A person is expected to adopt this set

when he is motivated to gain membership in the group,

threatened with loss of membership, and when he is informed

by the group that integrative behavior is necessary for

group survival. On the other hand, in a tafk set, the

person perceives other group members as "mediators of fact,"

or as instruments or standards to test his own judgments.

In this case he is concerned, not with maintaining a social

relationship, but with achieving cognitive clarity about

the environment.

One series of experimental studies has demonstrated the

influence of motivation on the way a person interprets the

interpersonal relations assumed to be present in a series of

pictures depicting various social situations. McClelland

gt al. (1949) have developed what appears to be a valid

method for measuring achievement motivation by means of the

content analysis of brief imaginative stories composed in

response to the Thematic Apperception Test. Achievement

motivation, similar to what has been called task-orientation,

was aroused by an experience of failure prior to responding

to the TAT pictures. It was found that those who were

scored as having high need for achievement showed a

significant increase in general achievement imagery,

achievement-related deprivation themas, successful instru-

mental acts, and anticipatory goal responses.

Similarly, Shipley and Veroff (1952) and Atkinson ej al.

(1954) have reported the effect of the affiliation motive

on shifts in perception and in the thematic content of

stories. The affiliation motive is parallel to what has

been called group set or interaction orientation. The cues

used to arouse this motive were describing others on

personality traits, being described by others, and choosing

personal friends. Behavioral sequences dealing with

positive affective relationships with other persons occurred

more frequently in the imaginative stories of the group in

which the affiliation motive had been aroused. The control

group was composed of subjects who performed a ten-minute

anagrams task following a task-oriented-type instruction.

The Ss above the median in need for affiliation score were

described as approval-seeking by other group members

significantly more frequently than Ss below the median score

in the need for affiliation.

These studies provide evidence that different sets,

established through instructions or experimental procedures,

influence the way a person views interpersonal behavior.

Personality characteristics

A third set of studies deals with orientation to group

situations as a function of personality patterns. Perhaps

one of the first attempts to investigate the effect of

member motivation upon group development was made in the

Human Dynamics Laboratory of the University of Chicago and

at the National 'Training Laboratory in Group Development in

Bethel, Maine. Stock and Thelen (1958) summarize the results

of this research in a recent publication. The two concepts

of emotional culture and valency are fundamental in this

theoretical approach. The term group culture refers to the

situation in the group as a whole. The concept of valency

refers to the relation of the individual and the group

culture. Valency is defined as an inherent property of the

individual that accounts for the nature of his participation

in the emotional aspects of group life. These emotional

aspects of group life, called the emotionality culture, have

been classified into three basic categories according to the

characteristic kind of affect expressed: pairing, dependency,

and fight-flight. A sentence completion test known as "The

Reactions to Group Situations Test" (EGST) was developed for

identifying the valency patterns of individuals. An analysis

of the test protocol provides information about the

individual's typical affective response to group situations.

The RGST has been applied mainly to training groups designed

to help members improve their understanding of group opera-

tion and of their own feelings and behaviors in group


Wolmaan (1956), in his study of the situational factors

in leadership, took as his starting point the emphasis on

perceptual factors in group situations. He states that the

total situation as such will not account for leadership

choice. It is the porcletion of the situation which is the


decisive factor in the making of leaders. He suggests a

threefold division of groups based on the goals of the

individual members. In the first class are those who join

groups having the satisfaction of their own needs in mind.

These are called "Instrumental Groups." In the second place,

persons join groups with the basic need for mutual accept-

ance. Hence the name "Mutual Acceptance Groups." Still

others join groups for the purpose of serving a goal outside

themselves. These are called "Vectorial Groups" because

they move in the direction of a common ideal.

From this assortment of studies it is possible to

abstract three basic kinds of orientation. These include

self-orientation, interaction-orientation, and task-

orientation. Bass (1961) has done some extensive work on

defining and measuring these three kinds of orientations.

He describes self-oriented persons as those who are

attracted to groups because of the expectation of obtaining

personal rewards and satisfactions regardless of the task or

interaction effectiveness of the group. Fouriezos, Hutt,

and Guetzkow (1950) centered attention on self-oriented

behavior, showing that it can be reliably assessed by

observers of discussion groups. Dependent behavior,

succorant behavior, status-seeking, and domination attempts

characterized self-orientation in discussion.

Interaction-oriented persons are described by Bass as

gaining their basic satisfactions from the interpersonal

relationships in the group. They seem to be less concerned

about getting the task done, and more concerned with forming

friendships, sharing with others, and providing personal

security. They wish to avoid any behavior that would bring

disharmony to the group.

Task-oriented group members are described as being

attracted to the group by the expectation of task achieve-

ment and reward. They are concerned about getting the job

done and tend to be persistent in overcoming any barriers

in the solution of the external problem.

These, then, are some of the theoretical considerations

that have been followed in defining and categorizing member

motivation in group situations. It is a common observation

that the patterning of interpersonal behavior changes over

time. The evidence reviewed above implies that structural

development in a new group does not depend solely upon

changes inevitably linked with time. These changes may also

depend upon conditions which may be different from the very

first, either in the original composition of the group or

in the situation which they face. The present investigation

is primarily concerned with the changes that occur because

of the composition of the group.

2. Origin and Nature of Group Structure

One important aspect of the social organization of

groups which endures over time is the fact that there

develops a patterning of relationships, that is, individual

members of the group can be located in relation to other

members according to some criterion of placement. There are

several studies which suggest that differential structure is

perceived almost immediately by the members of a newly

created group. Barker (1942) has demonstrated that a group

of strangers, after a few moments of being acquainted,

exhibits a high degree of unanimity, not only in describing

each other's behavior and appearance, but also in choosing

members for seatmates. Further, expressions of choice are

highly related to descriptions of behavior and appearance.

Suchman (1956) found that members of small experimental

groups, after thirty minutes of interaction in the perform-

ance of a task, were most accurate in estimating the

expressed feelings of those other members toward whom they

reacted favorably. TSguiri, Blake, and Bruner (1953) also

report that the members of experimental groups are able to

estimate each other's feelings with a higher degree of

accuracy than expected by chance.

Much has been written about the reasons that groups

become structured, and we shall not attempt to review all

of this literature. /It will be useful, however, to note

that at least three kinds of factors tend to produce stable

differentiations within groups. The first set stems from

the characteristic interaction patterns of individuals, the

second arises from the different abilities and motivations

of different individuals, and the third derives from the

requirements for efficient group performance.

Homans (1950), in his classic analysis of the human

group, shows that variance between members in the initiation

of interaction is the basic condition which underlies the

emergence of differentiated roles. He hypothesized that

"a person of a higher social rank than another originates

interaction for him." Homans also found that the higher a

person's social rank, the wider will be his range of


Jennings (1950) shows that/individuals differ not only

in the number of persons they characteristically choose as

interaction partners (emotional expansiveness) but also in

the number they contact as interaction partners (social

expansiveness). It appears that the capacity of the

individual to maintain choice relationships with other

persons is a highly stable characteristic. Although a group

may present equal opportunities for interaction to each

member, the opportunity is not utilized equally by the

members. They differ not only in the number of persons with

whom they interact but also in the number of members with

whom they can initiate and accommodate reciprocal choice

relationships. Borgatta and Bales (1953) also discovered

that each individual in a four-person group appeared to

exhibit an upper boundary which represented the limits of

his capacity to initiate interaction no matter how much

opportunity he had to participate.

Bales (1950) describes the process of role differ-

entiation as emerging from the kind of interaction

characteristic of each group member. Social -structure,

according to this viewpoint, can be understood primarily as

a system of solutions to the functional problems of inter-

action that become institutionalized in order to reduce the

tensions due to the unpredictability in the actions of the

group members.

On the basis of such a theory, Bales and Strodtbeck

(1951) have presented a method for testing the existence of

differentiated phases in group development. Briefly stated,

the phase hypothesis predicts that groups tend to move in

their interaction from a relative emphasis upon problems of

orientation, to problems of evaluation, subsequently to

problems of control, and that simultaneously with these

transitions the relative frequencies of both negative and

positive emotional reactions tend to increase.

Heinicke and Bales (1953) have observed the develop-

mental trends in the interaction structure of small problem-

solving groups. They found that initially the members

exhibited a high degree of task-oriented interaction, but

this preoccupation with the task declines in the second

session and thereafter. Comparatively small amounts of

disagreement and tension are exhibited in the first session,

but these show a sharp increase in the second session. From

the second session on, negative reactions decrease but

positive reactions tend to increase. Amount of agreement

declines sharply from an initial high level, but this is

compensated for by the sharp rise in solidarity and tension

release. It was observed that the second session was

generally the critical one for the differentiation and

establishment of structure. Philp and Dunphy (1959),

observing groups in Australia, found both similarities and

differences in the development of the interaction structure

of their groups as compared with their American counterparts.

They found that task-oriented interaction increased in the

second session from the first session, while there was a

decrease in the amount of positive affect in the second

session. These trends are just the opposite of those found

by Heinicke and Bales. It can also be noted that both

positive and negative reactions did not attain as high a

level throughout the sessions as did the groups of Heinicke

and Bales. However, Philp and Dunphy did observe that the

second session constituted somewhat of a crisis in the

development of the interaction structure. They attributed

the differences to the type of problem involved and the

differing resources of the members of the respective groups.

These studies provide strong evidence to support the

view that individuals differ in their capacity to initiate

and maintain interaction with other persons, and that these

differences account to a very large degree for the differ-

entiation of structure in the group. They also demonstrate

that systematic changes in the interaction structure over

time can be observed and measured.

In the second place, many writers have looked for the

origin of group structure in the characteristics of the

individuals composing the group. For example, Barnard (1946)

stresses the way in muich individual differences in ability

and temperament lead people to do certain group tasks

themselves and to give other tasks to other people. In like

manner, some persons like to assume responsibility while

others prefer to be told what to do. Jennings (1950)

reports that the emergence of status differences is related

to personality differences among the group members,

particularly in reference to spontaneity and personal

security. Carter (1954) found that functional differentia-

tion is related to differences in the group task and to

differences among the members in task-related skills.

Thirdly, the development of a differentiated group

structure stems from requirements for efficient group

performance. Heinicke and Bales (1953) have observed that

newly assembled groups are able to make only abortive

attempts to complete the group task so long as the role

structure remains undefined. The members seem compelled

to direct their efforts toward the differentiation of

function and status. Once they have developed and

acknowledged a role structure, the members are able to go

ahead with the group task. Bavelas (1948) has invented

designs which impose strictly controlled communication

channels upon experimental groups. A "circular" design

permits each member to communicate only with one member

immediately to his right and left. An "all channels" design

permits each member to communicate with every other member.

A "wheel" design makes one member a co-ordination center,

and all other members are able to communicate only with the

central, co-ordinative member. Groups with wheel designs

quickly develop clearly defined role structures with the

co-ordinate member occupying the high status position in

the group.

Other studies emphasize the role of the solution of the

internal and external problems of group life as the

prerequisite to the development of group structure. Bales

(1950) sees a group as having two basic tasks to perform:

(1) to solve the objective problem which confronts the group

and (2) to build, maintain, and regulate group life. Thus,

all group activities are directed to task goals and system

goals: task goals being the problem the group is trying to

solve or the task for which the group exists to undertake;

and system goals being those of integrating the group's

interpersonal relationships. The evolving group structure

is regarded as a product of the solutions to early

instability in interpersonal relationships, which become

more standardized or institutionalized so the group members

can be free to devote their full energies to the task.

Stock and Thelen (1958) have formulated a similar conception

of group structure. According to their point of view, a

group can be described at any point in its existence as

operating in a particular work-emtionality culture. In

other words, a group san be described in terms of its work

activities, the emotional basis on which it is operating,

and the dynamic relationship that exists between these two

aspects of group life.

Beniis and Shepard (1956) have elaborated a concept of
a six-phase sequence of group development derived from their

observations of training groups. They compare the develop-

ment of group maturity to the development of individual

maturity. They believe that group maturity involves the

overcoming of obstacles to valid communication among the

members, and the development of methods for testing group


consensus. The main obstacles or problems to be solved

before valid consensus can be arrived at are the problems

of authority and personal intimacy, love and power,

dependence and independence.

It is clear from the above that members of groups find

it necessary to develop a recognized role structure before

they can devote their efforts to effective task achievement.,1

The conception arrived at is that the activity or communica-

tion of group members is devoted to the solution of two

basic problems: the internal emotional or interpersonal

relationships, and the external task imposed on the group.

This activity or interaction is conceived of as being

distributed in time and among persons. The former assumption

is the basis for the phase sequence hypothesis of Bales and

Strodtbeck (1951) mentioned above. The latter assumption

implies that members will contribute differently to the

solution of these problems because of their unique pattern

of abilities and personality characteristics, and that

insofar as these differences are stable, they may be

referred to as roles. /Role" in this case is defined as

"a patterned sequence of actions and reactions performed

by a person in an interaction situation" (Sarbin, 1954). It

should be quite clear that the term role in the present

context of the small group situation is not the same thing

as the one found in the large scale social system. In the

latter context there is a recognition by the actor and by

others of the obligations, both duties and restraints,

involved in a role (Neiman, 1951). By role here is meant a

function or pattern of behavior which a group member more

or less exhibits in the group process.

The development of stability and the corresponding

emergence of roles is predicated on the notion that

stability fulfills certain important functions. Differ-

entiation in kind and amount of activity comes to serve

a useful purpose in the group, not only in terms of task

efficiency, but also in increasing the accuracy with which

members can predict when, how, and why other members are

going to behave. Group members thus come to have roles

which involve both very loose prescriptions and proscrip-

tions in behavior and also special relationships to the

external task. The role expectations have been analyzed by

several researchers in terms of actions, or task activities,

and gualitisj, or kind of affect exhibited in the group. In

summary, the/structure of the group may be described in

terms of the differential interaction, performances and

abilities, and expectations of the members./

In a recent book Parsons and Bales (1955) present a

theory of role differentiation in small social systems in

which the major generalization specifies that leadership

structures in all small groups are uniformly differentiated

alongthe above described instrumental-expressive axis. The

empirical data upon which this generalization is based are

primarily a series of laboratory studies by Bales and Slater

(1955) and Slater (1955). It is this theory that provides

the basic point of departure for the present experiment.

The basic theme is that the task demands and secial-


emotional demands of the group lead to the emergence of

persons who more or less specialize in one of these aspects

of group life. The task specialist is one who represents

the task values of the group members. The social-emotional

specialist is one who represents other values and attitudes

which are threatened or deemphasized by the task require-


Measures of how much a man talks are obtained from the

interaction records. Measures of how the members judge each

other's ideas, guidance ability, and how much they like each

other are obtained by a questionnaire after each meeting.

A simple measure of the degree to which these measures tap

different aspects of the differentiation of function is

obtained by asking the question: How many times are found

in which a person ranks first on one characteristic but does

not rank first on any of the other characteristics? A

"specialist" is then defined as one who achieves isolated

prominence in only one of these areas. In like manner, if

there is any other characteristic which achieves prominence,

then it might be considered to identify an "axis" of

differentiation. The principal type of differentiation is

revealed by the separation of the task rankings from the

rankings on liking. Role differentiation, then, seems to be

bipartite, with a task specialist and a social-emotional

specialist. Sometimes differentiation is tripartite, with a

person ranked high on total activity being prominent.

Usually, though, activity (talking) is highly correlated

with task status.

Grusky (1957) offers supporting evidence for the role

differentiation generalization of Parsons and Bales by

analyzing the behavior of a "natural" small group. The

object of the study was the staff of a psychological clinic

in a large mid-western university. A sociometric test was

followed by personal interviews. In the course of the

observations, a phenomenon called familial role differ-

entiation emerged. The investigator noted two specialized

roles that characterized the role structure, the "father-

like" role and the "mother-like" role. The father-like

figure had a strong sense of responsibility, a general

concern and interest in the control of the group process,

and was efficient and task-oriented. The function of the

father-like figure centered around the group goals. The

mother-like figure centered around the prolongation of

smooth and friendly interpersonal relations in the group.

She was characterized by her warmth and expressiveness in

her social relationships. The function of role differ-

entiation, according to the point of view here presented,

is that of setting up integrative roles which together act

to preserve the solidarity of the group.

Other studies seem to support the concept of three

basic roles which correspond to Bales' categories of

"activity," "task ability," and "likeability." A recent

study of Wispe (1955) is concerned with a set of ratings

and sociometric choices made of each other by a group of

insurance salesmen in an insurance organization. After a

factor analysis of his data he finds "a paradoxical

situation: the hard-driving agent, who may be a valuable

asset on one's debit, is not the person to invite home for

a relaxing evening, while the person with the compassionate

qualities, who makes a pleasant house guest, is not the

person to select if you have to make your daily quota of

sales. This conflict of expectancies reveals the agents's

dilemma. As insurance salesmen these men would like to be

successful and as human beings they would like to be

accepted. Yet, according to the analysis, the traits which

make for success as an insurance salesman preclude accept-

ance as friend."

Wispe found three orthogonal factors which he describes

as follows:

Factor A, insurance intelligence, has high
leading on items 4 and 7, which pertain to "help
with insurance problem" and "technical insurance
information" respectively. This factor seems to
describe the kind of person to whom men turn for
technical insurance information.
Factor B, sociability and rampathy, has
loadings on item 3, "choice for house guest," and
item 6, "the most sympathetic man in the district."
Factor C, aggressive salesu.mnship, has high
loadings on items 1, 2, and 5. Item refers to
"choices for an assistant for a day on the debit;"
item 2 refers to "someone to present a new sales
plan;" and item 5 is the selection of the "most
aggressive man in the district." This factor
seems to be the stereotyped aggressive salesman.

These three factors also bear a very strong resemblance

to the findings of a number of other factor analytic studies

in which the members of small groups rate and choose each

other on a wide variety of descriptive criteria, or are

assessed by others. Carter (1954) indicates the generality

of the factors in reviewing a series of factor analytic

studies. Carter describes the factors as follows:

Factor I. Individual prminence and achievement -
behaviors of the individual related to his efforts
to stand out from others and individually achieve
various personal goals.
Factor II. Aiding attainmen by. Ithe a rou --
behaviors of the individual related to his efforts
to assist the group in achieving goals toward
which the group is oriented.
Factor III. Sociability -- behaviors of the
individual related to his efforts to establish
and maintain cordial and socially satisfying
relations with other group members.

Both groups of three factors seem to represent under-

lying dimensions in the evaluations persons make of each

other, whether as observers or as fellow group members. But

the important thing to note is that these three factors,

which may be called "activity," "task ability," and

"popularity," have been found in general to be uncorrelated.

It is these three factors, which seem to represent more or

less different functions in the group, that form the

theoretical basis for research in the development of a

differentiated role structure in groups.

C. Basic Assumptions

The researcher who studies group interaction is faced

with data consisting of verbal statements and nonverbal

behaviors of individuals in a face-to-face situation. His

task is to understand what is going on -- to make sense of

the group situation. In this task he may be aided by (1)

some a priori view of the nature of group interaction,

and/or (2) certain procedures for data collection. Both

aspects of group research involve certain preconceptions


about group processes and methods of measurement. We turn

now to a brief summary of the assumptions made in the

present research regarding these elements.

The fundamental assumptions of this experiment may be

stated briefly: (1) the interaction of the small group

constitutes a system i:hlch has properties which to some

degree are regular and predictable, (2) the system can be

seen as a structure made up of roles, (3) the orientations

obtaining in the social system can be experimentally

manipulated and differential effects observed.

In its barest essentials, the fundamental conception of

group processes employed in this research is as follows:

The problem-solving small group faces two basic problems,

the internal or interpersonal problems, and the external or

the task demands made on the group. The internal problem is

not only a means to a solution to the external problem, but

interpersonal relations may become the chief problem,

resulting in the neglect or reinterpretation of the task

problem. In the solution of the external problem, or even

in the balancing of the emphases to be given to external and

internal matters, certain processes are essential, processes

which Bales (1950) has defined as the "functional problems"

of communication, evaluation, decision, tension reduction,

and reintegration. The first three relate to the process of

arriving at a group consensus; the latter three relate to

the social-emotional aspects of coordinated group effort.

The interaction arising from the attempt to solve these two

basic problems tends to be differentiated in terms of time,

of who performs the activity, and the quality of the

activity. This differential patterning of action and

reaction leads to the development of role structure in the


Not only can this patterning of social relationships be

observed directly, but they may also be studied by question-

ing each individual about his perceptions of others in the

social system. Each actor's definitions and expectations of

the other actors are constantly being tested and revised

through interaction. The questionnaire technique enables

the researcher to take the point of view of each actor in

turn. By proper fitting together the answers of all the

participant individuals, the experimenter can obtain

considerable insight into the system of relationships. It

is assumed, then, that each actor's definitions of the other

actors can be partially learned through the proper use of

questionnaires. In the present analysis the group structure

was inferred from consistencies in overt behavior, consensus

in rankings, and sociometric choice.

For the purposes of this experiment, questions con-

cerning the relative complexity, clarity, and permanence of

role-differentiation which may be measured, or considered

to take place in small ad hoc problem-solving groups, are

not of great importance. It is assumed that the develop-

ment of group structure can be caught in the making from

some minimal level, and that these observations will serve

as clues to the general forms of role differentiation and

development of group structure on a more complex level.

D. Hypotheses

In this research the initial formulation was not

planned around tight models of the hypothetical-deductive

variety, but was more empirically oriented, seeking to

discover and explore some variables assumed to be important

for the development of group structure. Since there is no

comprehensive theory dealing with the development of group

structure, i.e., no theory that incorporates all the

relevant variables that influence the development of group

structure, it is not possible to predict with precision all

the consequences of the interaction of the experimental

variables and the effect of other factors, such as kind of

task, frequency of meeting, etc. Some of the main effects

may be predicted, but others will have to be discovered

through analysis of the data, and these may obtain only

under the present experimental conditions. For example,

the exact patterning of the kinds of interaction can be

predicted in a gross way only. However, there are some

consequences that can be predicted, based on the results of

previous research and the author's interpretation of the

variables involved.

The basic hypothesis predicts that, if two groups are

differentially oriented, distinctive structures should

appear. The groups will be differentiated by their kind of

interaction and their style of informal leadership. Specific

hypotheses can be formulated, based on this broad generali-


Interaction trends

Hypothesis 1: Interaction-oriented around will
engage a mori Dpositiv social-emotional activity
than tak-oriented glounj.
Hypothesis 2: Tahk-orientad groups will enngae
ia moa negative social-emotional activity than
interaction-oriented aroulA..

These hypotheses are based on the known personality

characteristics of individuals who are members of the

respective groups. Task-oriented individuals tend to be

aggressive and competitive; interaction-oriented individuals

tend to avoid hostility and strive to attain harmony in the


Hypothesis 3: There will be a decrease in the
amount QL overt agreement expressemd in bot task-
oriented and interaction-oriented groups.

This hypothesis indicates that the orientation of group

members will not effect the decrease in agreement. As a

group develops, regardless of orientation, there is less

need to agree overtly because of the increase in solidarity

and accuracy of expectation. Through the building of a

common culture by interaction, agreement will become more

implicit than explicit.

Hypothesis 4: Interaction-oriented wrouaa
wll neasea aI greater amount gf task-oriented
Hypothesis 5: hme total amount of inter-
atila wzill kb greater in nteragtian-oriented
grouDs1 than in task-oriented WrQaus.
These two hypotheses are based on the definition of

interaction-oriented individuals as being persons who derive

their basic satisfaction from their interaction with other

persons. For them it should be supportive to engage in

conversation with others. The higher rate of activity would


also stem from the fact that they are more sensitive to the

stimulation of other individuals. 'hey may not be as

efficient taskwise as task-oriented individuals, thus

needing to interact more concerning the task.

Informal leadership

Hypothesis 6: TAlk functions and social-
emotional functions will be more clearly
separated in interaction-oriented rou.s than In
task-oriented groups.

This hypothesis is based on the assumption that task-

oriented individuals would like those persons best who help

them to attain their basic satisfaction in a group situation,

i.e., task achievement. In the interaction-oriented groups

the person who was the most supportive would be the best

liked. This would not usually be the task leader. There-

fore, these functions would tend to reside in separate




A. Subjects and Groups

'Ihe subjects were 80 students, 40 and 40 women, from

the introductory psychology course at the University of

Florida. They were selected from approximately 600 students

who formed the enrollment for the fall and spring semesters

1961-62. They were selected on the basis of their scores

on the SIT Inventory, a test which discriminates among three

different orientations that may be taken in group situations:

self-orientation, interaction-orientation, and task-orienta-

tion. The SIT Inventory consists of forced-choice questions

about personal preferences, values and projections. It was

constructed by Bass (1961) as a device for screening

populations for samples of these ideal types of orientation.

A report (Bass, 1961) was published that deals with the

construction of the SIT Inventory, its reliability, norms,

and validity. The final form of the test contained twenty-

seven triads. Each triad consists of an incomplete

statement followed by three alternative completions relating

to each of the three orientations. The examinee chooses

which of the three statements he agrees with most, and

the one he agrees with least. The test is scored by giving



a weight of two points for the alternative agreed with most,

one point for the alternative left blank, and no points for

the alternative agreed with least. The score on the three

scales is derived by summing the weights given to the items

corresponding to the three types of orientation. as who had

extreme scores in the high end of the distribution on the

interaction- and task-orientation scales were placed in

homogeneous groups according to their orientation pattern.

In this study. the category of self-orientation was not

included. There were eight groups in each orientation

condition and five Ss to a group. Four groups in each

condition were composed of men, the other four being

composed of women. The groups met twice a week for three

weeks, making a total of six meetings.

Due to the difficulties of obtaining and scheduling

subjects, an attempt to equalize the intervals between

meetings was abandoned. Casual observation of the behavior

of the groups did not reveal any great difference between

groups with differing time intervals between meetings.

Since most of the differences in the time intervals between

meetings were not more than a day, it is assumed that these

differences were negligible in their effects on the inter-

action and leadership structure of the groups.

B. Task

as were required to discuss a human relations problem

and come to a group consensus in order to answer several

suggested questions pertaining to the problem. The problem

was taken from a book of case histories of human relations

problems in business (Glover and Hower, 1957). The problem

and the task instructions may be referred to in Appendix D.

A single problem was used for all six sessions, instead of

one for each session, in order to observe the interaction

pattern of groups in solving a single problem over several


In order to keep the group stimulated to discuss the

problem additional information pertaining to the problem

was given to the As each time they met. This additional

information was classified into three categories: employer-

employee relationships (E), personality characteristics of

some of the prominent characters involved in the case (P),

and descriptions of two incidents in the story (I). Two

different orders of presentation of the additional material

were arranged so that the effect of order and type of

material on the interaction pattern of the groups could be

evaluated. In Order 1 the material was presented in PIE

sequence. The sequence was EPI for Order 2. This additional

information may also be referred to in Appendix D.

The basic information sheets and the additional informa-

tion sheets were mimeographed and distributed to Se each

session. Scratch paper was provided each S for keeping

notes on the proceedings. Both the problem sheets and the

note sheets were taken up by K each session. A questionnaire

regarding various aspects of the group process was given to

.s after each session.

C. Dependent Variable Measurements

Interaction trends

Bales' interaction process analysis categories (1950)

were used as a means of analyzing the interaction structure

of the groups. Dales has developed a set of observational

categories which represent a systematic set of carefully

defined concepts which can be used for observing and

analyzing the interaction of any kind of face-to-face group.

Each unit of observed behavior, which is the smallest

discriminable segment of verbal or nonverbal behavior to

which the observer can assign a classification, is classified

into one of a set of twelve categories. This set includes

behavior of four types: Positive affect: showing solidarity,

tension release, and agreement; task questions: asks for

direction, orientation, and opinion; task answers: gives

direction, orientation, and opinion; and negative affect:

disagrees, shows tension, shows antagonism. These may be

seen in Table 1.

The analysis did not include individual to individual

communication, but rather the interaction of the group as a

whole. To observe the groups, E sat behind an inclined

bench about eight feet from the Ss at the side of the room.

as sat around a round table about five feet in diameter.

This table was used to control the effect of position on the

leadership structure of the group. In order to control

observational bias by E, the two kinds of groups were coded

by a person other than E. The groups were then scheduled

according to the code. Thus, h was not aware of the

Table 1

Bales' Interaction Process Analysis Categories


A: Positive Affect

B: Giving Answers

C: Asking Questions

D: Negative Affect


1. Shows solidarity, raises other's
status, gives help, reward

2. Shows tension release, jokes,
laughs, shows satisfaction

3. Agrees, shows passive acceptance,
understands, concurs, complies

4. Gives suggestion, direction,
implying autonomy for other

5. Gives opinion, evaluation, analy-
sis, expresses feeling, wish

6. Gives orientation, information,
repeats, clarifies, confirms

7. Asks for orientation, informa-
tion, repetition, confirmation

8. Asks for opinion, evaluation,
analysis, expression of feeling

9. Asks for suggestion, direction,
possible ways of action

10. Disagrees, shows passive rejec-
tion, formality, witholds help

11. Shows tension, asks for help,
withdraws "Out of Field"

12. Shows antagonism, deflates
other's status, defends or
asserts self

orientation category of the groups being observed.

Informal leadership

Two measurements of the informal leadership in the

group were made: role differentiation, or the extent to

which group members distinguished between different kinds of

favorable evaluations, and status consensus, or the extent

to which group members agreed on the evaluations. Questions

were asked requiring the group members to rank each other's

ideas, guidance ability, and amount of participation. as

were also asked to choose the person they liked best and the

one they liked least.

Simple measures of the degree to which these rankings

may tap different aspects of role differentiation were

obtained by asking the following questions: How many times

is there found a rank-one person on one of the ranked

characteristics who is top man on that characteristic only?

How many times is there found a rank-one man on either task

ability or guidance ability who is also ranked first on

popularity? Are the rankings on the four characteristics

correlated or uncorrelated?

The measure used to represent status consensus, or the

amount of agreement on a given set of rankings of members

of each other, is based on Kendall's "Coefficient of

Concordance" which he calls "W." It is obtained from a

matrix of rankings, each individual (placed in vertical

order on a series of rows) ranking each individual in the

group (placed in horizontal order on a series of columns).

The formula is as follows:

W 12 S
m~ (n) -n)

S equals the sum of squares of the deviation of the column

totals from the grand mean, and n1 equals the number of

individuals ranked by m observers. In the rankings here

n1L, since everyone in the group ranks everyone else

including himself. When agreement is perfect, W is equal to

1.00; when there is no agreement, W is equal to .00.

Bole and decision satisfaction

Measurements of the degree of satisfaction S. had with

the role they played in the discussion and with the decisions

made in the group were made by having each a rate his

satisfaction on each of these factors on a seven-point

scale, which ranged from very greatly satisfied to very

greatly dissatisfied.



A. Interaction Trends

The interaction patterning of the groups was observed

by using Bales' interaction process categories. This

systematic set of twelve categories is shown in Table 1.

The twelve separate categories are combined further into

four basic classifications of interaction: A, positive

affect; B, giving answers; C, asking questions; D, negative

affect. The scores used for analysis represent the total

number of acts classified in each of the separate and

combined categories for each group per session. The effects

were evaluated by an analysis of variance.

Social-emotional trends

Positive affect.--Table 2 shows the total number of

acts for the four combined categories for both orientation

groups. The F values for the combined categories are

reported in Table 3. Although the interaction-oriented

groups (I) did express more positive affect than task-

oriented groups (T), the difference did not reach statistical

significance. Thus, Hypothesis 1 is not supported by the

data. It may be seen also that there was a slight increase

in positive affect over the six sessions. The first two


Table 2
Total Acts Per Session for Combined Categories

Categories Orienta- Sessions
tion 1 2 3 5 6

A: Positive Affect I 1214 1225 1374 1266 1428 1353
T 1148 1127 1370 1325 1248 1554
B: Giving Answers I 3940 4099 4039 4315 4169 4273
T 3716 3774 3710 4096 4175 4563

C: Asking Questions I 301 314 280 280 337 342
T 308 304 302 237 320 333

D: Negative Affect I 75 113 123 151 143 152
T 177 243 193 256 171 299

Table 3
F Values for Combined Categories

Sources of Variance Categoric.

Orientation .00 .27 .05 6.38*
Sex 2.64 .00 .92 .00
Order .32 3.18 .64 1.70
Sessions 1.86 7.02** 1.75 3.47*
Orient X Sex .03 .42 2.38 .05
Orient X Order .68 .02 1.44 .10
Sex X Order .54 .27 .54 .05
Seas X Orient .69 1.92 .27 1.24
Seas X Order .46 1.85 1.10 3.48*
Sess X Sex 1.55 3.01* .46 .48
Sess X Orient X Sex .68 1.60 .92 1.37
Sess X Orient X Order .52 .78 .53 3.09*
Sess X Sex X Order .93 .45 1.77 1.45
Orient X Sex X Order 1.13 .17 .01 .33
Orient X Sex X Order X Sess 1.15 .38 .15 1.19


sessions were approximately equal. There was a sharp

increase in session 3 followed by a decrease in session 4.

The I groups showed a sharp increase in session 5 followed

by a decline, while the T groups showed a decrease in

session 5 followed by a sharp increase in session 6.

Data for selected separate categories are reported in

Table 4. The F values for these categories are indicated

in Table 5. There appears to be very little difference

between the I and T groups in regard to the amount of tension

release expressed. It may be observed, however, that there

is a statistically significant amount of change over the six

sessions. The amount of tension release increases very

rapidly for the first three sessions, levels off some

through the fourth session, followed by a sharp increase in

both groups in sessions 5 and 6.

It may also be noted that there is only a small differ-

ence between the I and T groups with regard to the amount of

agreement overtly expressed. However, the I groups did tend

to express more agreement in the first session than the T

groups. It was predicted that there would be a decrease in

the absolute magnitude of agreement expressed over the six

sessions. Even though there was a decrease over time in the

total number of acts of agreement, the results are not

statistically significant. However, the F value obtained

barely missed the value for the .05 level. It may be

observed in Table 4 that three of the five categories show

an increase in the number of acts over the six sessions.

Two of these, tension release and gives opinion, show a

Thble 4
Total Number of Acts Per Session for
Selected Separate Categories

Categories Orien- Sessions
station 1 2 3 4 5 6

2. Tension Be- I 272 423 644 526 732 715
lease T 306 352 564 483 657 723

3 Agrees I 903 776 680 703 662 668
T 796 726 740 683 726 735

5. Gives Opinion I 2734 3026 3003 3225 3118 3248
T 2597 2780 2713 3020 3123 3310

6. Gives Orienta- I 1170 1024 1030 1059 1002 983
tion T 1022 949 959 1034 1033 1205
10. Disagrees I 51 91 83 87 104 77
T 131 131 119 153 119 178

Table 5
F Values for Separate Categories

Sources of Variance
2 3 5 6 lo

Orientation .03 .00 .44 .01 4.51
Sex 2.34 1.61 .26 .76 .36
Order .01 2.85 2.82 .80 .53
Sessions 4.42** 2.22 7.18** .95 .91
Orient X Sex .07 1.41 .51 .11 .55
Orient X Order .48 .44 .32 .93 1.04
Sex X Order .43 .50 .02 1.25 .12
Sess X Orient .59 .79 .72 1.71 1.35
Seas X Order .47 1.29 2.06 1.73 .23
Sess X Sex 1.44 .77 1.62 .96 .41
Sens X Orient X Sex .50 .92 1.22 .40 .74
Seas X Orient X Order .71 1.81 2.02 1.08 1.51
Sess X Sex X Order .22 .34 .29 .62 .33
Orient X Sex X Order .00 9.84* .10 .20 .02
Orient X Sex X Order X Sess 1.64 .17 .38 .51 3.16*



significant increase. Although these results do not support

Hypothesis 3 specifically they do suggest that overt agree-

ment became less and less important over the six sessions.

Negative affect.--Table 3 indicates that there was a

significant difference between the I and T groups in regard

to the amount of negative affect manifested in the group

process. The T groups engaged in much more negative emotional

interaction than the I groups. This confirms Hypothesis 2.

Along with the rise in positive affect there was a

concomitant rise in negative affect. This increase over

time was statistically significant. It may be seen in

Table 2 that sessions 2, 4, and 6 had the greatest amount of

negative affect in the T groups. Both groups showed a sharp

rise in the second session. For the I groups there was a

steady increase in negative emotional interaction, while the

T groups showed periods of decline following the sessions of

a large amount of negative emotion.

Table 6 shows the amount of negative affect in those

groups having different orders of presentation of the

additional information. In the first two sessions those

groups receiving the information in Order 2 (EPI) had the

greatest amount of negative affect. From session 3 to

session 6 those groups receiving the information in Order 1

(PIE) engaged in more negative emotional interaction. From

Table 3 it can be observed that this constituted a statis-

tically significant sessions X order interaction. Further

analysis revealed that it was basically the T groups that

Table 6

Total Number of Negative Emotional Acts in Those
Groups Having Different Orders of
Presentation of Information

1 2 3 4 5 6

1 101 167 205 272 189 265

2 151 189 111 135 125 186

contributed to this interaction effect. This resulted in a

significant sessions X order X orientation interaction.

This means that T groups having Order 2 engaged in much

greater negative emotional interaction during the first two

sessions. After session 2 the T groups having Order 1

expressed more negative affect. I groups having Order 1

expressed more negative affect in all six sessions, although

this difference at times was slight.

Although the T groups expressed more disagreement, the

difference for this separate category failed to reach

statistical significance. Disagreement constituted most of

the negative affect expressed by the I groups. This means

that the negative affect that was expressed in the I groups

was limited mainly to less emotionally charged ideational

conflicts. The effect of the different orders of additional

material is demonstrated in a significant sessions X order

interaction, shown in Table 5. Initially, those groups

having Order 2 expressed more disagreement than those having

Order 1. From the third session on those groups having

Order 1 expressed more disagreement. However, in this case


there was not a significant sessions X order X orientation


Task trends

The data for the combined task-oriented categories are

presented in Table 2. It was predicted that the I groups

would engage in more task-oriented interaction, especially

in category B, "giving answers," and in the separate

category "giving opinion." The results do not show signifi-

cant differences between the two orientation groups. 'Thus,

Hypothesis 4 is not supported by the data. However, the

trends are in the direction expected. For the first four

sessions the I groups do engage in slightly more task-

oriented interaction than the T groups. This is true for

both the combined category B and separate category "giving

opinion." The data do reveal a significant increase in

task-oriented acts over time. This may be noted in Tables

3 and 5.

There is only a very small difference between the I and

T groups in the combined category C, "asking questions."

The amount of interaction in this category tended to remain

stable over time. The greatest change occurred between

session 4 and session 5. In session 5 there was a rather

sharp increase in this type of interaction in both orienta-

tion groups. This increase continued into session 6.

In contrast to the increase in "giving opinion," there

was a decrease in the number of acts in the "giving orienta-

tion" category. The I groups decreased over the total six

sessions, while the T groups decreased for the first three

sessions and then showed a slight increase. This means that

orientation to the specific task assumed less importance in

the problem-solving process over time.

Total activity

In Table 7 the total amount of activity for each

session is reported. Analysis of the data revealed a sig-

nificant increase in the total amount of activity over time.

It may also be observed that for the first three sessions

the I groups were more active than the T groups. This

difference is not statistically significant. Even though

the trends are in the expected direction, the results do not

support Hypothesis 5.

'Thble 7

Total Amount of Activity for Each Session

Orientation Sessions
1 2 3 4 5 6

I 5530 5771 5871 6012 6077 6263

T 5349 5448 5575 5914 5914 6649

Total 10,839 11,219 11,446 11,926 11,991 12,912

B. Role Differentiation

Subjects in this experiment were ranked in four differ-

ent ways for each session. From a post-meeting questionnaire

it was possible to rank the individual on the perceived

quality of his ideas, his perceived ability to guide the

discussion, his amount of participation, and how well he was

liked. The interest in role differentiation stems from the

relationships of these rank orders to each other. For

purposes of communication these rank orders are referred to

as Talking, Ideas, Guidance, and Liking. They represent the

two basic aspects of group life: task and social-emotional.

Role differentiation is defined in terms of the separation

of these functions. That is, role differentiation is said

to occur if the task and social-emotional functions are

performed by two separate individuals.

In order to see whether or not these functions were

separated in the groups, a count was made to see how many

times out of the total number of 48 sessions (8 groups times

6 sessions) that a person who was ranked first on one

characteristic (e.g., Talking) was not ranked first on any

of the other three characteristics. These data are shown in

Table 8. It is clearly revealed that there is a separation

of the rankings on Liking from the rankings on the other

Table 8

The Number of Cases Out of a Total of 48 Sessions
in Which an Individual Held Top Rank on One
and Only One of Four Characteristics

Characteristic Orientation Combined

Talking 7 17 24

Ideas 13 6 19

Guidance 7 4 11

Liking 30 14 44

three measured characteristics. Using a Chi-square measure

of significance, it was found that the probability that

these results occurred by chance is less than .01 for the

T groups, less than .001 for the I groups, and less than

.001 for the groups combined. It may be further noted that

a greater separation of function occurred in the I groups.

This means that in the I groups the characteristic of Liking

achieved isolated prominence to a greater degree than in the

T groups. This supports Hypothesis 6. The factor of Talking

also achieved some isolated prominence in the T groups,

slightly more so than Liking. In the I groups the Idea

characteristic seems to be separated some from the

characteristics of Thlking and Guidance.

Further information may be obtained by changing the

question and asking how many times does the same person hold

top rank on two characteristics simultaneously? In Table 9,

the number of groups per session in which the same person

held top rank on both the Idea characteristic and the Liking

characteristic are reported for both orientation groups.

It may be seen that these functions tend to be separated

more in the I groups than in the T groups. Table 10 indicates

the number of groups out of eight in which the same person

held top rank on both the Guidance characteristic and the

Liking characteristic. This shows even more clearly the

greater separation of task and social-emotional functions in

the I groups.

The above techniques for determining the amount of

specialization among the various characteristics deal only

Table 9

Number of Groups Out of Eight in Which the Same
Individual Held Top Rank on Both the Idea
and Liking Characteristic Per Session

Orientation 4 5
1 2 3 4 5 6

T 3 5 6 2 8 5

I 3 4 2 3 2 1

Thble 10

Number of Groups Out of Eight in Which the Same Individual
Held Top Rank on Both the Guidance and the
Liking Characteristic Per Session

Orientation Sessions
1 2 3 4 5 6

T 3 5 6 5 4 5

I 0 2 0 1 1 2

with the top position on each characteristic. In order to

include the other ranks in the analysis, each individual's

mean rank for all six sessions for each of the characterin-

tics was obtained. Then a rank-order correlation was

computed for all pairs of characteristics. These eight

rank-order coefficients for each set of orientation groups

were transformed into z scores, averaged, and then trans-

formed back into rank-order coefficients. These correlations

are shown in Table 11. It may be seen, first of all, that

there is a higher correlation among the characteristics of

Talking, Guidance and Ideas than between Liking and either

Table 11

Mean Rank-Order Correlations Among Rankings on
Talking, Ideas, Guidance and Liking

Task-Oriented Groups


Talking .87 .90 .60

Ideas .91 .74

Guidance .98

Interaction-Oriented Groups


Talking .71 .92 .24

Ideas .86 .59

Guidance .29

Talking and Ideas. Secondly, a striking difference between

the I and T groups may be observed. There is a sharp

contrast between the two orientation groups in regard to the

correlation between Liking and Guidance. In the T groups

Liking is strongly associated with Guidance. For the I

groups there is very little association between being liked

and guiding the group. This is still further evidence that

there is a greater separation between the task and social-

emotional functions in the I groups. There is very little

difference between the two orientation groups as to the

interrelation of the various task characteristics. Talking,

Ideas, and Guidance are highly correlated in both groups.

It is interesting to note also the reasonably high

correlation between Liking and Ideas in the I groups.

Apparently, even for the I groups, the contribution of ideas

to the group discussion is perceived to be of value.

If the number of positive sociometric choices accruing

to individuals of the various task-related characteristics

are counted, some picture of the relationship between how

well the person performs and how well the other group

members like him can be obtained. For each session all

persons were listed in rank order for the characteristics of

Talking, Ideas, and Guidance. The number of choices for

being the best liked person were recorded for each rank.

Then the number of choices for all six sessions for all

rank-one individuals on each characteristic were pooled, and

so for all rank-two individuals, and so on for the five

ranks. This total number of choices was then averaged for

each characteristic for each set of differently oriented

groups. No distinction was made as to which meetings in the

series of six were represented. The identity of the

individuals was not preserved from meeting to meeting. The

fact that John Smith might have been rank-one person in the

first meeting, rank-two person in the second, and so on, was

ignored. The data are presented in Figures 1, 2, and 3. As

will be seen in Figure 1, the most talkative individual (a

composite fiction, it will be remembered) is not the most

popular man in either of the orientation groups. For the T

groups the rank-two person is best liked; for the I groups,

rank-three person. When sociometric choice is related to

rankings on Ideas, a further difference between the I and T


groups may be seen (Figure 2). In the T groups ranks 1-3

are clearly differentiated as to liking. In the I groups

there is very little discrimination among the first three

ranks. In fact, there is much less discrimination among all

the ranks in the I groups as compared to the T groups. This

may also be seen in Figure 3 in which ranking on Guidance is

associated with liking. It would appear that there is here

a very interesting phenomenon which seems to be the conse-

quence of the two types of orientation toward groups. Not

only is there less differentiation among the group members

in the I groups, but the range is more restricted. That is,


o 1

0 9


o 7





e 4


- -




1 2 3 4 5
Fig. 1.--Relation of sociometric choice to rank on



u 12


a) 10

.. 9
0 8

> 7

0 5

o 4




o 12
F 11


v) 8



a) 5


1 2 3 4 5
Fig. 2.--Relation of sociometric choice to rank on

1 2 3 4 5
Fig. 3.--Relation of sociometric choice to rank on




the perceived similarity between the first-ranked individual

and the last-ranked individual is greater in the I groups.

C. Status Consensus

Table 12 reports the mean W (Coefficient of Concordance)

for the Idea and Guidance rankings for both orientation

groups for each session. There appears to be very little

difference between the two kinds of groups with regard to

the amount of agreement on leadership rankings. However,

one interesting observation can be made. The T groups show

a decrease in amount of agreement in sessions 2, 4, and 6.

This particularly true for the Guidance rankings. In these

same sessions the I groups show an increase in amount of

agreement. These sessions are the ones in which different

classification of additional information was given to the

group. 'hey are also the ones in which the greatest amount

of negative affect was expressed.

Table 12

Mean W for Guidance and Idea Rankings
Per Session

Rankings Orien- Sessions
tation 1 2 3 4 5 6

Guidance T .62 .53 .66 .54 .67 .59
I .68 .76 .60 .76 .66 .69

Ideas T .59 .55 .67 .47 .56 .45
I .60 .52 .50 .45 .58 .69

Combined T .61 .54 .67 .51 .61 .52
I .64 .64 .55 .61 .62 .69

D. Role and Decision Satisfaction

After each session each group member was asked to rate

his satisfaction with the role he played in the group during

that session and with the decisions that were made. The

ratings were made on a seven-point scale. The results are

presented in Table 13. The differences between the I and T

groups are not statistically significant. However, it is

interesting to note that the I groups on the average rated

themselves as being more satisfied with their role and with

the decisions made than the T group members in the first two

sessions. Initially the I group members seemed to be more

satisfied with the group process. Analysis of the data

revealed that there was a statistically significant decrease

in the amount of satisfaction both with the role played and

with the decisions made over the six sessions in both

orientation groups.

Table 13

Mean Ratings for Role and Decision Satisfaction

Rating Orien-sons
tainn 1 2 3 4 5 6

Role T 4.6 4.4 4.6 4.6 4.3 4.3
I 5.2 5.0 4.7 4.8 4.6 4.1
Decision T 4.6 4.4 4.3 4.7 4.4 4.3
I 5.2 4.9 4.6 4.6 4.6 4.0



Against this background of findings we may now attempt

an interpretive summary. The evidence suggests that member

orientation to the group is an important factor in the

development of the structure of a group. That is, the data

indicate that, if members of two groups perceive the group

differently, distinctive interaction and role structures

emerge. However, the data also reveal some striking

similarities as well. These similarities suggest that there

may be underlying developmental patterns common to most

small problem-solving groups regardless of the orientation

of the members.

A. Interaction Trends

The only statistical significant difference between the

I and T groups with regard to the interaction structure was

that the T groups expressed more negative affect. The

members of these groups tended to argue, disagree, show

antagonism, and become defensive to a greater degree than

I group members. This was to be expected on the basls of

the personality characteristics associated wit thiAs

particular orientation. However, the data su QPt that the



difference in personality characteristics alone is not the

only factor determining this difference. The nature of the

task also was a contributing factor to the response of the

T group members. It will be remembered that T individuals

were described as persons who derive their basic satisfaction

from progress in task solution. The task given the group

members in this experiment was ambiguous to some extent,

i.e., the path to the goal was not clear. Even though .s

were instructed to make inferences from the information

given, T group members were reluctant to do so and attempted

to document conclusively every interpretation drawn from the

material. The differential effect of the task may be seen

in the significant sessions X orientation X order inter-

action. In the first two sessions those T groups having the

additional information given in Order 2 showed the most

negative affect. Beginning with seEsion 3 this was reversed.

Those T groups having the additional information in Order 1

expressed more negative emotion. The I groups having

Order 1 expressed more negative affect than those having

Order 2 in all six sessions. In addition, the difference

between Order 1 and Order 2 in the T groups was greater than

this difference within the I groups. A plausible explanation

for this may lie in the fact that the information given in

Order 1 ranged from specific information to more general

information. Hence those groups receiving information in

Order 1 perceived that they were getting less and less task-

relevant information. This was particularly an important

factor for the T groups.

Further evidence concerning the effect of the task on

the T groups is found in the fact that there was a sharp

increase in negative affect in sessions 2, 4, and 6. These

were the sessions in which a new classification of material

was given to the groups. This effect also occurred in the

I groups, showing that this response was not completely due

to the T group members. However, the T group members

demonstrated a greater increase in negative emotion in these

sessions. It seems that the introduction of new material

causes some conflict of interpretation, etc., thus increasing

the negative emotional interaction. For the T groups the

amount of agreement on leadership rankings decreased in

sessions 2, 4, and 6. The decrease in agreement on leader-

ship rankings and the increase in negative affect in

sessions 2, 4, and 6 indicate that the task partially con-

tributed to the differences between the orientation groups

in the amount of negative affect displayed. That is, the

personality characteristics of the T group members inter-

acted with the particular kind of task to produce the

differences observed.

It was predicted that the I group members would engage

in more positive emotional interaction than the T groups.

The results indicated that there was no significant differ-

ences between the two kinds of orientation groups in this

eategery. However, initially the I groups did display

slightly more positive emotion. As the groups progressed

over time this difference became smaller. This may be

explained on the basis of Bales' (1955) equilibrium theory.


According to this theory a social system will survive only

if a steady state is maintained in the system. Any dis-

turbances introduced into the system tend to create forces

that establish equilibrium within the system. For example,

it might be argued that negative reactions in a group

introduce a disturbance and reduce the satisfaction level

of the members. Positive reactions are then needed to

maintain the steady state of the group. Bales assumes that

a preponderance of positive reactions over negative reactions

is a condition of maintenance of the steady state. If this

be true, then the greater amount of negative reactions in

the T groups would result in comparatively more positive

reactions in the latter half of the six sessions. This

would balance the comparatively more positive reactions of

the I groups in the first half of the six sessions.

In addition to these differences there are some

striking similarities between the I and T groups as to their

interaction development. In general both orientation groups

initially began with a heavier emphasis on the task-oriented

types of interaction, with inhibition of the more affective

types of interaction. In session 2 there was a sharp

increase in negative reactions but positive reactions

remained steady. In session 3 the trend toward greater

affect relative to the task-oriented interaction continues

in both orientation groups. However, in session 4 there is

a slight increase in task-oriented interaction.

Two kinds of trends are evident in the area of positive

reactions. Overt showing of agreement shows a downward

trend, while tension release shows an upward trend. That

overt sho.:ing of agreement would decrease over time is not

obvious to one who is not familiar with interaction trends

in groups. This downward trend in overt agreement may be

explained by the fact that when individuals interact for the

first time in a group they do not know what response to

expect from one another. It can also be assumed that there

is some motivation for insuring and maintaining the solid-

arity of the group. Expression of agreement in the face of

unstructured relationships may be an attempt to maintain

solidarity. As the group members interact over time they

become more accurate in their expectations of the behavior

of the other members and therefore feel freer to disagree.

There also develops a common culture in which agreement may

become more implicit than explicit. As the group develops

in the process of solving the problem, and more information

about the problem is known, there is a greater basis for

disagreement. More time is needed in the later sessions

than in the earlier sessions to formulate and build arguments

for various points of view. Solidarity comes to be expressed

in other ways, e.g., tension release, Joking, and praise.

All of these factors converge to produce an interaction

structure in which overt expression of agreement assumes

less and less importance.

There are also two kinds of trends in the task-oriented

categories. "Giving opinion" shows a significant increase

over the six sessions with a concomitant decrease in "giving

orientation." The fact that the present task was somewhat


ambiguous may have forced the group members to persist in

trying to solve the problem. This would necessitate the

contribution of more ideas and the evaluation of the ideas

presented. As the group members became more familiar with

the task, the giving of orientation became less important

relative to other kinds of interaction. In general all of

these trends are consistent with the findings of Heinicke

and Bales (1952). This provides some evidence that there

is an underlying developmental interaction trend regardless

of member orientation. That is, there seems to be an under-

lying pattern of development on which differences due to

member orientation are imposed.

Contrary to common sense expectations, the I group

members engaged in more task-oriented interaction and

manifested more total activity. This is particularly true

for the first three sessions. It is possible that, ir-

respective of manifest content, social interaction is of a

social-emontional nature. If one seeks information and

receives a response, that might indicate to him that his

relationship with others is supporting him against a

demanding environment. Perhaps only after exploration, when

one is more confident of a response, will a trusting

relationship develop. It also might be suggested that an

interaction-oriented, socially sensitive person might be

more responsive to the stimulation of other group members.

B. Role Differentiation

Both I and T groups show a separation of task functions

from social-emotional functions. Slater (1955) explains

this divorcing of the task functions from the social-

emotional functions by two basic kinds of reasons:

sociological and psychological. The sociological reasons

revolve around the incompatibility of the task and social-

emotional roles. The group member who presses for the

solution, contributes ideas, and attempts to direct the

problem-solving process forces the other members to make

continual minor adjustments in their behavior and evaluations

of their ideas and values. The individual who concerns

himself with the social-emotional aspects of the group

process, on the other hand, tends to be supportive in his

responsive to the ideas and behavior of the other group

members. It is assumed that the latter behavior is more

appealing to the group members when they are asked to indi-

cate whom they like best. The task specialist will not

necessarily be disliked, but his task emphasis will tend to

arouse some negative feelings. These feelings may not be

expressed overtly, but they may enter into the evaluations of

the group members so as to neutralize any positive feelings

for him.

The second set of reasons is called psychological.

These deal with the individuals' predispositions to assume a

particular role. That is, some persons may have a "need" to

avoid conflict. They may avoid task activity because of the

threat that this activity may hold for them. On the other

hand, the person who emphasizes the task functions may seek

to avoid involvement in interpersonal relationships to any

depth. Needs to express hostility may be expressed in

aggressive problem-solving behavior.

Bales and Slater (1955) have also related another

factor to the development of role differentiation In groups.

Tils is the factor of status-consensus. They have found in

general that those groups who end their first meeting with

low status-consensus (as measured by Kendall's W) usually

develop sharply differentiated role structure. That is, in

groups with low status-consensus the task functions and

social-emotional functions are performed by separate

individuals. Conversely, in high status-consensus groups

there is only a moderate amount of separation of function.

Sometimes it was found in high status-consensus groups that

there was no differentiation, i.e., there was a one-to-one

correlation with rankings on task activities and popularity.

However, this was found to be a rare phenomenon.

If the results of this study are valid, then the above

description and explanation will have to be modified. There

is no indication in the present research that there is a

relation between orientation, as defined here, and status-

consensus. The amount of status-consensus was high for both

orientation groups. Therefore, low status-consensus is not

necessarily the only factor related to sharp role-differ-

entiation, and high status-consensus does not guarantee lack

of differentiation of roles. The evidence also indicates

that a high correlation between Liking and either Ideas and

Guidance may not necessarily be as rare as has been sug-

gested. The task and social-emotional functions are not

necessarily incompatible. They may be incompatible only in

some kinds of groups, e.g., randomly assembled groups. In

certain kinds of composed groups there may be a high

correlation between rankings on task ability and rankings on

popularity, e.g., task-oriented groups.

It is at this point that the effect of the differences

in orientation on the development of role structure may be

observed. The particular personality characteristics

associated with task-orientation, coupled with the ambiguous

task, seemed to produce a different set of values for the T

group members. It will be remembered that they find their

greatest satisfaction in progress toward the solution of the

problem. On this basis it seems reasonable to assume that

those members who made the most contribution to the solution

of the problem, and thus reduced tension and gained satis-

faction for the other group members, would tend to be the

best liked persons. Thus there would be less separation of

the task and social-emotional functions. This is demonstrated

in the fact that there is almost a one-to-one correlation

between the Liking characteristic and the Guidance character-

istic. The best liked person was the one who guided the

group through the discussion. This is true to a lesser

extent for the Idea characteristic.

In the I groups, on the other hand, task functions have

a low association with the Liking characteristic, i.e., the

task functions and social-emotional functions tend to be

performed by separate individuals. It will be remembered


that I group members are oriented basically toward the group

process. Their behavior is determined by inner needs such

as the need to be loved and accepted, and a need to avoid

conflict, disharmony, and hostility (Bass, 1961). Thus, the

members who satisfy this need would tend to be the best

liked. Even though I group members gain satisfaction from

the interaction itself, they still need someone to con-

tribute to the tash solution. They still need to make some

progress in the solution of the problem if the group is to

satisfy the requirements of the experimenter. Ihus the task

specialist tends to be someone other than the social-

emotional leader.

It may be concluded, then, that personality factors

enter into the matrix of factors which determine the develop-

ment of role structure. There is not always a sharp differ-

entiation of task and social-emotional functions. At times

there may be a high correlation between them, depending on

the orientation of the group members. Another implication

of the findings is that the configuration of personality

factors operating to produce interaction orientation are not

the same as those producing low status-consensus, even

though both resulted in sharp role differentiation. Some

personality factors may lead to a sharp separation of task

and social-emotional functions without interfering with the

agreement of the group members on rankings of task ability.

This is an area that needs further experimental exploration.

Apparently different groups emphasize task and social-

emotional functions in varying proportions and the

distinctiveness of the separation of these functions

depends on the weight given to each.

The differences in orientation also produced another

interesting phenomenon. It was observed in the I groups

that there was generally less discrimination among the group

members on the Liking characteristic. The range between the

best liked and the least liked was comparatively smaller

than in the T groups.

The meaning of this tendency to make undifferentiated

ratings on Liking has been explained by Slater (1955) on the

basis of its relationship to scores on the California

F-Scale. He found that high F-scores were found more fre-

quently among members of low status-consensus groups. High

F-scores were made also by those ranked high on Liking and

comparatively low scores were made by those persons ranked

high on either Ideas or Guidance. Both high F-scores and

undifferentiated ratings were interpreted as reflecting a

tendency toward a rigid and oversimplified apporach toward

interpersonal relationships. Fine perceptual discrimination

are not expected to be made by those who score high on the


This phenomenon may also be related to the work of

Fiedler (1954) on psychological distance in interpersonal

relations. As a measure of psychological distance he used

an index which he calls the Assumed Similarity between

opposites, or ASo. An individual is asked to choose persons

who are at the opposite poles of some continuum. Then the


individual is asked to make a judgment about how similar

these Dersons are to each other. Fiedler reports the

results of two studies in which the influence of ASo on

teamwork was investigated. It was found in both studies

that the most preferred man on the best teams (basketball

and surveyor) tended to assume relatively little similarity

between his own preferred and rejected co-workers. He

infers from this that those persons who assume very low

similarity between their best and poorest co-workers are

highly task-oriented. This is consistent with the findings

of Slater (1955) that those individuals ranked high on task

status showed more discrimination among group members on

Liking. The results of the present investigation are also

consistent with this concept. In the task-oriented groups

there is a greater differentiation made between the best

liked and the least liked person.

Fiedler (1953) also found that Assumed Similarity

scores are related to liking. Subjects perceive those whom

they like as more similar than those whom they dislike.

This can be reversed and stated that subjects who show a

greater differentiation among persons in regard to liking

will perceive very little similarity between them. And

conversely, those who show lesser differentiation among

persons on the liking continuum will perceive more

similarity between them. It has been stated previously that

interaction-oriented persons are basically oriented toward

the group process. They are more concerned with the inter-

personal relations in the group than with the task aspects.

On this basis it may be assumed that interaction-oriented

individuals would tend to perceive group members as similar

to avoid the anxiety of rejecting and being rejected. To

admit that one likes some members more than others would

imply that the individual may not be liked by other members.

This would constitute a threat to interaction-oriented

persons. This then t:ould be expressed in a less fine

differentiation among the group members on the Liking




This is an experimental study of small problem-solving

groups designed to show the effects of member orientation

upon the development of group interaction and role structure.

It was predicted that different interaction and role

structures would develop in groups with differing orienta-

tions. Member orientation was defined as a set of pre-

dispositions which determine how an individual will interact

with other group members in building and maintaining a

pattern of interpersonal behavior. Three different types

of orientations were isolated and discussed: self, inter-

action, and task. A test, the SIT Inventory, which

discriminates among these three orientations, was used to

screen out two of these ideal types of orientation.

Subjects who scored high on either the task or interaction

scale of the SIT Inventory were placed in homogeneous groups

representing each orientation. There were eight groups in

each condition, half of which were composed of men, and half

composed of women.

The group members wore given a human relations problem

to discuss for six 40-minute sessions. The interaction was

observed by Bales' interaction process categories. Following


each meeting a questionnaire was given which was designed to

elicit information about three types of roles: activity,

task ability, and popularity. The analysis of the data was

organized in terms of six main hypotheses, three of which

received substantial support from the data:

Hypothesis 1 Interaction-oriented aruOs aill eneaze
in more Dositive social-emotional activity than task-
oriented srouos.

There was no significant difference between the two

oriented groups in the expression of positive affect. The

hypothesis was not confirmed.

Hypothesis 2 Task-oriented egrouns L enia1 e in morse
negative social-emotional activity than interaction-oriented
arouns .

This prediction was substantiated. There was a

statistically significant difference in the amount of

negative affect expressed in the task-oriented groups.

Hypothesis 3 There &ll hb. a decrease ain the amount
Sovert t gre exp ressed in both task-oriented and
interaction-oriented gXrous.

This prediction was interpreted as being partially

supported. The hypothesis refers to the decrease in

absolute magnitude of agreement. There was a decrease over

time, but this barely missed significance at the .05 level.

However, the significant increase in interaction in other

categories coupled with the decrease in agreement lends some

support to the basic idea inherent in the hypothesis, viz.,

that the importance of overt agreement will decline over


Hypothesis 4 Interaction-oriented rouAs willl sengae
in a arater amount Qf task-oriented interaction.


Hypothesis 5 he& total amount oL interaction. wll 1e
greater in interaction-oriented aroius .han I tansk-oriented

there was no significant differences between the two

orientation groups with respect to the amount of task-

oriented interaction and total activity. However, the

differences were in the expected direction.

Hypothesis 6 Tfsk functions and social-emotional
functions will be more clearly separated JI interaction-
oriented groups than In task-oriented groups.

This prediction was supported by the data. The inter-

action-oriented groups demonstrated a sharper differentiation

of task and social-emotional functions than the task-oriented


In summarizing the results it may be seen that there

are both similarities and differences between the two types

of orientation groups. The interaction-oriented groups

present a picture of being more positive in emotional tone.

Initially they engaged in slightly more task-oriented

interaction and showed more gross activity, and seemed to be

a little more satisfied with the role they played in the

group and with the decisions made. They also demonstrated

a sharper differentiation of role.

On the other hand, the task-oriented groups present a

picture of being more negative in emotional tone. They

seemed to be more concerned about the task itself, which

caused a lot of frustration. Their ratings on role and

decision satisfaction were initially lower than those of the

interaction-oriented groups. However, there was very little

difference in the latter three sessions. The task-oriented

groups showed a much higher correlation between rankings on

task ability and rankings on popularity. Hence there was

less clear separation of the functions.

The basic similarities between the interaction- and

task-oriented groups reside in the development of the

interaction structure. In both types of groups there was a

rise in the social-emotional activity concomitant with an

initial slight decrease in task-oriented activity. The

expression of agreement declined over the six sessions for

both groups. Tension release and giving opinion increased

at a significant rate. In fact, the total activity of both

orientation groups increased significantly over time.

There are several general conclusions that can be drawn

from the results:

(1) It is possible to discriminate among orientations

persons have toward groups by the use of a paper and pencil

test. This experiment provides partial validation for the

SIT Inventory. Persons classified on the basis of this test

will behave in predictable ways when they interact in a


(2) The use of the concept of orientation toward the

group is a fruitful approach to group development. Groups

whose members have differing orientations can be expected to

develop differing interaction and role structures. It is

possible to measure these differences.

(3) The relation of personality to role and interaction


structure in groups constitutes a very large problem area

in which there are probably few simple answers. The

measurements of this study are over-simplified, gross

measurements. Further research is needed to tease out all

the important factors or conditions determining group

development. For example, the influence of various types of

tasks on the developing patterns of groups is not precisely

known. Another example might include the relation between

status-consensus and orientation.

(4) Research is needed to investigate whether or not

the various orientations produced by different determining

conditions have similar consequences. For example, does a

group develop the same when interaction-orientation is

produced by a procedure-oriented leader and when this

orientation is an expression of personality organization?

Similarly, there is a need to study the interaction of

internal and external determining conditions. For example,

would interaction-oriented and task-oriented persons respond

similarly to task-oriented instructions? What differences

would there be?

It is hoped that these results and conclusions will

provide a basis for exploring further the factors that

effect the development of group structure.


Atkinson, J. W., Heyns, R. W., and Veroff, J. The effect of
experimental arousal of the affiliation motive on
thematic apperception. J. abn. soc. Psychol., 1954,
4L, 405-410.

Bales, R, F. Interaction process analysis. Cambridge:
Addison-Wesley Press, Inc., 1950.

Bales, B. F., and Slater, P. E, Role differentiation in
small decision-making groups. In T. Parsons, and
R. F. Bales (eds.), Family, socialization and inter-
action process. Glencoe: The Free Press, 1955.

Bales, R. F., and Strodtbeck, F. L, Phases in group
problem-solving. J. abn. soc. Psychol., 1951, 46,

Bales, R. F. The equilibrium problem in small groups. In
A. P. Hare, E. F. Borgotta, and R. F. Bales (eds.),
Small groups. New York: Alfred A. Knopf, 1955.

Barker, R. G. The social interrelations of strangers and
acquaintances. Soclometry, 1942, j, 169-170.

Barnard, C. I. Functions and pathology of status systems
in formal organizations. In W. F. Whyte, Industry and
society. New York: McGraw-Hill, 1946.

Bass, B. M. Behavior in groups. Contract N7 ONR 35609,
Group Psychology Branch, Office of Naval Research,
Tech. Rpt. 25, January, 1961.

Bavelas, A, A mathematical model for group structures.
Apl. Anthron., 1948, 2, 16-30.

Benne, K. D., and Sheats, B. Functional roles of group
members. J. see. Issues, 1948, 4., 41-49.

Bennis, W. G., and Shepard, H. A. A theory of group
development. Hum. Relat., 1956, 9, 415-437.

Borgatta, E. F., and Bales, R. F. Interaction of individuals
in reconstituted groups. ScLQouetrv, 1953, 16., 302-320.

Carter, L. F. Recording and evaluating the performance of
individuals as members of small groups. Personnel
Psvlhol., 1954, 2, 477-488.

Cartwright, D., and Zander, A. Group dynamics. Evanston:
Row, Peterson and Company, 1953.

Fiedler, F. E. Assumed similarity measures as predictors
of team effectiveness. J. abn. soc. Psychol., 1954,
4, 381-388.

Fiedler, F. E. The psychological-distance dimension in
interpersonal relations. J. of Personality, 1953,
2., 142-150.
Fouriezos, N. T., Hutt, M. L., and Guetzkow, H. Self-
oriented needs in discussion groups. J. abn. soc.
Psyvhol., 1950, 4i, 682-690.

Glover, J. D., and Hower, R. M. The administrator.
Homewood: Richard D. Irwin, Inc., 1957.

Gouldner, A. W. Patterns of industrial bureaucracy.
Glencoe: The Free Press, 1954.

Grusky, O. A case for the theory of familial role differ-
entiation in small groups. Soc. Forces, 1957, 15,

Heinicke, C., and Bales, R. F. Developmental trends in the
structure of small groups. Sociometrv, 1953, 16, 7-38.

Homans, G. C. The human group. New York: Harcourt, Brace
and Company, 1950.

Jennings, H. Leadership and isolation. New York: Longmans,
Green, and Company, 1950.

Kahn, R. L., and Katz, D. Leadership practices in relation
to productivity and morale. In D. Cartwright, and
A. Zander, Group dynamics. Evanston: Row, Peterson,
and Company, 1960.

Lewis, H. B., and Franklin, M. An experimental study of the
role of the ego in work. II. The significance of task-
orientation in work. J. exn. Psychol., 1944, 34,

Lipset, S. M., Trow, M. A., and Coleman, J. S. Union
democracy Glencoe: 'The Free Press, 1956.

Marcus, P. M. Expressive and instrumental groups: toward a
theory of group structure. Amer. J. Sociol., 1960, 66,

McClelland, D. C., Clark, R. A., Roby, T. B., and Atkinson,
J. W. The projective expression of needs. IV. The
effect of the need for achievement on thematic apper-
ception. J. exp. Psychol., 1949, 3., 242-255.

Neiman, L. J. Problem of the concept of role. Social
Forces, 1951, 3., 141-149.

Olmsted, M. S. Orientation and role in the small group.
Amer. social. eev., 1954, la, 741-751.

Parsons, T., and Bales, R. F. Family. socialization and
interaction process. Glencoe: 'Ihe Free Press, 1955.

Philp, H., and Dunphy, D. Developmental trends in small
groups. Sociomfetry, 1959, 22, 162-174.

Sarbin, T. R. Role theory. In G. Lindzey (ed.), Hadbook
of social pyscholvy. Cambridge: Addlson-lesley
Publishing Company, Inc., 1954, Vol. I.

Shipley, T. E., Jr., and Veroff, J. A projective measure
of need for affiliation. J. exp. Psychol., 1952, 41,
Slater, P. E. Role differentiation in small groups. Amer.
sociol. Rev., 1955, 2a, 300-310.

Stock, D., and Thelen, H. A. Emotional dynamics and group
culture. New York: New York University Press, 1958.

Stogdill, R. M. Individual behavior and group achievement.
New York: Oxford University Press, 1959.

Suchman, J. R. Social sensitivity in the small task-oriented
group. J. abn. soc. Psychol., 1956, 52-, 75-83.

Swanson, G. E. A preliminary laboratory study of the acting
crowd. Amer. soclol. Rev., 1953, 1L, 522-533.

Tagiuri, R., Blake, R. R., and Bruner, J. S. Some deter-
minants of the perception of positive and negative
feelings in others. J. abn. soc. Psych1o., 1953, i.,
Thibaut, J. W., and Strickland, L. H. Psychological set and
social conformity. J. Personality, 1956, 2., 115-129.

Wispe, L. G. A sociometric analysis of conflicting role-
expectations. Amer. J. Sociol., 1955, 61, 134-137.

Wolman, B. Motivational objectives of group members.
J. soc. Psvchol., 1956, 43, 11-25.




Table 14
Analysis of Variance for Combined Category A,
Positive Affect

Source df SS MS F P

Orientation 1 81.58 81.58 .00
Sex 1 50,142.04 50,142.04 2.64
Order 1 6,048.37 6,048.37 .32
Sessions 5 14,926.22 2,985.24 1.86
Orient X Sex 1 590.04 590.04 .03
Orient X Order 1 13,020.04 13,020.04 .68
Sex X Order 1 10,343.88 10,343.88 .54
Seas X Orient 5 5,559.54 1,111.91 .69
Seas X Order 5 3,673.49 734.70 .46
Seas X Sex 5 12,461.08 2,492.22 1.55
Sess X Orient X Sex 5 5,441.21 1,088.24 .68
Sess X Orient X Order 5 4,135.95 827.19 .52
Sess X Sex X Order 5 7,465.40 1,493.08 .93
Orient X Sex X Order 1 21,420.37 21,420.37 1.13
Orient X Sex X Order X Sess 5 9,208.13 1,841.63 1.15
Within 8 152,143.69 19,017.96
Error 40 64,182.31 1,604.56
Total 95 380,843.34

Table 15

Analysis of Variance for Combined Category B,
Giving Answers

Source df SS MS F P

Orientation 1 6,683.34 6,683.34 .27
Sex 1 52.51 52.51 .00
Order 1 77,577.51 77,577.51 3.18
Sessions 5 66,109.98 13,221.99 7.02 .01
Orient X Sex 1 10,188.76 10,188.76 .42
Orient X Order 1 372.09 372.09 .02
Sex X Order 1 6,583.59 6,583.59 .27
Sess X Orient 5 18,075.28 3,615.06 1.92
Sess X Order 5 17,397.41 3,479.48 1.85
Sess X Sex 5 28,323.41 5,664.68 3.01 .05
Sess X Orient X Sex 5 15,061.17 3,012.23 1.60
Sees X Orient X Order 5 7,375.09 1,475.02 .78
Sees X Sex X Order 5 4,231.78 846.36 .45
Orient X Sex X Order 1 4,069.01 4,069.01 .17
Orient X Sex X Order X SeEs 5 3,586.37 717.27 .38
Within 8 195,413.30 24,426.66
Error- 40 75.262.20 1.881.56
Total 95 536,362.80

able 16

Analysis of Variance for Combined Category C,
Asking Questions

Source df SS MS F P

Orientation 1 26.04 26.04 .05
Sex 1 513.38 513.38 .92
Order 1 360.38 360.38 .64
Sessions 5 995.71 199.14 1.75
Orient X Sex 1 1,335.04 1,335.04 2.38
Orient X Order 1 805.04 805.04 1.44
Sex X Order 1 301.04 301.04 .54
Sess X Orient 5 152.21 30.44 .27
Sass X Order 5 625.37 125.07 1.10
Sess X Sex 5 262.12 52.42 .46
Sess X Orient X Sex 5 524.45 104.89 .92
Sess X Orient X Order 5 302.70 60.54 .53
Sees X Sex X Order 5 1,006.45 201.29 1.77
Orient X Sex X Order 1 7.04 7.04 .01
Orient X Sex X OrderX Sess 5 83.99 16.80 .15
Within 8 4,479.33 559.92
Error 40 4,542.67 113.57
Total 95 16,322.96

iTble 17

Analysis of Variance for Combined Category D,
Negative Affect

Source df SS MS F P

Orientation 1 3,528.37 3,528.37 6.38 .05
Sex 1 .16 .16 .00
Order 1 953.19 953.19 1.70
Sessions 5 1,596.21 319.24 3.47 .05
Orient X Sex 1 30.38 30.38 .05
Orient X Order 1 54.00 54.00 .10
Sex X Order 1 26.04 26.04 .05
Sess X Orient 5 573.00 114.60 1.24
Sess X Order 5 1,604.68 320.94 3.48 .05
Sess X Sex 5 220.95 44.19 .48
Sess X Orient X Sex 5 628.74 125.75 1.37
Sees X Orient X Order 5 1,424.69 284.30 3.09 .05
Sess X Sex X Order 5 666.83 133.37 1.45
Orient X Sex X Order 1 181.50 181.50 .33
Orient X Sex X Order X Sess 5 549.59 109.92 1.19
Within 8 4,422.17 552.77
Error 40 3.684.93 Q2.12
Totol 95 20,145.33

Table 18
Analysis of Variance for Tension Release

Source df SS MS F P

Orientation 1 404.26 404.26 .03
Sex 1 28,462.59 28,462.59 2.34
Order 1 90.09 90.09 .01
Sessions 5 34,597.18 6,919.44 4.42 .01
Orient X Sex 1 810.84 810.84 .07
Orient X Order 1 5,843.76 5,843.76 .48
Sex X Order 1 5,236.26 5,236.26 .43
Sees X Orient 5 4,651.43 930.29 .59
Sees X Order 5 3,707.60 741.52 .47
Seas X Sex 5 11,223.60 2,244.72 1.44
Sess X Orient X Sex 5 3,927.09 785.42 .50
Seas X Orient X Order 5 5,578.67 1,115.73 .71
Sess X Sex X Order 5 1,709.98 341.99 .22
Orient X Sex X Order 1 52.51 52.51 .00
Orient X Sex X Order X Sess 5 12,831.38 2,566.28 1.64
Within 8 97,239.94 12,154.99
Error t40 62575. 6 1964.39
Total 95 278,942.74

Table 19
Analysis of Variance for Agreement

Source df SS MS F P

Orientation 1 2.04 2.04 .00
Sex 1 3,408.17 3,408.17 1.61
Order 1 6,048.38 6,048.38 2.85
Sessions 5 4,634.59 926.92 2.22
Orient X Sex 1 2,992.67 2,992.67 1.41
Orient X Order 1 925.04 925.04 .44
Sex X Order 1 1,006.67 1,006.67 .50
Sese X Orient 5 1,656.33 331.27 .79
Sess X Order 5 2,688.74 537.75 1.29
Sees X Sex 5 1,619.45 323.89 .77
Sess X Orient X Sex 5 1,923.20 384.64 .92
Sees X Orient X Order 5 3,778.83 755.77 1.81
Sess X Sex X Order 5 703.51 140.70 .34
Orient X Sex X Order 1 20,886.00 20,886.00 9.84 .05
Orient X Sex X Order X Sess 5 355.34 71.07 .17
Within 8 16,985.00 2,123.13
I~rror 40 16,733.00 418.4';
Total 95 86,411.96

Table 20
Analysis of Variance for Giving Opinion

Source df SS MS F P

Orientation 1 6,851.26 6,851.26 .44
Sex 1 3,991.26 3,991.26 .26
Sessions 5 62,097.18 12,419.44 7.18 .01
Order 1 43,818.76 43,818.76 2.82
Orient X Sex 1 7,975.26 7,975.26 .51
Orient X Order 1 5,002.59 5,002.59 .32
Sex X Order 1 243.84 243.84 .02
Sess X Orient 5 6,228.66 1,245.73 .72
Sess X Order 5 17,816.66 3,563.33 2.06
Sess X Sex 5 14,003.36 2,800.67 1.62
Sess X Orient X Sex 5 10,577.67 2,115.53 1.22
Sess X Orient X Order 5 17,433.16 3,486.63 2.02
Sess X Sex X Order 5 2,526.85 505.37 .29
Orient X Sex X Order 1 1,512.10 1,512.10 .10
Orient X Sex X Order X Sees 5 8,313.98 1,662.80 .96
Within 8 124,281.26 15,535.16
Error 40 69 203.24 1730.08
Total 95 396,874.50

Table 21
Analysis of Variance for Giving Orientation

Source df SS MS F P

Orientation 1 35.04 35.04 .01
Sex 1 2,053.00 2,053.00 .76
Order 1 2,166.00 2,166.00 .80
Sessions 5 2,882.21 576.44 .95
Orient X Sex 1 294.00 294.00 .11
Orient X Order 1 2,521.50 1,521.50 .93
Sex X Order 1 3,384.38 3,383.38 1.25
Sees X Orient 5 5,179.96 1,035.99 1.71
Sess X Order 5 5,247.00 1,049.40 1.73
Sess X Sex 5 2,922.25 584.45 .96
Sess X Orient X Sex 5 1,202.49 240.50 .40
Sess X Orient X Order 5 3,273.74 654.75 1.08
Sess X Sex X Order 5 1,893.61 378.72 .62
Orient X Sex X Order 1 532.04 532.04 .20
Orient X Sex X Order X Sees 5 1,563.74 312.75 .51
Within 8 21,683.01 2,710.38
Error 40 2429Q8.99 607.47
Total 95 81,132.96

Table 22
Analysis of Variance for Disagreement

Source df SS MS F P

Orientation 1 1,190.04 1,190.04 4.51
Sex 1 96.00 96.00 .36
Order 1 140.17 140.17 .53
Sessions 5 212.71 42.54 .91
Orient X Sex 1 146.09 146.09 .55
Orient X Order 1 273.38 273.38 1.04
Sex X Order 1 30.83 30.83 .12
Sessions X Orient 5 314.83 62.97 1.35
Sees X Order 5 52.33 10.47 .23
Sess X Sex 5 96.37 19.27 .41
Seas X Orient X Sex 5 171.78 34.36 .74
Sees X Orient X Order 5 351.24 70.25 1.51
Sess X Sex X Order 5 76.79 15.36 .33
Orient X Sex X Order 1 5.03 5.03 .02
Orient X Sex X Order X Ses. 5 734.29 146.86 3.16 .05
Within 8 2,111.01 263.88
Irror 40 1 860.9Q .46.52
Total 95 7,863.83



Each of the 27 statements below have three completions.
Indicate the one with which you agree the most by placing a
plus sign (+) in the blank to the left of the completion;
indicate the one with which you agree the least by placing a
minus sign (-) in the blank to the left of the completion.

1. One of the greatest satis-
factions in life is:
__Recognition for your
The feeling of a job
well done
.'The fun of being with

2. If I played football, I
would like to be:
._The coach whose planning
pays off in victory
The star quarterback
__Elected captain of the

3. The best instructors are
those who:
__Give you individual help
and seem interested in you
__Make a field of study in-
teresting, so you want to
know more about it
__Kake the class a friendly
group where you feel free
to express an opinion

4. The worst instructors are
those who:
_Are sarcastic and seem to
take a dislike to certain
_Make everyone compete with
each other
_Simply can't get an idea
across and don't even seem
interested in their subject

5. I like my friends to:
Want to help others
whenever possible
Be loyal at all
_Be intelligent and in-
terested in a number
of things

6. My best friends:
Are easy to get along
Know more than I do
_Are loyal to me

7. I would like to be known
A successful person
I__ An efficient person
A friendly person

8. If I had my choice, I
would like to be:
A research scientist
A good salesman
A test pilot

9. As a kid, I most enjoyed:
Just being with the gang
-The feeling of accom-
plishment I had after I
did something well
__Being praised for some

10. Schools could do better
jobs, if they:
_Taught children to
follow through on a job
__Jncouraged independence
and ability in children
uttless emphasis on
getting along with

11. The trouble with an organi-
zation like the Army or
Navy is:
._The rank system is un-
_The individual gets lost
in the organization
_You can never get any-
thing done with all the
red tape

12. If I had more time, I would
like to:
Make more friends
_Work at my hobby or be
learning something new
and interesting
_Just take it easy, with-
out any pressure

13. I think I do my best when:
.I work with a group of
people who are congenial
_I have a job that is in
my line
My efforts are rewarded

14. Whst I like best is:
_Being appreciated by
Being satisfied person-
ally with my performance
_Being with friends with
whom I can have a good

15. I would rather that a
story about me appear
in the newspaper:
_Describing a project
I had completed
_Citing the value of
my actions
Announcing my elec-
tion to a fraternal

16. I learn best when my
__rovides me with
individual attention
_Stimulates me into
working harder by
arousing my curiosity
Makes it easy to dis-
cuss matters with him
and with others

17. Nothing is worse than:
Having your self-
esteem damaged
__Failure on an impor-
tant task
___Losing your friends

18. I like:
Personal praise
._Cooperative effort

19. I am disturbed consider-
ably by:
Hostile arguments
-Rigidity and refusal
to see the value of
new ways
PPersons who degrade

20. I would rather:
__e accepted as a
friend by others
_Help others to com-
plete a mutual task
___Be admired by others

21. I like a leader who:
Gets the job done
Makes himself respected
by his following
Makes himself easy to
talk to

22. I would rather:
__Have a committee meeting
to decide on what the
problem is
___.ork out by myself the
correct solution to the
__Be valued by my boss

23. Which type of book would you
rather read:
__A book on getting along
with people
__An historical romance
A how-to-do-it book

24. Which would you prefer:
Teach pupils how to play
the violin
._Play violin solos in
Write violin concertos

25. Which leisure-time activity
is more satisfying to you:
__atching westerns on 'IV
__Chatting with acquaint-
__Keeping busy with inter-
esting hobbies

26. Which would you prefer,
assuming the same amount of
money was Involved:
__Plan a successful contest
Win a contest
Advertise the contest and
get others to participate

27. Which is the most important
to you:
To know what you want to
To knew how to do what you
_To know how to help others
to do what they want



1. Please rank the group members according to the quality of
ideas contributed to the group discussion. Assign rank 1
to the person who contributed the best ideas, rank 2 to
the next best . and the rank of 5 to the person who
contributed the least good ideas. Include yourself.




2. Please rank the group members according to their contri-
bution in guiding the group through the discussion.
Include yourself.

Members HaMk


3. Please rank the group members according to the amount of
participation in the group discussion. Give rank 1 to
the person who participated the most, rank 2 for the next
most participation, rank 5 for the least participation,
etc. Include yourself.




4. Fill in the blanks with the letter of the person of your

a. I interacted with most; least.
b. I liked _most; least.
c. I agreed with post; least.
d. The major spokesman (leader) for the group was .

5. Rate your satisfaction with the role you played in the
group discussion by checking the appropriate description.

Very strongly

_ Neutral

Very strongly

6. Rate your satisfaction with the decisions that were made
in the group.

Very strongly


Very strongly



In May, 1956, C. A. Vance, a graduating student at the
Harvard Business School, had an interview with M. N. Northey,
personnel manager, in connection with a possible position
with the Liddick Company, Columbus, Ohio. The results of
the interview were favorable, and Mr. Vance was asked to
come to the head office of the company at a later date for
further interviews. He was subsequently hired.
The company is one of the largest food processors in the
country. Its products were nationally advertised and well
established. When Vance went to the head office, he knew the
following facts about the company. The company's profit
reached a peak in 1951 but in 1952 had fallen more than 50%.
Since that time recovery in earnings had been steady, but
total profits remained below the 1951 top. At the present
time the company faces the problem not only of maintaining
and improving its merchandising efforts on its several
products, but also of improving its production and purchasing
methods. Its professed policy of "no speculation' in the
materials used in production is not regularly followed.
Vance's father, a wholesale grocer, had long been a
customer of the Liddick Company. He was intimately acquaint-
ed with the company's local representative and its divisional
sales manager in the territory where he did business. He
had met the company's sales manager and two or three other
executives from the head office on several occasions.
In June, 1956, at the end of the Business School year,
Vance was interviewed again, this time by H. S. Wellington,
top executive, and A. B. Randles, general manager, who pro-
posed that he join the company as a salesman, study their
merchandising methods, and later be returned to the head
office to work in merchandising. Vance accepted the position
and was ordered into the field sales division. He worked as a
salesman calling on the retail grocery trade until October 9,
1956. On that date he received a telegram ordering him to
report to Randles at the head office the next morning.
At the head office Handles told him that they were
contemplating an addition to their purchasing department
and were thinking of assigning some person to make a study
of the principal commodities purchased by the Liddick Com-
pany, to analyze the prices and trends of the prices, and
to make recommendations to Handles and Fairing, the chief
purchasing agent. Handles went into considerable detail as
to how he wanted the job done. It appeared to Vanee that
the company had been doing its purchasing on a very haphazard
basis. Handles told him that they wanted to make their
purchasing more scientific and to coordinate their usage,
storage, and purchasing. Handles made it plain that he did
not expect any tangible results for five or six months,

because he realized the large amount of research and pre-
liminary work that Vance would have to do before producing
worth-while information.
Vance was assigned a desk in the middle of the steno-
graphic department, since the offices were very crowded.
Fairing, the chief purchasing agent, instructed him to begin
work on rice. He immediately began collecting files on the
rice industry and the factors affecting grain prices.
Correspondence was begun with the Department of Agriculture,
various state agriculture departments, trade associations,
the Chicago Board of Trade, and other institutions.
Vile this material was coming in and being assembled,
Vance was setting up a commodities library in an attempt to
broaden his knowledge of the ccmmodlties purchased. In this
he was assisted by some of his former professors. Further-
more he induced Fairing to join the National Association of
Purchasing Agents, so that the Liddick Company might receive
the benefit of that Association's work. He had to use
considerable persuasion to accomplish this result.
There was so much to be done and Vance was so eager to
get his department under way that each afternoon he dictated
from 10-15 letters, besides reports and memoranda. throughh
November and December he continued to assemble and complete
the rice report, which included statistical studies of
prices, yields, crops, and a comprehensive analysis of the
factors affecting prices, together with a general routine
for buying and handling the rice on the most economical
basis. He also kept posted on the prices of all the
commodities bought by the company and reported them to
When Vance reported to work after the Christmas holidays,
he was called into Fairing's office and bluntly told that
"there is no statistical department." He was told that the
company had decided that his work was not so important to the
company as having an experienced grain buyer, who could keep
posted personnaly on corn and wheat and act directly on his
own authority, thereby lessening Fairing's work. When Vance
inquired about whether or not his work had been satisfactory,
he was assured that his work was done as instructed and his
dismissal was only because the budget of the purchasing
department would not permit his retention. At this point
Vance asked for a letter from Fairing stating that his
dismissal from the company was not due to unsatisfactory
work but to the changed requirements of the purchasing
department. Fairing agreed to write such a letter and
promised Vance would receive it in a day or two.
From the time Vance took the job to the time of his
dismissal, he was given no indication that his work was in
any way unsatisfactory, except on two occasions. Both of
these arose when he was warned by Fairing that he was giving
the stenographic department too much work. Fairing stated
that Vance w?s writing so many letters that another stenog-
rapher would have to be hired, but the budget would not
permit it. The second time Fairing warned him about too
much work being given to the stenographers, Vance was told

that he should not have written individual letters to his
Business School professors, but that he should have written
one letter, instead, to one of them, and merely mention the
others in the letter.
On Friday, December 30, Vance received a copy of a form
mailed to the State Unemployment Commission, signed by the
office manager, and giving the reason for his dismissal as
"Work Unsatisfactory." Vance immediately telephoned Fairing
and asked him why he had received such a paper, when he had
previously been informed otherwise. Fairing appeared
surprised that Vance had received it and said that it was a
mistake and that he would go to the office manager to
correct the error, and would call back immediately. Fairing
did not call back.
On January 3, Vance called Fairing and stated that he
had not received the letter he had requested. Fairing
explained that he had not had time to dictate it, and that
Vance would not receive it until it was passed by the
management. Vance asked Fairing why he had said that his
work was satisfactory and then repudiated the statement.
When Fairing made no reply, Vance asked him directly what
particulars about his work was unsatisfactory. Fairing
replied that Vance had "caused trouble in the office" and
that his work had been "inaccurate." Fairing then stated,
"Some people around here say that you cost the Liddick
Company 410,000." In answer to why the charge had been
made, Fairing replied that Vance had written letters to the
rice millers and associations seeking information on the
rice market in general, and this had caused the millers to
become alarmed, with the result they raised the price of
riee, believing the Liddick Company was in the market.
Vance replied by indicating that the Liddick Company buys
less than 1% of the total U.S. production of rice. Also he
reminded Fairing that these letters had gone across his desk
before they were mailed.
The next day Vance called Northey at his office and
told him about the letter that Fairing had promised and had
not written. Northey said that he would call back in a few
minutes. When he called back in about 30 minutes, he said
a letter could not be written, but if Vance wanted employ-
ment, he could ask the prospective employee to write him
for information about Vance. Before the interview was
terminated, Northey volunteered to write Vance's father and
tell him the reasons for his dismissal. He has not yet
written the letter.
Vance wrote the dean of the Business School and
requested him to write the Liddick Company to inquire as
to the reasons for his dismissal. Vance does not believe
that the real reasons have as yet appeared.


Your task as a group will be to study and discuss a
human relations problem in a business situation, to make
a comprehensive report concerning the matter, and to
formulate recommendations for action. The problem centers
around a recent Harvard Business School graduate,
C. A. Vance, who was abruptly dismissed from his job with
a large food processing company. You are to place yourselves
in the role of a state Fair Employment Practices Commission
who has been asked to investigate the incident, and report
your findings and recommendations. Included in the report
should be factors such as the relationships among the
company board members, the relationships between Vance and
his superiors, his attitude toward his work, the quality
of his work, mistakes he may have made, and the reasons
for his dismissal. The report should include any and all
factors you feel have a bearing on Vance's experience with
the company and his subsequent dismissal. An investigator
has been sent from the FEPC to the company in order to
collect more relevant information concerning the case,
and to send reports back to the FEPC. The information in
these reports will be made available to you during several
of the sessions. Your final report will be evaluated
according to the number of relevant factors considered,
your realism and imagination in dealing with them, and the
clarity and cogency of expression.

You will meet twice a week for four weeks, making a
total of eight sessions. You are to take the entire time
to discuss the problem and make your report. There will
be no other problems to discuss. You will have 45 minutes
per session for discussion. Place your names on the problem
sheet so you can get the same one back each session. This
is your sheet and you may take notes on it if you desire.
Additional scratch paper will also be available. At each
chair around the table you will find a card with a letter
on it. The letter is placed there for purposes of
identification, so please sit at the same place each

After you have read the instructions, you may begin
reading the problem. If you have any questions, you may
ask them after everyone has read the problem. No questions
will be answered about group procedure, and no questions
will be answered after discussion has begun.


(Personality Descriptions)

Within the last few years A. B. Randles had been hired
as general manager. Randles was formerly connected with
a management engineering firm and in that capacity had
advised the Liddick Company from time to time. He has held
this position for five years. Within the past five years
the company has been passing through several upheavals in
its personnel. Randles has been making numerous changes
in the higher personnel since his appointment as general
manager, and most of these changes have bean to the
advantage of the company. At the present time, among his
other duties, he is sales manager. He has refused to
delegate any responsibilities in the running of the company
and, as a result, is taxing himself with the entire burden
of operations. He realizes the need for delegating
responsibilities but finds it extremely difficult to choose
men who will be acceptable not only to himself but also to
Wellington. Consequently, no one in the company can make
a move in any direction without first selling his idea to
Bandles and Wellington.
Mr. H. S. Wellington owns the controlling share of the
business. In the past he had made most of the decisions
that affected the business in any important way. Wellington
built up the company by clever and unusual merchandising and
really ran the entire business himself. It was this
consistently good merchandising program that has enabled
the organization to grow to its present size. As two of
the older employees explained, "We made money despite our
mistakes." In fact the company made a fortune for
Wellington. Wellington insists that all activities of the
company be highly organized and rigidly controlled.
However, he is now approaching the age of retirement and
has been taking steps to turn over to other hands the
active management of the company.

(Personality Descriptions)

M. N. Northey holds the title of personnel manager.
He is in charge of all personnel, both sales and production.
You will recall him as the Liddick agent who interviewed
Vance at the Business School. With good reason, Northey
is very unpopular with all but two of the higher officials,
but he has managed to establish himself in H. S. Wellington's
confidence. He is described as a little shy and slow in
getting to know people. He is regarded as "hard-boiled" and
one who demands everything be done to perfection.
The purchasing department includes Fairing, chief
purchasing agent, who buys grains and supervises all
purchases; F. E. Dewick, who buys paper and advertising
material; D. C. Croft, who is responsible for the purchase
of machinery, oils, and greases; and P. 0. Berry, who
purchases office supplies and miscellaneous items. The
purchases of commodities other than the foregoing is split
up among Fairing, Dewick, and Croft. This is the entire
purchasing department personnel, and they perform the
complete purchasing function as best they can. For his
sources of information, Fairing depends solely upon brokers'
advice, the Chicago Tribune, and the Chicago Journal of
Commerce. The company receives no government reports of
any nature. The purchasing library consists of several
books and catalogues on machinery purchases. Vance reports
that he discovered Fairing to be a man who did not
understand what he (Vance) was talking about at least half
of the time. Fairing seems completely ignorant of the
commonest statistical terms and it was necessary for Vance
to phrase reports to him very carefully so that he could
understand them at all.


During his conversation on the telephone with Northey
following his dismissal, Vance asked Northey if the Liddick
Company had not made a mistake in its survey of what his
assignment was to be since in a short time the officials
had discovered that an experienced grain buyer was more
urgently needed than Vance was. Northey replied, "Oh no.
Our mistake was in hiring you at all." Evidently before
Northey had called Vance back, Fairing had given him a
great deal of information concerning Vance.
Vance made an appointment with Northey for the follow-
ing week to discuss the reasons for his dismissal. Northey
then explained that one reason was that Vance was disliked
by the workers in the department. Vance felt that this
was so far-fetched that it was laughable. He felt that
everyone was his friend and indicated that everyone had
shown his friendship in many ways, by invitations for
fishing and hunting trips, visits to their homes, and an
expensive wedding gift from the office force when he was
married. Another reason mentioned was that Vance had been
"indiscreet." Simply that, with no explanation of his
indiscretion. Northey also mentioned a report that Vance
had not been courteous to the grocery trade when he called
on members. Vance thought he was a little late with that
criticism since he had not been on the road for nearly three
months. He felt that the last reason had as little basis
as the first.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs