LEADERSHIP ORIENTATION AND SELF-PRESENTATION: A LEADER'S
IMPRESSION MANAGEMENT UNDER STRESS
BRUCE T. CAINE
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOC THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
The author expresses his thanks to all those who supported and
assisted in this project. To the members of his review committee, he
extends special thanks for their encouragement and patience. To the
members of the University of Florida Reserve OFficers Training Corps
and their cadre, he sends his wishes for continued success. But the
most sincere and heartfelt gratitude must be reserved for his wife,
Pam and his children, Scott and Alyssa, for without their support,
understanding and love this project would never have been completed.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS. . . . .
LIST OF TALES. . . . . .
LIST OF FIGURES . . . . .
ABSTRACT. . . . . . .
INTRODUCTION . . . . . . . .
A Role flaking Model of Leadership. . . .
Social Power and the Leadership Process. .
Leader Behavior--the Application of Social
Social Roles and Social Exchange . . .
Role Miaking . . . . . . . .
The Self-PresentaLional Process. . . .
Establishing Credibility . . . ..
Defining the Situation . . . . .
II PREDICTORS OF SELF-PRESENTATION.
. . . . . 22
Interpretation of the Scales . . . . . .
Hypotheses Relating the Predictor Scales. . . .
III SELF-PRESENTATION AND ROLE BUILDING. . . . . .
Measures of Self-Presentation. . . . . . .
Analysis of the Leadership Situation . . . .
The Korten Stress Model . . . . . ..
The Life Cycle Theory. . . . . . . .
The Contingency Model . . . . . . ..
The Experimental Situations. . . . . . .
Hypotheses Concerning Situational Self-Presentation.
IV EXPERIMENTAL DESIGN. . . . . . . . ...
Method . . . . . . . . . . . .
Subjects . . . . . . . . . . . 61
Experimental Ovarview. . . . . . . ... 61
Procedure. . . . . . . . . . .. 62
V RESULTS. . . . . . . . . . . . .. 66
Predictor Scale Relationships. . . . . . .. 66
LPC/Self-Concept Relationship. . . . . . 66
LPC/Behavioral Survey Relationship . . . . 67
Self-Esteem/Behavior Survey Relationship . . . 69
Situational Evaluations . . . . . . .. 70
Verification of the Contingency Model Situational
Dimensions . . . . . . . . . . 70
Verification of the Situational Stress
Manipulation . . . . . . . . . 70
Prediction of Self-Presentation. . . . . ... 71
Contingency Model and Prediction of Self-
Presentation . . . . . . . . . 72
Evaluation of Leader Behavior. . . . . .. 75
VI DISCUSSION . . . . . . . . . . . 80
Predictor Scale Relationships. . . . . . .. 80
The Prediction of Self-Presentation . . . ... 82
Self-Presentational Predictability and
Effectiveness Ratings. . . . . . . .. 83
Evaluation of the Research Design. . . . ... 84
Conclusion . . . . . . . . . . . 86
APPENDICES. . . . . . . . . . . . . .. 89
A LEADERSHIP BEHAVIOR SURVEY SCALE . . . . . . 91
B LEAST PREFERRED CO-WORKER SCALE. . . . . . .... 94
C INFORMATION EXCHANGE WORKSHEET . . . . . . 96
D POST-EXPERIMENTAL RATING FORM. . . . . . .. 98
REFERENCES. . . . . . . . . . . . . .. 102
BIOGRAPHICAL SKETCH . . . . . . . . . . . 108
LIST OF TABLES
1 SYSTEMATIC PRESENTATION OF THE RELATIONSHIP OF LPC AND
LEADER BEHAVIOR AS DESCRIBED IN FAVORABLE AND
UNFAVORABLE SITUATIONS (FROM FIEDLER, 1972). . . ... 38
2 ANALYSIS OF THE STRUCTURE OF TACTICAL SITUATIONS
MEAN RATINGS . . . . . . . . . . . 53
3 PREDICTED EFFECTS ON SELF-PRESENTATION OF THE LPC
DESIGNATION BY FUNCTIONAL DIMENSION EMPHASIS
INTERACTION. . . . . . . . . . . .. 56
4 SELF-PRESENTATION SCORES . . . . . . . . . 74
5 COMPARISON OF LEADERSHIP PERFORMANCE RATINGS AND
PREDICTABILITY . . . . . . . . . . . 76
LIST OF FIGURES
1 Johnson and Johnson's Behavior Survey Grid. . . . 68
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
LEADERSHIP ORIENTATION AND SELF-PRESENTATION: A LEADER'S
IMPRESSION MANAGEMENT UNDER STRESS
Bruce T. Caine
Chairman: Dr. Marvin E. Shaw
Major Department: Psychology
Organizational leadership has been viewed both as a social role bounded
by norms and expectations, and as a dynamic social influence or social
exchange process. In both, the contingencies of the situation and the
personality of the leader are seen to interact in the production of
leadership behavior. The evaluation of that behavior depends on the degree
to which it satisfies the expectations of the organization and of the group
members and fosters goal achievement for both.
The organizationally appointed leader's objective is to achieve
maximum performance with a minimal use of formal sanctions. One method
the leader has to achieve this is through the establishment of a working
relationship with each subordinate. This process has been labeled "role-
making" and involves the exchange of information and expectations. In
the initial phase of "role-making" self-presentation or impression
management is seen as crucial. This "taking of a line" or establishing
of "face" is the individual's way of trying to control the definition of
the situation and the informal social exchange agreements that are formed.
For the appointed leader, self-presentation is seen as a means of estab-
lishing initial credibility and indicating one's suitability for the
leadership role so as to gain a high degree of voluntary compliance
from subordinates early in the interaction.
Three scales were considered as possible predictors of a leader's
self-presentation: Johnson and Johnson's Leadership Behavior Survey,
Fiedler's Least Preferred Co-Worker Scale, and Ziller's Self-Other
Orientation Scale. Relationships among these scales were evaluated and
their predictability for a structured self-presentation analyzed in a
laboratory study using decision making groups.
Forty-eight freshman ROTC cadets participated in 4-person groups
in which each served as the group leader for one of four tactical
situations. Situational stress was manipulated by emphasis on the task
responsibilities, the human relations responsibilities, or both of these
functions of appointed leaders. Each member prepared a structured self-
presentation by rating himself on the degree to which each of 33 personal
attributes was representative of him. This description was read to the
group by the member prior to his tenure as group leader. At the con-
clusion of the fourth situation, group members rated each other on
performance and the predictability of their self-presentation for their
behavior as leader.
Certain hypotheses concerning the relationships among the scales
were supported. Low LPC leaders were found to be higher in self-esteem,
as predicted. Against predictions, most leaders described themselves
as active on both dimensions of the Leadership Behavior Survey rather
than emphasizing the functional area corresponding to their secondary
motivation as inferred from their LPC score.
A strong main effect of LPC was found for the favorability of self-
presentations. Low LPC leaders described themselves more favorably. In
particular, low LPC leaders who received the experimental manipulation
emphasizing both dimensions of leadership described themselves most
favorably on the structured self-presentation. High LPC leaders in this
condition, in contrast, described themselves less favorably. These low
LPC leaders' self-presentations were seen as highly predictive of their
future behavior while these high LPC leaders were seen as performing
very differently than they said they would. Leaders in both these cells,
however, achieved high leadership performance ratings.
This investigation provides some evidence that the LPC scale may be
a useful predictor of an individual's self-presentation style and that
different styles may be equally effective in fostering attributions of
effectiveness for the individual in the leadership role.
People come to occupy the social role of leader in a variety of
ways. Some emerge from the group by virtue of activity or resource
to become "first among peers." Others conduct open campaigns for
recognized positions and attain the leader role through election by
the group to be led. Still others are altercast by the group into
the role of leader seemingly against their will and inclination. A
large percentage of those who occupy formal positions of authority,
however, are appointed not by the group members themselves but by
the larger organization or corporate hierarchy of which that group
is a part. The production foreman, the school principal, the police
sergeant, the corporate manager, and the military officer all occupy
positions of authority by virtue of organizational appointment. The
problem of predicting how effective an appointed individual will be in
the leader role has intrigued students of the leadership process for
centuries. Single factor leadership theories ranging from the "Great
Man" and trait hypotheses to situational force prescriptions have
failed to attain substantial predictive validity. Interactionist
theories, as exemplified by Fiedler's Contingency Model (1967) have
achieved greater success by considering the complex interactions
between the unique characteristics of the leader and the intricacies
of the situation.
Leadership is viewed as both a social role bounded by expectations
and as a dynamic social influence/social exchange process. The specific
role expectations and exchange agreements that come to operate within
a formal group are derived from the interaction of general social norms,
organizational presumptions and the characteristics of the group members.
Recently, Fiedler (1974) has proposed that organizations can improve
their chances of securing effective leadership by the manipulation of
certain situational variables and/or the appropriate matching of a pro-
spective leader to situations which favor that individual's leadership
"style." In his formulation, Fiedler (1974) proposes two contrasting
"styles" based on the construct of motivational hierarchy and utilizes
three situational factors to analyze the degree of uncertainty or stress
confronting the leader. Based on over one hundred empirical studies,
Fiedler concludes that "a leader's performance depends as much on the
situation assigned him (her) as on his or her own personality" (1974,
p. 70). The organization may most readily change the situation by
altering the structure of the group task and/or by adjusting the level
of position power authorized. Changes in the most significant of
Fiedler's three situational dimensions, however, depend primarily on
the leader, for it is he or she who must establish person to person
working relationships with subordinates and superiors.
It is the objective of this paper to present an investigation of
the processes of self-presentation and role-building as they relate
to the establishment of such operational relationships between the
appointed leader and others within a formal group. The ability to
establish one's credibility and suitability for the role of leader
is seen as crucial to the formation of working relationships that will
facilitate goal achievement and ascriptions of effectiveness. Based on
this assumption, the central questions then are (1) is there a way to
predict a leader's self-presentational emphasis, (2) do situational
as well as personal variables effect self-presentations, and (3) do
the leader's early efforts at defining his/her role affect later
evaluations of his/her performance on-the-job?
A Role Making Model of Leadership
Most classic and contemporary psychological theories of leadership
have concentrated on the precursors and results of leadership behavior.
Classic theories, based on the presumed traits of leaders and principles
of leadership performance, assumed that possession of the former and
knowledge of and adherence to the latter would foster acceptance of the
individual as "leader" and enhance his/her effectiveness. Having the
"correct" complex of traits coupled with somewhat flexible behavior
patterns (based on knowledge of "the principles") was viewed both as
the predictor of and criteria for judging suitability for leadership.
Situational theories of leadership and management, in contrast,
have concentrated on the environmental precursors and on-going social
forces. Conceptually, as the situational forces vary, required or
preferred leadership behaviors also vary. A number of models have
been proposed and tested to varying degrees which suggest the matching
of leadership behavior to the demands of such situational variables
as time constraints, communications system, imposed organizational
structure, task demands, group composition, and stress.
Numerous definitions of leadership have been proposed that suggest
leadership should be viewed as an on-going group process. Allport
(1924) proposed that leadership requires direct, face to face contact
between leader and followers as it is a process of "personal social
control." Nash (1929) advanced his belief that leadership "implies
influencing change" in the behavior of people, while Tead (1929)
defined it as the activity of influencing others to cooperate in
achieving some mutually desirable goal. For Schenk (1928) leadership
is the "management" of men and women by inspiration and persuasion
rather than by coercion or the threat of coercion. French and Raven
(1959) and others have defined leadership in terms of differential
power relationships operating among the members of a group.
Social Power and the Leadership Process
Bertrand Russell, writing in 1938 stated, "The fundamental concept
in social science is Power, in the same sense in which Energy is the
fundamental concept in physics" (cited in Clark, 1969, p. 511). Russell
proposes that the "laws" of social dynamics can be understood only in
terms of power and considered it the job of social sciences to discover
these laws. Clark (1969) defines "social power" as "the force or energy
required to bring about, to sustain, or to prevent social, political,
or economic change" (p. 515). To have meaning in a social context,
power must be demonstrable. Claims to power are not the same as power
itself. Being in the position of leader and having the authority to
determine the direction of group activities is a potential source of
power. It is not, however, actual power. Performing in the role of
decision maker is an exercise of power only when the decision is put
into action by group members.
The social phenomenon called leadership is the process of gaining
willing compliance with decisions and insuring the execution of
directions by others without resort to coercive force. Therefore, as
the central attribute of the leadership process, interpersonal influence
is social power in operation. As the primary concern of this paper is
the investigation of organizationally appointed rather than emergent
or elected leaders, the following definition of leadership will be
Leadership is "the process of influencing human behavior
so as to accomplish the goals prescribed by the
organizationally appointed leader."(Bons, 1976, p. 6)
Within an organization, the authority to direct and control the
behavior of others is derived from the normative and functional
structures of the organization and is exercised by selected members
who occupy those positions in the hierarchy responsible for the
achievement of organizational goals. Members comply with the direction
of those occupying positions of authority because they recognize the
right of the organization to influence their behavior on certain tasks
.and in certain situations. In exchange for this granted power, the
organization provides rewards or remuneration to its members. However,
as noted by Katz and Kahn (1966), the essence of effective organizational
leadership is the achievement of desired behavior beyond simple com-
pliance with "routine directives" or minimally acceptable performance.
Effective leaders are those who are able to exercise some forms of
social power beyond the position-dependent legitimate power acceded to
all who occupy that organizational position.
French and Raven (1959) proposed four bases of social power in
addition to legitimate power. Reward and coercive power are in large
measure dependent on the degree of legitimate power and distributable
resources provided to the leader by the organization for use as his/her
discretion. Personal resources also play a role but are seen as
instrumental in determining the degree of referent and expert power
attributed to the leader by followers. Identification with the leader
is the basis for referent power while acknowledgement of task relevant
skills and knowledge forms the foundation for the exercise of expert
power. Leadership, then, is the ability to acquire and to apply
appropriately the various forms of social power so as to attain the
prescribed goals of the group through coordinated action of the members.
Leader Behavior--the Application of Social Power
An incredible amount of research effort has been expended in quest
of the definitive list of leadership/management/supervisory behaviors.
Most present classifications of leadership behavior can be traced, at
least in part, to Halpin and Winer's (1952) report of Hemphill's 1950
Ohio State University study of aircraft commanders. These authors
suggested four factors: consideration, initiation of structure,
production emphasis, and sensitivity or social awareness.
Bowers and Seashore (1973) reviewed the research on leadership and
compared a large number of conceptualizations of leadership functions.
While they prefer use of four factors, the dual aspects of group
maintenance functions (consideration) and goal achievement functions
(task orientation) preferred by Cartwright and Zander (1968) and
Stogdill (1974) seem to have gained general acceptance. Fiedler (1971)
calls the identification of the two factors of initiation of structure
and consideration "one of the most important achievements of leadership
research" (p. 7). This may well be true, yet military theorists since
Sun Tzu, writing for his emperor at the end of the sixth century B.C.
(Griffith, 1963), have emphasized these two factors.
Present day military leadership theory can be summarized in the
concise statement from the Armed Forces Officer (DOD, 1960), "The
winning of battles is the product of the winning of men" (p. 70).
The mission, always of prime importance, can be accomplished most
effectively by gaining the willing compliance and support of those men
who must work together to accomplish it. Leadership, as "the winning
of men," is an influence process, one that is based on the building of
credibility, confidence, trust, and mutual respect between the leader
and the led. The "winning of battles" is essentially a management
process, concerned with the efficient and effective application of
resources, both human and material, to the accomplishment of a given
task. However, without the former, the latter process will suffer from
internal power struggles which will vie with task related challenges
for the attention and energies of the group or organization and its
members. The would-be leader is the military in particular, but also
in industry, business, and government is counseled that mission
accomplishment, whether that mission is victory in battle, high pro-
ductivity, maximum profits, or national survival, is paramount. This
goal is, according to the United States Army's Field Manual 22-100
Military Leadership, "fully compatible with a leadership approach that
includes reasonable consideration of the men being led" (1973, p. i).
Taking care of your subordinates, men and women, is a cardinal principle
of leadership; "It is a paramount and overriding responsibility of every
officer to take care of his men before caring for himself" (The Armed
Forces Officer, 1960, p. 30). The leader's dual responsibilities of
mission accomplishment and concern for subordinate's welfare are
intimately interwoven. And yet, the descriptions of leadership behavior
and the classification systems of that behavior that have come into
common use in the psychology and management literature have succeeded
in isolating these responsibilities and for the most part ignoring
Research with small problem solving groups conducted concurrently
with Hemphill's study resulted in the development of the Interaction
Process Analysis technique (Bales, 1950). From this research base,
Bales proposed that leaders may be classified on the basis of their
most frequent behaviors as "task specialists" or as "social specialists."
Suggesting that groups need both types of leadership, Bales acknowledged
that both types might be provided by the same person. While it may not
be reflected in reality, the military leader is expected to do just
that--perform as both task and socio-emotional leader for his unit.
This behavioral specialization is often referred to as role
differentiation and has been conceptually tied to the concepts of
social exchange, social power, and reciprocal social influence in
an effort to more fully describe the leadership process.
Social Roles and Social Exchange
Homans (1958) is generally recognized as having first conceived
of human interaction and communication in groups as an exchange process.
Each member is seen as having certain expectations for the interaction,
expectations both of process and of outcome. Many of these expectations
are often tied to the normative structure of the group and society at
large through the medium of social roles.
There are three different aspects of a social role. The perceived
role is that complex of behaviors viewed by the position holder or
"actor" as appropriate for one occupying that position. Other social
group members similarly have a perception of the appropriate behaviors
for a position occupant which may be termed the expected role. Thirdly,
the actual behavior of a position occupant comprises the enacted role,
which may be variably correspondent with the other two conceptualizations.
Role conformity is defined as a high degree of correspondence between
expected and enacted roles. In-role performances (enacted roles) are
judged both on their degree of correspondence and on the outcomes they
Playing a role well often means conforming to the expectations of
the group which defines the expected role components. Sarbin (1968)
views role expectations as a cognitive concept through which positions
in the social structure are defined in terms of actions and qualities
expected of the person who at any time occupies the position. Sarbin
describes it thus:
The conduct expected of the occupant of the position, the
exercise of rights and privileges, and the fulfillment of
duties and obligations, applies to the person who at any
time is assigned this role. (Sarbin, 1968, p. 498)
Sarbin (1968) proposes three judgmental questions concerning role
enactment. First, is the conduct (behavior/enacted role) appropriate
to the social position granted to or attained by the actor? That is,
does the performance fit the ecological setting? Secondly, is the
enactment proper? That is, does overt behavior meet normative demands
and standards (expected role) which serve as evaluation criteria for
the observer? And thirdly, is the enactement convincing? That is,
does the enactment lead the observer to declare unequivocally that
the incumbent is legitimately occupying the position? The answer to
each of these questions will determine in part the costs and rewards
incurred by the actor as a result of that behavioral sequence. This
is especially true if the behavior is an influence attempt made by an
individual in a leadership position.
As Hollander and Julian (1970) put it, "...the leader who fulfills
expectations and helps achieve group goals provides a rewarding resource
for others which is exchanged for status, esteem and greater influence"
(p. 35). This is the essence of the leader-follower exchange agreement
or "informal contract." In return for his/her "leader pay" (defined by
Jacobs, 1971, as higher degrees of status wisdom, esteem, competence,
etc.) the individual performing the leadership role is expected to facili-
tate the satisfaction of member needs and aspirations. Therefore, the
outcome of group activities is an important determinant of the final
social exchange ratio.
The personal determinants of role enactment are (1) past experience,
(2) physical and intellectual ability, (3) the number of roles and
positions simultaneously salient to the actor, (4) organismic involvement
(effort), and (5) preemptiveness (time and priorities). Cameron (1950)
proposed that the more roles in an individual's repertoire (i.e., past
experience) the better prepared this actor will be to meet the exi-
gencies of social life. A diverse repertoire would insure possession
of such assets as adaptability and flexibility, which are manifest in
the ability to reciprocate to novel input and to accommodate stress
from both sudden change and from long unremitting strain within the
social context. In this regard, Biddle and Thomas (1966) employ the
concept of "specialization" which is defined as the amount and variety
of particular differentiated behaviors engaged in by a person. The
individual possesses both a level of within-role specialization and a
broader level of role-summed/position-summed specialization. In the
social exchange of group interaction, role appropriate abilities, skills,
knowledge, techniques, physical traits, and past experience become
"currency" available to the individual.
In the special case of the position of leader, role enactment centers
around the exercise of influence. Bass (1961) has divided the process of
social influence into three stages. He labeled an individual's efforts
to alter or direct the behavior of others as "attempted" leadership.
In both informal groups and formal organizational settings, numerous
suggestions, proposals, and wishes are expressed. Of these attempted
leadership acts, only those that actually produce the suggested or
desired change in others can be classified as "successful" leadership.
"Effective" leadership exists in Bass' model only when the change in
behavior results in reinforcement or reward for the complying individuals.
A member's action or effort that leads the group to execute a course of
action that results in goal achievement is, therefore, an act of
effective leadership, regardless of whether this action was taken by
the formally recognized "leader" or a hierarchical subordinate. It must
be noted that Bass' formulation has been primarily applied to the concept
of emergent leadership in "leaderless groups." It suggests, however,
that judgments of the effectiveness of an individual occupying an
institutional position of leader (titular leader) will be made on the
degree to which his/her decisions, directions, and guidance result
first in compliance and then in rewarding outcomes for the organization
and its members.
The social exchange process involves the expectations that each
individual has for himself and for each of the other members of the
group. Argyle and Kendon (1967) propose that one of the major
determinants of these expectations and subsequently one's performance
in a situation is one's conception of the "kinds of others" he/she
is dealing with. If we perceive our fellow workers and/or subordinates
as self-motivated, enthusiastic, cooperative and competent our behavior
towards them will vary radically from that which we would manifest if
we saw them as unmotivated, antagonistic, and incompetent. In 1964,
Dr. Robert Ziller proposed that the extent to which the leader dif-
ferentiates among the members of his group is positively related to the
variance of the leader's performance ratings of these members as well
as the the "satisfaction and productivity" of the group members. He
cites a series of studies conducted by Fred Fiedler (1960) which
indicated that those leaders described as "psychologically distant"
were more effective in promoting the productivity of task groups than
are leaders with "psychologically closer interpersonal relations."
Fiedler assumed that his "similarity of opposites" and "least preferred
co-worker" scales were the measure of "psychological distance," a term
Ziller suggests is a misnomer for the concept being measured. Ziller
concludes that in reality, the leader is being asked to discriminate
between two persons. Apparently then, the leaders who were effective
in discrimination between "good" and "poor" workers in some general
sense were more effective in promoting the productivity of task groups.
He goes on to state:
...Fiedler's results suggest that leaders of highly
productive groups evaluated the members as individuals
rather than as non-differentiated parts of a greater
whole, the group; that is, members of the more
productive groups were individuated by their leaders. (p. 355)
Erving Goffman (1959) suggests another factor influencing the
social exchange bargain or informal contract arranged between members
of a task group. He states:
Regardless of the particular objective which the individual
has in mind and of his motive for having this objective, it
will be in his interest to control the conduct of the others,
especially their responsive treatment of him. This control
is achieved largely by influencing the definition of the
situation which the others come to formulate, and he can
influence this definition by expressing himself in such a
way to give them the kind of impression that will lead them
to act voluntarily in accordance with his own plan [emphasis
added].' (p. 3)
For the leader of a military, industrial, or governmental organi-
zation, this statement has special importance. Tasked with the res-
ponsibility of directing the group's activities and accountable for the
group's performance on its assigned tasks, the leader must ".control"
the behavior of others by gaining the maximum amount of "voluntary"
compliance. Such voluntary compliance will reduce the leader's costs
incurred through demands for critical observation and detailed super-
vision, threatened and applied sanctions, and promised and granted
incentives. The leader can control the definition of the situation in
large measure by influencing the expectations followers come to form
about his or her suitability to serve in the leadership position and by
prescribing the terms of the informal contract or exchange bargain that
must be negotiated between the leader and each subordinate. Graen (1975)
calls this process of defining the situation and establishing a working
consensus for the group "role making."
Graen (1975) defines role making as a set of processes by which two
functionally interdependent individuals work through how each will behave
in certain situations (interlocking behavior) and agree upon the structural
characteristics of their relationship (constructing relationship norms).
This process occurs against the background of the society, the culture,
and the formal or informal organization to which these individuals
Graen and Cashman (1975) have proposed.that the functional rela-
tionship between a person in a leader position and one in a follower
position constitutes a special case of role making. The occupants of
these two hierarchical positions from a vertical dyad and their inter-
action constitutes a vertical dyad linkage which is circumscribed by
the development of both behavioral norms and relationship norms between
the leader and the individual group member.
Graen and Cashman (1975) summarize the role-making process between
a newly appointed leader and group member as follows:
The model which emerges from our role making studies of
managerial dyads within formal organizations is one of
team building within units. Leaders of managerial units,
when faced with the task of developing new reporting
relationships with most of those they are responsible
for leading, respond in manners which serve to differentiate
their units. With only some of their subordinate managers,
leaders attempt to develop special exchange relationships
which transcend the formal employment contract. All
subordinate managers so selected may or may not accept
such a special exchange relationship. But, those who do
consummate such an exchange, promise to develop into
members of the 'trusted' in-group of the leader. In
contrast, those who either are not given the opportunity
or who decline the opportunity of the special exchange
become members of the out-group of the leader. Thus, the
unit becomes differentiated over time into two distant
subgroups, an in-group and an out-group. (p. 17)
The first stages of role-making must include the exchange of
information between the members of the dyad. Graen calls this the
"Sampling Phase" which begins with the first encounter and the impressions
made by each member. Initial observations and assessments are matched
against expectations for physical appearance, bearing, speech patterns,
task relevant abilities, talents, interests and motivations. It is
during this Sampling Phase that the leader's self-presentational skill
will be particularly important.
The Self-Presentational Process
Well before Goffman (1959) proposed his theatrical analogy of
social behavior, Dexter (1950) wrote, "The most important instrument
with which the leader has to work is himself--his own personality and
the impression that he creates on other people" (p. 592). Anyone desiring
to assume leadership, he counsels, "will discover that his behavior,
gestures, mannerisms, styles of clothing, all may enter into his effec-
tiveness or ineffectiveness--and that in selling his ideas, he must also
be concerned with the impression he himself makes" (p. 593). Dexter
further proposes that "the innovator or leader must study, to achieve
maximum effectiveness, what role he ought to play and, within the limits
of possibility, adapt himself to that role, realizing always that
changing situations may call for a change in roles" (p. 593).
The above statements constitute a fair description of the concepts
and processes involved in social phenomena called self-presentation.
Behavior ranging from formal written self-description such as a resume
to the simplest mannerism presents our "public face" to the world
(Goffman, 1967). Argyle and Kendon (1967) suggest, "Generally speaking,
people present a somewhat idealized or at least edited version of
themselves for public inspection" (p. 82), and propose that the ability
to "create perceptions of and attitudes towards self on the part of
others present is a subtle social skill, though one that is usually
practiced quite unconsciously" -(p. 82).
Goffman (1959) calls this ability to "create perceptions"
impression management which he defines as the process by which people
select those personal attributes for presentation that will facilitate
the creating of a certain impression in hopes of improving their chances
for success in the situation. A consistent and well supported self-
presentation is what Argyle and Kendon (1967) would call a "skilled
performance." They define "performance" as a stream of action con-
tinuously under the control of sensory input, and a "skill" as an
organized, coordinated activity in relation to an object or situation
involving sensory, central nervous system, and motor methanisms. The
elements of a skilled self-presentational performance would include
self-descriptions supported by actual behavior; performance predictions
that could be met or exceeded; and situationally relevant abilities
offered to the group and then provided. Consistency of one's pre-
dictive self-presentation and one's actual behavior will foster
credibility as a communicator (Burgoon, 1974) and as a legitimate
source of influence.
If we wish to make a certain impression, we must carefully control
the appropriate aspects of our public behavior. Our public image will
determine in large measure how others will react to us and, therefore,
we take care to present a "face" that will facilitate social interaction
and/or the achievement of our personal goals for the ongoing interaction.
As noted by Dexter (1950) and frequently by Goffman (1959, 1967),
changing situations call for a change of roles and, therefore, of self-
presentation. The individual in an interaction creates specific
expectations in others through his/her self-presentation. The self-
presentation conveys information about how an individual sees him-/
herself and how he/she would .like to be seen by others. Since the
individual portrays each social role with a degree of uniqueness,
observation of that person's overt behavior is essential to determining
how that individual defines the situation and what position he/she
intends to occupy in the interaction. Goffman says it best:
Every person lives in a world of social encounters,
involving him in either face to face or mediated contact
with other participants. In each of these contacts, he
tends to act out...a line--that is, a pattern of verbal
and nonverbal acts by which he expresses his view of the
situation and through this his evaluation of the partici-
pants, especially himself. Regardless of whether a person
intends to take a line (make a self-presentation) he will
find that he has done so in effect. The other participants
will assume that he has more or less willfully taken a stand,
so that is he is to deal with their responses to him, he
must take into consideration the impressions they have
formed of him. (1955, p. 213)
Numerous studies of persuasion processes including propaganda,
advertising, and attitude change using both direct person to person
and mass communications, have shown the source's credibility or
believability to be of central importance (see reviews by Kiesler,
Collins, & Miller, 1969; Burgoon, 1973).
Source credibility is a perceived phenomenon; it is "in the eye
of the beholder." McCroskey and his associates have suggested five
attributes or dimensions influencing the perception of source credi-
bility. These are competence, character, composure, sociability, and
extraversion (McCroskey, Jensen, & Todd, 1972). Conflict between
dimensions, however, reduces the overall credibility of the source.
Great expertise without trustworthiness (good character) may not
produce high credibility, while extroversion with composure, sociability,
and apparent trustworthiness may overcome a task competence deficiency
and permit influence to be exerted by the source.
Perceptions of credibility are tied to the source's self-presentation
and degree of self-disclosure. Goffman (1959) sees maintaining composure
as crucial to preventing "loss of face," while that class of behaviors
labeled "extroverted" are a means of gaining and holding a central
location (focal point) in the group and consequently the attention of
that audience. Informing others of task related skills and abilities
through self-description and actual performance provides them with
information upon which to make a competence estimate. Perceptions of
trustworthiness result when an individual fulfills or surpasses an
ability self-description or a previously stated intention to act.
Bandura suggests that the consequences for self and others of
behavior also influences the perception of source credibility. If we
observe a source or potential model threaten or promise a particular
consequence (for another individual) contingent upon a prescribed
behavior and then fail to follow through on the threat or promise, the
source's credibility drops. Inconsistencies in "follow through"
behavior by leaders and other holding reward/coercive power makes
their future behavior unpredictable and reduces their credibility.
As noted above, one of the variables involved in the perception
of communicator credibility is competence. In the case of the
appointed leader, the presentation of evidence concerning competence
for the position is of central importance, as referent and expert
power (French & Raven, 1959) are both related to demonstrated competence.
For example, Goldman and Fraas (1965) ran task-oriented groups with
four types of leaders: (1) those appointed by the experimenter for
ability on the task; (2) those artibrarily appointed; (3) those elected
by the group; and (4) no leader (leaderless groups). In this study,
groups worked best for the leader appointed for proven ability, that is,
the ones perceived by the group as having the ability or competence to
lead on the task. Supporting this is the fact that the second best
groups were led by elected leaders, who it can be assumed were seen as
reliable and effective.
Self-presentation therefore can be seen as an integral part of the
building of relationships within a formal group. Each member, but the
new leader in particular, presents his/her definition of the role he/
she is to play and the personal attributes they see as appropriate to
that performance. Subsequent evaluations of performance may well be
influenced by the degree of correspondence between the leader's initial
self-description and subsequent on-the-job behavior. This subsequent
behavior is usually referred to as the leader's "style."
Defining the Situation
The role building model presented assumes that the appointed leader
will make an analysis of the situation and attempt to alter certain
facets to increase his/her control. In particular, the leader's initial
self-presentation is seen as an attempt to generate certain expectations
that will foster attributions of competence, credibility and legitimacy.
Such favorable attributions should increase the leader's ability to
influence the behavior of his/her subordinates toward the accomplishment
of prescribed goals.
In addition to the expectations of subordinates, other situational
variables also have an impact on the leader's behavior. Korten (1962),
for example, proposes that variations in stress will alter the acceptability
of different styles of leadership. He contends that in high stress
situations there is pressure for a clarification of goals and of goal
paths (means to attain goals). He goes on to state that:
...the more compelling and/or the more clearly structured
the goal, the greater will be the desire to take a direct
approach to the attainment of the goal. Pleasant socializing
is replaced with more intense emphasis on achievement. (p. 225)
Such a "direct approach" to task achievement he labels authoritarian
leadership which he defines as centralized decision making. In less
stressful situations, or when a task is well in hand, Korten suggests
that groups prefer a more democratic or person-oriented rather than
Korten's model, however, considers only one side of the leadership
effectiveness question. According to Fiedler (1974), certain types of
leaders will behave differently in a given situation. In describing
his Contingency Model, Fiedler states:
The theory holds that the effectiveness of the task group
or an organization depends on two main factors: the
personality of the leader and the degree to which the
situation gives the leader power, control and influence
over the situation or, conversely, the degree to which
the situation confronts the leader with uncertainty.
Accurately estimating the uncertainty and perceiving the demands of the
situation is a challenge to even the most socially adroit. For the
newly appointed leader this challenge is particularly salient for at
no other time are the many variables effecting the leadership provess
less under the leader's control than in the early stages of his/her
association with a group. It is at this point that the interaction
between a leader's personality and the situation is particularly
evident. Not only immediate reactive behavior of others but future
effectiveness as well may depend on the leader's ability to rapidly
gain control of the situation and reduce the level of uncertainty for
him/herself and for others by establishing credibility and legitimacy.
A carefully considered self-presentation is seen as the first step in
defining the situation and building the parameters of one's role.
In summary, a role-making model of leadership has been presented
that incorporates the concepts of social power, social exchange, leadership
functions, and situational pressures. The initial phase of an encounter
between the appointed leader and the group is a period of testing. Both
the group environment and the other members of the group are investigated
and evaluated. Members attempts to clarify their roles and offer their
expectations for the leader, who in turn attempts to define the situation
in terms favorable to his/her "style" of leadership. The result is a
working consensus upon which the group builds its subsequent interactions.
PREDICTORS OF SELF-PRESENTATION
For the formal organization, the selection and placement of
leaders and managers is a crucial problem. Fiedler (1974) contends
that leaders should be matched to the situation in which they must
function. This match may be achieved by altering the situation to
fit the leader; by altering the leader to fit the situation; or by
selecting leaders who by personality and motivation are suited to the
existing situation. This latter course of action requires the use of
some means of analyzing the leader and the situation. A number of
models have been devised and tested to varying degrees in an effort
to uncover a reasonable predictor of leadership performance and
effectiveness. However, none has directly addressed the leader's self-
presentational behavior. The capability to predict the "live" or
"face" a leader is likely to assume for him/herself in a given situation
would be a forward step in the interpretation and explanation of
differential leadership performance. If we could predict the most
likely self-presentation and judge its effectiveness in fostering
attributions of competence and credibility, then we might well be able
to help potential leaders increase their impression management skills.
Three scales were selected as potential predictors of self-
presentational emphasis--Johnson and Johnson's (1975) Leadership Behavior
Survey, Fiedler's (1967) Least Preferred Co-Worker (LPC) Scale, and
Ziller's (1973) Self-Other Orientation Scale. In addition to their
differences in format, the scales also differ in the object rated.
The Behavior Survey consists of 20 statements describing behavior in
group settings. The items are equally divided between task functions
and group maintenance functions. The respondent is instructed to rate
him/herself on each item so as to reflect how he/she usually behaves
in groups. A five point frequency scale (always to never) is used. A
total score for each of the functional groupings (task and maintenance)
is computed and these scores are then used as "coordinates" to permit
plotting of the individual's self-report on the "task-maintenance grid."
Johnson and Johnson assume this placement is a veridical indicator of
the individual's leadership style. A (1,1) person is inactive and non-
directive; a (1,10) person places high value on maintaining good inter-
personal relations within the group; a (10,1) person places maximum
emphasis on task accomplishment with minimum concern for group main-
tenance; a (5,5) person, sitting in the center of the grid, balances
task and maintenance needs of the group through compromise and yet
neglects the "creative integration" of these two needs; finally, the
(10,10) person is seen as the "ideal" leader--creative, democratic
and committed to both the task and his/her subordinates.
The LPC scale consists of 16 to 22 bipolar adjectives separated by
an eight-unit equal interval scale. Respondents are requested to think
of the individual with whom they have had the most difficult time working
on a cooperative task and then to describe this individual using the
scale items. Total scores are computed with subjects divided on the
basis of their score into high and low LPC categories or high, middle
and low LPC designations.
The Self-Other Orientation Scale consists of a set of topological
representations on which the respondent is asked to place letters
representing various combinations of persons including self, friends,
parents, etc. The scale is intended as a measure of self-concept and
includes items to assess self-esteem, self-centrality, social interest,
interpersonal power, social marginality (reflecting competing group
demands) and inclusion (identification with a group). Respondents are
scored on each item individually resulting in a composite self-concept
Interpretation of the Scales
Johnson and Johnson's (1975) Leadership Behavior Survey is an
updated form of Blake and Mouton's (1964) Managerial Grid Approach.
It was designed as a behavioral analysis system to facilitate the
teaching of diagnostic skills and behaviors to be used in improving
the effectiveness of a working group under respondent's leadership.
Johnson and Johnson (1975) define leadership as the "performance of acts
*that help the group reach its goals, maintain itself in good working
order, and adapt to changes in the environment, and these acts are
group functions" (p. 22). In support of their approach, Johnson and
Johnson (1975) state, "The functional approach to leadership assumes
that leadership is a learned set of skills that anyone with certain
minimal requirements can acquire," and they contend that an individual's
orientation to leadership can be analyzed and enhanced by training.
The Behavior Survey represents a class of self-descriptive scales
that may be easily manipulated by respondents (Campbell, Dunnette,
Lawler, & Weick, 1970). Its items are all positively stated and each
represents a behavior that is, by definition, desirable for the "ideal
leader." As such, responses'may well reflect an idealized rating rather
than the assumed veridical perception of the individuals' actual
behavior in groups.
In contrast to the Behavior Survey, the LPC scale is not readily
connected to an idealize role description. Whether a person's rating
of the least preferred co-worker is in fact a perception of a single
individual or of a class of people is not in question here. Fiedler
(1972, 1974) considers the LPC scale to be an indicator of a dimension
of an individual's personality. Specifically, Fiedler considers the
LPC score to be "an index of a goal or motivational hierarchy" (p. 11,
1971) with a high score indicative of a social relationship motivated
individual and a low score suggesting task achievement motivation. He
contends that for an individual this hierarchy remains relatively stable
with one aspect consistently predominating and cites test-retest
reliability for military leaders of .72 over an 8-month period (Bons,
1974) and .67 for faculty members over a 2-year period (Prothero, 1974).
Fiedler (1971) states that the high LPC person (an individual who
rates his least preferred co-worker high on the majority of the items)
"does indeed have as his basic goal the establishment and maintenance
of close interpersonal relations--he needs to be related and socially
connected to others" (p. 11). If, however, good relations exist in the
group, this primary goal is satisfied and the high LPC individual will
seek satisfaction of secondary goals which Fiedler claims "appear to be
the approval and admiration of others and the attainment of a position
of prominence" (1971, p. 11). The high LPC person will strive to attain
this recognition through concentration on the task-related aspects of
On the other hand, Fiedler (1971) states that the low LPC person
(an individual more willing to criticize another, as indicated by ratings
on the negative side of the dimensions), "has as his major goal the
accomplishment of a task or assignment" (p. 11)...the low LPC person
appears to derive his principle reinforcement and self-esteem from task
accomplishment. However, when the task is under control, the low LPC
person pursues his secondary goal of good interpersonal relations with
his co-workers, believing good relations to be conducive to task
As is implicit in the above explanation, Fiedler does not consider
the LPC scale to be an independent predictor of leadership behavior. In
contrast to Johnson and Johnson's (1975) assumption of an "ideal leader,"
Fiedler (1974) states that "it is not accurate to speak of a 'good' or a
'poor' leader, rather, a leader may perform well in one type situation
but not in another" (p. 71). The behavior of a leader is, for Fiedler,
a product of the interaction of that individual's motivational system
and demands of the situation.
Ziller's (1973) Self-Other Orientation Scale is ideally suited to
an analysis of the self-concept as it relates to leadership behavior.
Like the LPC scale, it is not an obvious self-descriptive measure and
therefore less prone to conscious manipulation and social desirability
influences. Items included in the scale are assumed to tap many
dimensions of one's interpersonal orientation that might be related to
the role building process and to behavior in groups in general.
Epstein (1973) defines the self-concept as "a dynamic organization
that changes with experience" (p. 402) and as "a subsystem of internally
consistent hierarchically organized concepts contained within a broader
conceptual system" (Epstein,. 1973, p. 407). It is proposed that pro-
minent among these concpets are the social roles that the individual
has accepted as characteristic of himself and which act as anchoring points
for the individual in social interaction. Similarly, Ziller (1964) views
the self-concept as a product of a life-long socialization process which
reflects the constant conflict between the needs of one's social group
and one's own need satisfaction. For a person in a leadership position,
the competing demands of self and group are particularly pronounced.
Leadership success is judged primarily on the achievement of group goals,
as imposed by the organization and on the satisfaction of subordinates'
needs; not on the leader's own level of need satisfaction. The reported
self-concept is seen as a reflection of the degree to which one has been
able to cope with the competing demands in social life.
Hypotheses Relating the Predictor Scales
Hypothesis One An inverse relationship exists between LPC score
and those aspects of the self-concept that reflect a positive regard
for self and a sense of control over the environment. Specifically, it
is predicted that (1) high self-esteem, (2) majority identification, (3)
high interpersonal power, and (4) high social interest will be more
characteristic of low LPC persons than of high LPC persons. As a
corollary to Hypothesis One, a direct relationship is predicted between
LPC score and measures of marginality, self-centrality and inclusion.
Evidence in support of these predictions is largely tangential as
these constructs have not previously been directly compared to the best
knowledge of this writer. Research supporting the hypothesis as it
concerns each of the above constructs will be presented in turn.
Self-esteem. Wylie (1968) indicates that all recent personality
theories place importance on "self-referent constructs." The term
"self" is used to refer both to an "agent," "doer," or "mover" of
behavior, and as the object of the person's own knowledge and evaluations.
As a "mover" or motivational agent, however, the self-concept may also
reflect what Fiedler (1971) defined as the individual's motivational
hierarchy tapped by the LPC scale. The construct of such a motivational
hierarchy may best be defined as a manifestation of the individual's
past history of reinforcement (success, satisfaction, gratification)
in groups, (e.g., Epstein, 1973; Hilgard, 1949). It may be inferred
that an individual will emphasize those aspects of a situation that
have been rewarding in the past, or those he/she perceives as most likely
to result in future positive reinforcement. As has been noted, Fiedler's
(1971, 1974) motivational system dichotomy parallels the generally
accepted functional division of leadership behaviors into task and
maintenance activities. Fiedler infers from his many studies that those
scoring high on the LPC scale may be classified as "relationship motivated."
These individuals have as their "basic goal the establishment of close
interpersonal relations" (Fiedler, 1971, p. 11) and manifest a need to
feel socially connected to others. Fiedler (1974) states that this
basic goal is particularly apparent in uncertain and anxiety-provoking
situations. Under such conditions the high LPC person "will seek out others
and solicit their support" (p. 65). The high LPC person's secondary
motivation, which becomes apparent in more secure and relaxed conditions,
comes from a need for the esteem and admiration of others, especially
from immediate superiors.
In contrast to the high LPC person, the low LPC person is seen as
principally motivated by task achievement as a tangible demonstration
of his or her worth. In uncertain and anxiety-provoking situations,
the low LPC person concentrates on completing the task. In situations
where the task is reasonably under control, the low LPC person's secondary
goals reflect concern for subordinate's feelings and satisfactions. As
Fiedler (1974) phrases it, for the low LPC leader it is "business before
pleasure, but business with pleasure whenever possible" (p. 66).
Given the above conceptualization, the satisfaction of one's basic
goals or needs in an interpersonal setting should be supportive to one's
self-concept and maintain or enhance one's self-esteem. Threats to the
achievement of one's basic goal or need must be overcome if self-esteem
is to be maintained. For a leader whose primary motivation comes from
achieving good interpersonal relations, i.e., a high LPC person, the
foundations of his/her self-concept would be belief and confidence in
his/her social/human relations skills. For a leader whose primary
motivation is task achievement, i.e., the low LPC person, his/her
self-concept anchors should be in task-related skills proven by
experience to result in positive reinforcement, that is, task or
Ziller (1973) states that persons with high self-esteem are less
dependent on external sources of reinforcement. High self-esteem is
related to the integration of the self-system and Ziller sees this
integration as providing the individual with a degree of insulation
from environmental pressures. In contrast, the low self-esteem person
does "not possess a well-developed conceptual buffer for evaluative
stimuli" (p. 4). Ziller (1973) proposes that "Low self-esteem is
associated with short term adaptation...whereas high self-esteem is
associated with long range adaptation" (p. 6).
Fiedler, Bons, and Hastings (1975) report on the results of a
number of studies which indicate (given an initially moderately
favorable situation) that the performance of low LPC leaders improve
with experience on the job. These authors report positive correlations
between performance and experience for task-motivated leaders but a
negative relationship for relationship-motivated leaders. They state:
This suggests that relationship-motivated leaders who
were moved more rapidly from job to job performed
better than did those who remained on the same job
longer. Total experience on the job, and in this
case, no doubt, in similar jobs, enhanced the
performance on the task-motivated leader. (p. 241).
It would appear therefore that low LPC leaders made their adaptations for
the long term whereas high LPC leaders reflect a short term adaptation
to the job environment.
As has been noted, both high and low LPC persons are seen as placing
high value on establishing good interpersonal relations. However, in
the case of the high LPC person, acceptance by others is predominant
and central to their feelings of competence. Fiedler (1972) reports
on a study by Bishop (1964) in which the reactions of high and low LPC
leaders to success and failure in interpersonal relations and assigned
tasks were measured. Three independent variables were scored from
self-reports: adjustment, operationalized as the individual's
satisfaction with the group and their reported anxiety and tension;
subjective interpersonal success ratings; and subjective task success
as measured by ratings of past and future group performance. Fiedler
Bishop's study showed that the adjustment of high LPC
persons increases as a function of experiencing inter-
personal success, while the adjustment of low LPC
persons increases as a function of experiencing
task success. (p. 398).
The high LPC leader's need to be socially connected to others may
be most clearly reflected in his/her reluctance to criticize even the
least preferred co-worker. Wylie suggests that an individual 's
evaluations of others are a positive function of one's own level of
self-evaluation. Rogers (1951) found in a clinical setting that people
with high self-regard tend to be relatively accepting of others. Dittes
(1959), however, argues from laboratory research that persons low in
self-esteem have a greater need for approval from others, will be more
threatened by disapproval, and thus show a strong tendency to report
liking people who accept them and dislike those who reject them.
Walster (1965) reported similar findings. These discrepant results
and interpretations were reviewed by Berscheid and Walster (1969).
These authors concluded the variations were the result of situational
differences. Experimenters, in an effort to exert control, make the
evaluations of others clear to their subjects. Clinicians, in contrast,
work in situations where their clients are not sure whether others
accept or reject them. This ambiguous interpersonal situation is
probably the norm in most "real life" situations where "control" is
exercised by all parties to some degree, and where our social norms
of courtesy, tact, and emotional control often conceal honest feelings.
Under such ambiguous conditions as exist when people first meet, people
with high self-esteem are more likely to assume that others like them
and such assumptions lead to reciprocation of liking. Persons low in
self-esteem will tend not to make such assumptions of liking and may
even expect to be disliked. Thus they may be defensive and distant
or they may work hard to "being liked."
The above analysis permits the inference that a person with low
self-esteem is more likely to be a high LPC person. Given Dittes'
(1959) interpretation of a low self-esteem person's need for approval
and Berscheid and Walster's (1969) belief that this same person will
not expect to be liked at the outset of an interaction, it is possible
to see how the leader with lower self-esteem might concentrate his/her
efforts on improving interpersonal relations as a first step in securing
his/her leadership position. In contrast, the low LPC person may well
be high in self-esteem, being confident in his/her abilities on both
the task and, when the opportunity arises, in interpersonal relations.
.The low LPC person is able to criticize others since he/she is secure
in work group situations and sees no threat to his/her own self-concept
from the act of criticizing another.
Majority identification. Ziller (1973a) sees majority and minority
identification as alternate rather than competing or mutually exclusive
coping mechanisms. He states "Identification with the majority is
presumed to be associated with a sense of control over the environment"
(p. 15) and proposes that "majority identification and self-esteem may
be directly associated" (p. 15). Identification with the minority may
reflect a degree of personal distinction or feeling of uniqueness.
Ziller (1973a) reports majority identification as positively related
to low social interest (r = 0.36). Despite their apparent similarity,
majority identification is related to low inclusion (r = 0.24).
Interpersonal power. As this item is presented, a high score
indicates a low sense of personal power vis a vis others in a social
context. Ziller (1973a) reports strong positive correlations.for
power with low social interest (r = 0.50), low inclusion (r = 0.42),
majority identification (r = .33) and low self-centrality (r = .37).
The low LPC leader who, as Fiedler (1967, 1972) reports, performs well
in highly unfavorable situations (i.e., poor leader-follower relations,
unstructured task and low position power) must draw on a personal sense
of power to obtain compliance from the group. In contrast, the high
LPC leader performs better in situations that provide a moderate degree
of favorability for the leader. Such situations, it may be seen,
provide a power base of some sort for the leader either in existing
good interpersonal relations or from a structured task and strong position
power (Fiedler's Octants IV and V of the contingency model). Given these
results, it is inferred that feeling of low personal power (high score
on the items) will be more representative of high LPC persons.
Social interest. Ziller (1973a) reports a low negative correlation
(r = .09) between low self-esteem and social interest. He describes
social interest as the tendency to include the self with others rather
than holding oneself apart from others. Bass, Fiedler and Kreuger
(1964) report that responses to the Bass Personality Orientation
Inventory (Bass, 1961) revealed that high LPC persons were more self-
oriented than low LPC persons. High LPC persons described themselves
as seeking self-enhancement, positions of prominence, and public
recognition for their achievements. This would appear to reflect a
need to be apart from and above rather than included in the group as
Marginality. This construct is designed to measure feeling of
being a "marginal man," the person "who stands on the boundary between
two groups" (Ziller, 1973a, p. 18), and is unable or unwilling to fully
associate himself with either. Likert's (1967) concept of the linking
pin function of leaders is a positive manifestation of this boundary
position. The organizationally appointed leader, in particular, has
two competing loyalties. He must insure the achievement of directed
goals but must also respond to the needs of his subordinates. Fiedler
(1972) sees the low LPC leader as handling this role-conflict in a
reasonable manner as a result of his fostering of good interpersonal
relations as a support to task achievement. In contrast, the high LPC
leader finds the demands for task achievement and group maintenance to
be incompatible. It is proposed, therefore, that the marginality
responses will be more characteristic of high than of low LPC leaders.
Self-centrality. Ziller (1973a) defines this construct as "the
perception of the social environment largely from the point of view
of the perceiver" (p. 7). Again, the report Ly Bass, Fiedler and
Frueger (1964) of the self-orientation of the high LPC leader is cited
in support of the prediction of a direct relationship. Ziller (1973b)
reports that children who moved frequently between communities exhibited
greater self-centrality than did those who lived in the same community
throughout their life. These responses may be tangentially related to
the better performance of high LPC leaders who changed jobs frequently.
Frequent changes in envirnoment may require a self-orientation which
will permit a degree of evaluative consistency.
In support of the implied inverse relationship between self-
centrality and self-esteem in its above predictions, Ziller (1973b)
reports that individuals rated as sociometric isolates placed themselves
in the central position more frequently than did sociometric stars.
Inclusion. Ziller (1973a) considers his measure of inclusion to
tap the individual's perception of whether or not a relationship exists
between self and others. The person low in inclusion is seen as wishing
to avoid obligation and intimate social contact. While in the extreme
such a motivation would be detrimental to leadership, the task-motivated
leader may be less inclined to establish emotional ties between self
and followers, at least in the early stages of group activity. However,
Ziller (1973a) reports a strong positive correlation between inclusion
and social interest (r = 0.45) which has been predicted to be more
representative of the low LPC leader.
Hypothesis Two Situation free self-descriptions, while generally
favorable, will reflect an emphasis on that functional dimension of
leadership that corresponds to the individual's secondary motivations
as identified by the interpretation of LPC Scale scores.
On the basis of this hypothesis, it is predicted that (1) low
LPC leaders will score themselves as "always" or "frequently" performing
most of the behaviors described in the Leadership Behavior Survey,
thereby achieving a high total score on both dimensions of task and
interpersonal relations. However, low LPC leaders will emphasize the
latter over the former; (2) high LPC leaders will score themselves as
performing these behaviors less frequently but will score themselves
higher on the task dimension than on the interpersonal relations
Fiedler and Chemers (1974) propose that, even in laboratory situations,
leadership is a highly ego-involving activity. As a result, when a person
is asked to describe his or her past leadership behavior, the individual
is likely to render a reasonably favorable report, especially in situations
where his/her claims are not likely to be directly challenged. As pre-
sented, the Johnson and Johnson (1975) Leadership Behavior Survey is a
non-situationally-related self-report which Fiedler (1972) suggests is
likely to reflect idealized behavioral descriptions. Such descriptions,
Fiedler states, are often found to be inconsistent with actual behavior
observed under controlled conditions. These inconsistencies can be
resolved when the interaction between the leader's motivational system
and the situation is considered.
Fiedler (1972) states that on self-report measures low LPC persons
tend to describe themselves as concerned with developing or maintaining
good relations, while high LPC persons tend to describe themselves as
self-oriented, that is, concerned with attaining positions of promi-
nence..." (p. 392). Such self-descriptions appear to reflect how these
individuals behave in favorable situations, while experimental situations
may, in fact, be seen as unfavorable. Fiedler states:
Finally, we find a complex interaction between LPC and
situational favorableness in determining interpersonal
behavior. In favorable situations in which the leader
has a relatively high degree of influence and control
high LPC leaders behave in a task-relevant manner,
while low LPC persons are seer as considerate and
concerned with good interpersonal relations. The
opposite is the case in unfavorable situations. (p. 392).
Fiedler presents a table to summarize the relationships which is
reproduced as Table 1. In explaining Table 1, Fiedler (1972) proposes
An individual's self-description is more likely to
reflect how he sees himself in situations in which
he has relatively great control and influence, i.e.,
when he is secure, poised, and sure of himself, than
when he is insecure, anxious, and humbling. In
these secure situations, he is likely to seek
attainment of his secondary, as well as, his primary
goals and motives. (p. 394).
He points out that experimental conditions are designed to involve
unknown, uncontrolled (by the subject) elements that place the individual
under stress (as, of course, is the case in many non-experimental
leadership situations as well). Add to this situational stress the
evaluation apprehension present in the lab setting, and the situation
becomes relatively insecure for the participants. As a result, the
individual is likely to behave in a manner that will insure attainment
of his/her primary goals. That is, the high LPC person generally
concentrates on improving interpersonal relations while the low LPC
person concentrates on task accomplishment.
The prediction that low LPC leaders will be more likely to describe
themselves as highly active on both dimensions is based on the mutually
supporting nature of the low LPC person's primary and secondary moti-
vations. In highly favorable situations (i.e., Octant I of the
Contingency Model) low LPC leaders have been found to be most effective
(Fiedler, 1967, 1972). In such situations, the task is "under control
and the low LPC leader concentrates his efforts on maintaining group
cohesiveness and improving working relationships. Further support for
this prediction is provided by a study of combat engineer squads by
SYSTEMATIC PRESENTATION OF THE RELATIONSHIP OF LPC AND
LEADER BEHAVIOR AS DESCRIBED IN FAVORABLE AND
UNFAVORABLE SITUATIONS (FROM FIEDLER, 1972)
High LPC Concerned with
Low LPC Concerned with
*Note.--Presumes situation is
Description by Others
Favorable Situation Unfavorable Situation
Concerned with Concerned with
Concerned with Concentrates on
GOOD RELATIONS TASK
favorable to the leader.
Julian, Bishop, and Fiedler (1966). Results indicated that the high
LPC leader is at a disadvantage when forced by events to perform both
leadership roles as socio-emotional leader and as the initiator of
structure. In contrast, the low LPC leader was able to fulfill both
roles and perform both functions with relative ease. Fiedler (1972)
in reviewing this study summarized the relationship in the following
Thus, the high LPC leader, who attends to the task, does
so at the cost of poor relations with his group. The
low LPC leader can, however, be both task-oriented as
well as having good relations with his unit. (p. 399).
Two studies provide direct support for the predictions. Bass,
Fiedler and Kreuger (1964) report on one of the few self-report measures
to yield a reliable relationship with LPC, the Bass Personality Orien-
tation Inventory (Bass, 1961). They found, contrary to their pre-
dictions, that high LPC persons were more self-oriented than low LPC
persons, while low LPC persons showed a distinct interaction orien-
tation. High LPC persons described themselves as seeking self-
enhancement, positions of prominence, and public recognition of their
achievements. Good personal relations and social interaction were
identified most frequently by low LPC persons as their principle
objectives. Fiedler (1972) relates an unpublished study by S. M.
Nealey in 1968 in which college students were asked questions con-
cerning a list of various traits or personality adjectives. High LPC
students rated more task-relevant adjectives as important than did low
LPC students, while low LPC students rated the relationship relevant
items as more important. When asked to imagine themselves as the
foreman of a work group in serious difficulties, Nealey's high LPC
subjects predominantly said they would concentrate on the task, while
low LPC subjects reported they would concentrate on interpersonal
relations. These responses do not correspond with the observed behavior
of high and low LPC leaders in actual stressful situations. It appears
that in this case, imagining oneself in a difficult or stressful
situation was not sufficient to precipitate self-reports of intended
behavior that correspond to reality and to the individual's primary
goal motivation as measured by the LPC scale.
Hypothesis Three There is a direct relationship between the level
of self-esteem and the overall favorability of one's non-situational
self-description when these measurements are made at the same time.
However, recognizing the impact of aspirations, a leader who perceived the
ideal role performance for a leader,as primarily either task direction
or group maintenance, may also have high self-esteem. It is predicted
that (1) self-esteem will correlate positively with overall scores on
the Behavior Survey, and (2) that those persons strongly emphasizing
one dimension will be found to report high self-esteem.
William James (1950) proposed that self-esteem equals the ratio of
success over pretensions (or aspirations). Likewise, self-descriptions
are reflections of one's behavior in successful encounters in the past,
for as Heider (1958) suggests, people will attribute favorable char-
acteristics to themselves whenever the situation will allow. For
aspiring leaders, the items on the Behavior Survey constitute a series
of behaviors that may be valued differently as a result of past training
or experience. However, all are positively stated and appear to be
appropriate behavior for leaders.
Rosenberg (1968) proposes that a person's "global self-esteem is
not based solely on his assessment of his constituent qualities; it is
based on his self-assessment of qualities that count" (p. 339). Based
on the concept of psychological selectivity, Rosenberg proposes that a
person will come "to value those things at which he considers himself
good and to disvalue those qualities at which he considers himself
poor" (p. 340). Perceiving oneself to be an "ideal" leader (in Johnson
and Johnson, 1975, terms a 10, 10 leader) and therefore valuing such
role behavior should in the absence of conflicting evidence, be
supportive of high self-esteem.
SELF-PRESENTATION AND ROLE BUILDING
To achieve a desired status or position within a group, an individual
must demonstrate certain aspects of self seen as appropriate to assumption
of that role by the group. In some cases, an ascribed status or bio-
logical characteristic, such as ethnic group or sex, is sufficient to
either foster acceptance or precipitate rejection. In other cases, an
achieved status or complex of past performance is required, such as an
academic degree or a number of years of experience. Argyle and Kendon
(1967) state that a considerable amount and variety of research has been
directed at specific social skills related to perceptions of competence.
Frequently such research has compared "good" and "bad" performers on
some objective set of outcome criteria, for example, high and low pro-
ductivity (Argyle, Gardner, & Cioffi, 1958). This research, however,
has not touched on the self-presentational processes employed by the
leaders of the groups evaluated. These authors propose that achievement
of the state of competence in professional and social skills involves the
successful attainment of some combination of group or task related goals.
Acknowledging that different social tasks require different skills, they
suggest some common ones, i.e., perceptual sensitivity, warmth, flexibility,
energy, and a large repertoire of social techniques that may be cross-
situational for any position of social prominence.
A number of theoretical frameworks have been reviewed in an effort
to establish a foundation for a role-building/self-presentational model
of leadership in which the aspiring leader's selective presentation of
his or her "credentials" is seen as the foundation upon which that leader
builds his or her efforts toward acceptance, cerdibility and authority
within the group. As has been suggested, self-presentations must be
fitted to the demands of the situation if they are to serve the leader's
ends. Based on his or her past history of success (as measured by the
predictor scales) and his or her perception of the situational demands
to include follower expectations, (as manipulated in the experimental
situation) it has been proposed that the leader will adjust his initial
self-presentation to foster perceptions of competence. This initial self-
description is seen as an integral part of the Sampling Phase of Graen and
Schiemann's (1976) role making process model. It is at this point that
the leader begins to establish the "working consensus" (Goffman, 1959)
between his perceptions of the leadership role and his follower's expec-
tations for one filling the role. As part of this sampling process, it
is assumed that the leader will attempt to lay the foundation for later
recognition of referent and expert power by selective claims of skills,
abilities and personal characteristic felt to be both appropriate to the
situation and fully supportable by later behavior. In natural groups,
the appointed leader's self-introduction to the members of his/her new
work group is one opportunity to make these selective claims.
Rosenberg (1968) reviews the construct of "psychological selectivity"
and its relation to the self-concept (particularly self-esteem) and social
behavior. He suggests that selectivity:
...is particularly free to operate under two conditions:
(1) where the situation is unstructured and ambiguous,
and (2) where the range of options is wide. These
conditions are particularly characteristic of self-
evaluations. (p. 339).
The initial phase of a person's tenure in a leadership position is
a period of self-evaluation and can be seen to reflect both of these
conditions. Therefore, it is likely that new leaders will strive to
maintain self-esteem and gain recognition through a carefully monitored
self-presentation. This assumption is supported by studies conducted
by Snyder (1974) who proposes that "Individuals differ in the extent to
which they monitor (observe and control) their expressive behavior and
self-presentation" (p. 536). His results indicate that individuals
scoring high on his "Self-Monitoring Scale" were seen by peers as more
self-controlled and better able to create the impressions they see as
Measures of Self-Presentation
The self-presentational device employed in this study is a modifi-
cation of one devised by Schlenker (1975) and included in a "personal
information questionnaire." Schlenker explained the purpose of this
questionnaire to his subjects by emphasizing that a feeling of "groupness"
would develop more rapidly if the members exchanged personal information.
Subjects were led to believe that after completing the questionnaire in
private, they would sit around a circular table in the middle of the room
and read one another their responses. One of the pages of the questionnaire
contained the self-presentational measure. It consisted of 35 personal
attribute items, each of which was to be rated on a five point scale from
"completely representative" to "completely unrepresentative."
Schlenker considers his scale of 35 attributes to be a "behavioroid"
measure of self-presentation rather than merely a self-report form when
it is used in situations where subjects expect to be required to read
their self-evaluations to their group prior to beginning the task. As
such, he assumed it is a measure of how a person actually expects to act
in the upcoming situation rather than some idealized description of how
one would like to act in such a situation or how one would like people
to believe we will behave.
Because of the prominence of his/her position in the group, a leader's
self-presentation is constantly subject to the evaluations of both group
members and those outside the group but observing the group in action.
This being the case, a leader, provided an opportunity like Schlenker's
(1975) subjects to make a verbal self-presentation prior to the beginning
of the group task, is likely to exercise care in describing his personal
attributes to the group. Newly appointed, assigned, or elected leaders
are commonly provided this opportunity to make a self-introduction in
non-laboratory settings. Therefore, it is a logical extension of Schlenker's
design to provide subjects expecting to lead a problem solving group with
a structured self-presentational form and then provide them the opportunity
to actually present that self-description to the group they would sub-
sequently lead (an opportunity not provided to Schlenker's subjects). The
final form used in this study consists of 33 personal attribute items.
All are value-laden, culturally biased terms, selected from both Schlenker's
(1975) scale and listings of "leadership traits" compiled by Stogdill (1974).
Each item was rated on a five point scale of "representativeness."
This form, labeled "Information Exchange Worksheet" is a principle
dependent measure of this study. Its use is supported by the central
assumption that "first impressions" can and will be manipulated by those
who are aware of their impact on the behavior of others especially during
initial encounters. It is proposed that this structured self-presentation
is the experimental equivalent of an organizational leader's self-
introduction to the members of his/her new work group.
In describing the process of self-presentation/impression management,
Goffman (1955) uses the term "face" which he defines as "the positive
social value a person effectively claims for himself by the line others
assume he/she has taken during a particular contact" (p. 214). It is the
totality of the behaviors displayed and the facts and fictions revealed
about him/herself by the actor through his verbal and nonverbal communications.
For Goffman (1955), to "have," "be in," or "maintain face," a person's line
must effectively present an image of him/her that is internally consistent.
That is, the image must be supported by judgments and evidence conveyed by
other participants and confirmed by evidence transmitted through the
impersonal agencies of the situation. In effect, maintaining face is the
result of effective, conforming role enactment. Therefore, in this study
(as in life) self-presentation will not simply be considered one's self-
description but is viewed as all of those verbal and nonverbal behaviors
one portrays in fulfilling a social role as perceived by others.
In this regard, a series of questions are included inthe post-
experimental evaluation form to measure (1) perceptions of behavioral
emphasis (task-oriented and relationship-oriented behaviors); (2) the
internal consistency of the individual's self-presentation as measured
by a question of the self-description's predictability for in-group
behavior; (3) and ratings of performance which are viewed as indicators
of the individual's credibility and acceptability for leadership. The
relationships among these various ratings made by peers and in some cases
made about oneself are also investigated.
The relationships between the predictor scales and the individual's
total self-presentation were also assessed. This includes comparisons of
each scale with the structured self-description and with perceived in-role
behavior ratings. These analyses in particular are directed at answering
the first major question of this study concerning the predictability of
Analysis of the Leadership Situation
Prior to addressing the predictions made concerning the prediction and
assessment of self-presentations, the relationship of the situation to
leadership behavior must be reviewed.
Bogardus (1934) presents an interactionist view of leadership that has
only recently gained general acceptance as a result of the work of such
theorists as Fiedler and Hollander and Julian. Considering the situation,
Social situations are never static. They are ever changing;
the idea of process is implicit. Social situations call
now for one set of leadership qualities but tomorrow
perhaps for another set of traits. (p. 278).
Given the dynamic nature of the situation, Bogardus proposes:
The development of leadership depends on studying situations
and on acquiring skill in controlling them. In order to
'learn' leadership, a person analyzes situations and develops
appropriate techniques for controlling them. By anticipating
situations one person may become a leader while other persons
are 'running around in circles.' (p. 269).
As has been noted, Johnson and Johnson's (1975) prescription for the
"ideal" leader foregoes any consideration of the situational variability.
This conceptualization is challenged by Korten, Hershey and Blanchard
and, most prominently, Fiedler.
The Korten Stress Model
Korten (1962) presents a model based on the variables of situational
stress and need for goal structure. Simply stated, under conditions of
high stress there exists a demand for clear goals and a direct approach
to goal achievement. In such a situation, the group requires a leader
who can provide the needed clarity and who will facilitate goal-directed
activity. Korten refers to this type of leader as authoritarian, one who
through an orientation or the task will overcome internal disagreements
and control the group's behavior. Conversely, when the situation is free
of ambiguities and stress, group members prefer a higher degree of personal
control and involvement in decision making. In such an environment, the
leadership style Korten calls "democratic" or participativee" will be
preferred by the group. Korten concludes that by understanding the effects
of the situation on the group, dysfunctional pressures can be avoided or
The Life Cycle Theory
Hershey and Blanchard (1969) propose that the most effective mix of
task and relationship behaviors depends in large measure on the level of
"maturity" of one's subordinates. They define maturity as (1) the capacity
to set high but attainable goals, (2) the willingness and ability to take
responsibility, and (3) a high degree of task related education and/or
experience. With motive subordinates, Hershey and Blanchard propose that
the effective leader will be the one who manifests a low frequency of both
task and relationship behavior toward subordinates. With immature followers,
Hershey and Blanchard's "Life Cycle Theory" suggests that the successful
leader must show a high frequency of task behaviors but minimize rela-
tionship behaviors. With "average maturity" subordinates, the leader
should behave in a high task/high relationship manner until members
achieve a reasonable degree of task competence. At this point to remain
optimally effective the leader should reduce the frequency of task
direction while maintaining a high level of group maintenance behavior.
The Contingency Model
Both of these models contribute to an understanding of the factors
influencing leader behavior. However, unlike Fiedler's Contingency
Model, neither considers the interaction of situational forces with
leader personality. As has been noted, the Contingency Model utilizes
a dimension of situational favorability for the leader composed of three
dichotomous variables: leader-member relations, task and structure,
and position power. Fiedler (1974) considers that this dimension reflects
the degree of control, power and influence over the situation potentially
available to the leader. A situation reflecting good leader-member
relations, a structured task, and strong position power for the leader is.
defined by Fiedler as low in stress (Octant I). A situation lacking all
of these qualities would be high in stress (Octant VIII) while those
characterized by a combination represent gradations of favorability
Low LPC (task-motivated) leaders have been found to be most effective
in highly favorable (low stress) and highly unfavorable (high stress)
situations (Fiedler, 1967). Korten's (1962) prescriptions for leader
behavior match the reported performance of low LPC leaders in these two
types of situations. Under stress, the low LPC leader's primary motivation
of task achievement matches the needs and expectations of the group for
task direction and goal clarification. In non-stress situations, the
low LPC leader's secondary motivation for good interpersonal relations
becomes operational and again this matches group expectations.
Fiedler (1967, 1972) reports that high LPC leaders perform best in
moderately favorable/moderately stressful situations (Octants VI and V).
In Octant VI situations with favorable interpersonal relations secured,
the high LPC leader's primary motivations are satisfied, permitting a
concentration of task-related behaviors. The high LPC leader in such
cases may draw on referrent power derived from his/her relations with
subordinates to overcome a lack of position power. For the high LPC
leader, this type situation may be a low-stress/low anxiety one. Octant
V situations provide for a structured task and strong position power but
relatively poor interpersonal relations. In such a situation, the high
LPC leader's primary motives are threatened indicating a potential for
greater stress and anxiety. To this contingency, the effective high LPC
leader responds with an emphasis on interpersonal relations.
The Experimental Situations
Results of studies considering leadership performance in Octants III
and VII of the Contingency Model have not reflected the clear and con-
sistent types of relationship between LPC designation and effectiveness
that characterize the Octants reviewed above. These Octants, however,
are ideal for a study of organizationally appointed leaders since such
leaders generally have a recognized level of position power yet may often
find themselves faced with an unstructured task. These Octants differ
only on the variable of leader-member relations. Due to the lack of
consistent results from past studies, it was decided to attempt to increase
the stress in the experimental situation through a manipulation of the
instructions provided to the subjects. Variable emphasis on one or the
other or both of the dimensions of leadership was seen as a possible
source of additional stress. It was intended that the prescriptions of
the priority of mission accomplishment or of taking care of one's
subordinates would be perceived as a threat or a support to the individual's
primary motivational orientation. A combined "mission and men" pre-
scription reflecting the actual doctrinal expectations for a military
leader was also used.
Since subjects were to be drawn from freshmen ROTC classes, it was
decided to present each leader with a military tactics problem. The
tactical situations were selected from a group of 18 previously used with
junior and senior ROTC cadets in a study of responsibility attributions
(Caine, 1975) and were evaluated by the members of the ROTC staff and
other commissioned officers in residence at the University of Florida
prior to use. In many ways, these situations and the procedure for solving
them parallel a number of similar tasks analyzed by Shaw (1963). For
example, Shaw's tasks number 53, 56, and 71 all concern military related
problems, where analysis of a number of possible options was required.
Each of these tasks reflects a low scale value on the "decision varifi-
ability" dimension, high scores on both "goal path multiplicity" and
"solution multiplicity" dimensions, and a medium scale value on "goal
clarity" (a dimension considered by Shaw to show considerable variability).
This pattern of scale values on these dimensions fit Fiedler's (1963)
definition of an "unstructured task." IL was felt that the four platoon
tactical situations also fit this definition. To check this belief,
ratings were obtained from 10 officers using the definitions provided
by Shaw (1963) for each of the four tactical situations on the following
dimensions: difficulty, decision verifiability, goal path multiplicity,
intrinsic interest, solution multiplicity, and intellectual-manipulative
requirements. Ratings were on scales from one (low) to eight (high).
Mean ratings are presented in Table 2. In general, the situations do
appear to fit the "unstructured" category.
Strong position power (full decision authority) was built into the
experimental situation. Overall, subjects perceive it as such, as indicated
by their responses on a post-experimental manipulation check question. It
was assumed that at the outset leader-member relations would be essentially
neutral as subjects would be relative strangers. It was assumed that
subjects would strive for good relations. Therefore, post-exercise ratings
of interpersonal relations were taken to indicate the end-state rather
than the initial condition of group relations. Using these responses, it
was possible to determine the Contingency Model Octant that applied to
the experimental situation.
Hypotheses Concerning Situational Self-Presentation
Hypotehsis Four Responses on the Behavioral Survey will be predictive
of structured situational self-presentations and peer ratings of behavior
"on-the-job" only in low stress situations. Based on this, it is predicted
that in the low stress experimental conditions, (1) subjects who report
themselves as frequently performing in both leadership functional dimensions
of the Behavior Survey (high scores on both dimensions) will describe
themselves to others in highly favorable terms, and (2) will be perceived
ANALYSIS OF THE STRUCTURE OF TACTICAL SITUATIONS MEAN RATINGS
Dimension* One Two Three Four
Difficulty 6.3 6.8 5.9 7.1
Decision Verifiability 2.4 3.1 2.2 2.9
Goal-Path Multiplicity 7.0 6.9 7.2 6.5
Intrinsic Interest 5.2 4.9 4.7 5.1
Solution Multiplicity 6.0 6.5 7.1 6.8
Intellectual Manipulation 4.9 5.7 5.3 6.1
*From Shaw, 1963.
as active on both dimensions by peers. Likewise, subjects that report
themselves as performing less frequently on one or both dimensions
(variable scores), will (3) describe themselves in less favorably and
be perceived as differentially active by peers.
If Fiedler's (1972) reasoning that responses to self-report scales
(such as the Behavioral Survey) represent idealized leadership behavior
is correct, then such measures will be meaningful predictors of overt
behavior only in situations favorable to the individual. The above
predictions are supported by evidence previously presented during
consideration of the predictor scales (see Chapter II). However, it
must be noted that the consistency of what might be considered a measure
of behavioral intentions (the Behavior Survey), with one's later behavior
in a group setting is supported by a study of a scale designed to measure
"intentions to achieve individual prominence" (Shaw, 1960). The scale is
composed of items with obvious meanings related to a desire for individual
prominence and high scorers were found to volunteer for the prominent
role as group recorder in an experimental situation more frequently than
low scorers. In this case, the obvious self-report was a valid predictor
of later behavior.
Hypothesis Five An individual's predictive self-presentation and
group behavior are the products of the interaction of one's motivational
hierarchy and the degree of situational stress. It is predicted that (1)
high LPC leader will emphasize interpersonal relations traits in high
stress situations and task competence traits in _DW stress situations;
(2) low LPC leaders will emphasize interpersonal realtions traits in low
stress situations and task competence traits in high stress situations.
For this study, high stress was presumed to result from instructional
emphasis on the functional dimension corresponding to the leader's
secondary motivation. As has been noted previously, it is assumed that
emphasis on both functional dimensions is seen as a favorable situation
for the low LPC leader and as unfavorable for the high LPC leader.
Therefore, it is predicted that in the "Both" experimental condition
high LPC leaders will emphasize interpersonal relations traits while
low LPC leaders will present a highly favorable and balanced self-
description. These predictions are summaried in Table 3.
As a correllary to Hypothesis Five, it is proposed that overall
favorability of self-presentations will differ as a result of the
personality situation interaction. It is predicted that low LPC subjects
in the "Both" condition will present themselves most favorably followed
by low LPC leaders receiving the interpersonal relations manipulation.
The high LPC leaders in the "Both" condition are expected to present
themselves least favorably.
These predictions are based onthe assumption that individuals assigned
to a leadership position will attempt to establish their credibility and
suitability for that position by selectively presenting those "credentials"
that will permit them to make a favorable "first impression" on others
and to control the definition of their role and the group situation in
general. In Goffman's (1955) terms, this self-description is the. first
indication of the "face" that a leader projects to the group. As "the
positive social value a person effectively claims for himself by the line
others assume he/she has taken during a particular contact" (Goffman, 1955,
p. 214), one's social "face" should reflect those characteristics of self
PREDICTED EFFECTS ON SELF-PRESENTATION OF THE LPC
DESIGNATION BY FUNCTIONAL DIMENSION
LPC Designation Emphasis Condition (Manipulation)
Task IPR Both
Degree of Low Stress High Stress Low Stress
tS s A uetm d
Degtres Asumed High Stress Low Stress High Stress
Self-Presentational IPR Task IPR
of which the individual sees as appropriate to the situation. Jones,
Gergen, and Jones (1963) found that a person's goal in an interaction
will affect the type of self-presentation he/she will make. If one's
goal is to be liked (for example by a subordinate), modesty may be the
best course of action, while if one's goal is to impress another, self-
enhancement, especially on skills or attributes unlikely to be challenged,
may be the course to select. Therefore, given an opportunity to present
a self-description to another at the outset of a face to face interaction,
the individual must decide what claims to make to build the most favorable
expectations on the part of the other. These claims must be made care-
fully. The individual must feel capable of meeting the expectations he/she
creates, or risk what Goffman (1955) calls "loss of face," or embarrassment
and loss of credibility.
Goffman (1959) describes this process in the following way:
The expressiveness of the individual (and therefore his
capacity to give impressions) appears to involve two
radically different kinds of sign activity: the
expression that he gives and the expression that he
lives off. The first involves verbal symbols or their
substitutes... .The second involves a wide range of
action that others can treat as symptomatic of the
actor, the expectation being that the action was
performed for reasons other than the information
conveyed in this way. (p. 2).
The impact of the situation on self-presentation can be seen in the results
of two recent studies. Schneider (1969) investigated the effects of
success and failure on self-presentation. Schneider's procedure centered
on a subject's interview with a clinical psychology graduate student. When
subjects expected evaluative feedback from the interviewer, those who had
experienced success were quite modest in their claims of personal attributes.
This was interpreted as an effort to preserve the positive evaluation
they had previously received. Subjects that failed and expected feed-
back presented themselves quite positively. This was seen as an effort
to regain some self-esteem lost as a result of their performance failure.
It seems likely that situations that appear to threaten the leader's
chances of success, a similar effect might be found.
Schlenker (1975) also utilized a success/failure manipulation to
analyze self-presentation. Subjects completed a task and then received
bogus feedback on that task leading to expectations of doing extremely
well or very poorly. A control group was given no information on which
to form a specific performance expectation. Half of the subjects were led
to believe that the quality of their future performance would be unknown
to the other group members. The other half of the subjects were led to
expect their performance to be public. Once these conditions were
established, participants were asked to complete a personal information
questionnaire containing the self-presentational measure.
Schlenker reports that both performance expectations and anonymity/
public manipulations were successful. A factor analysis of the self-
presentational data produced two factors. An analysis of variance revealed
a reliable anonymity by performance expectations interaction for Factor One,
competence. Subjects in the public condition presented themselves con-
sistently with their performance expectations, that is, when expecting
success they presented more positively than when expecting failure. There
was no difference by expectation for the anonymous condition with all self-
presentations fairly positive. The no feedback public condition only
differed from the failure public condition. Schlenker concludes: "A lack
of certainty about their upcoming performance coupled with hopes for
success probably explains the fairly positive self-presentations of the
no feedback-public condition subjects" (p. 10).
Analysis of Factor Two, interpersonal relations, did not produce a
significant anonymity by performance expectations interaction. Given
the conditions of Schlenker's study which eliminated all face to face
interaction during the task by the use of individual booths, public
condition subjects appear to have been less concerned with presenting
"accurate" views of their interpersonal skills. These attributes would
not be subject to refutation as a result of the isolation.
Hypothesis Six There is a direct relationship between the degree
to which one's self-presentation is seen as predictive of later in-role
performance and the evaluation of that performance. Specifically, it is
predicted that peer ratings of leadership performance will be positively
correlated with ratings of the predictability of structured self-
As a corollary to Hypothesis Six, it is proposed that a highly
favorable self-description, if seen as not highly predictable, will
produce strongly negative attributions of effectiveness. As has been
suggested by Goffman (1959) and Argyle and Kendon (1967), having once
established a particular "face," it is in the individual's best interest
to try to meet or exceed the expectations created. The degree of
correspondence (predictability) between one's initial self-presentation
and later behavior will have an impact on how that behavior is evaluated.
Attributions of effectiveness and contribution to the group may well be
lower than actually deserved if one fails to "come up to expectations."
Hypothesis Seven A weak but positive relationship exists between
self-esteem and the overall favorability of one's self-presentation. As
the focal aspect of the self concept, the level of self-esteem should
effect how one presents him/herself. However, the impact of the experi-
mental situation is seen as moderating any direct relationship .that may
exist between self-esteem and the structured self-presentation. Therefore,
while self-esteem may effect self-presentational favorability, the demands
of the situation must be met as well.
A study by Stires (1970) provides indirect support for this hypothesis.
Stires informed the members of certain experimental groups that they
differed in task related ability, while other groups were told that members
did not differ in ability. Individuals who previously expressed confidence
in their ability, attempted to gain leadership of the group through
modesty. Those who reported being uncertain about their ability attempted
to gain the respect of others through self-enhancement.
A total of 53 freshman ROTC cadets at the University of Florida
completed the three predictors scales. Of these, 48 volunteered
to participate in the small group decision making exercise. Cadets
were randomly assigned to groups with the limitation that no two
individuals from the same class or drill unit were permitted in a
group. Each group consisted of two low and two high LPC persons.
Participating cadets were released from a drill period to compensate
Analysis of the relationships among the three predictor scales were
made independent of the experimental procedure. The LPC scale was
selected as the major experimental independent variable on the bases of
its demonstrated relationship to predictions of leadership effectiveness
when applied in the contingency model. The experimental manipulation
of situational stress was included in the instructions to subjects.
Three different descriptions of a leader's responsibilities were
prepared; one emphasizing the task-achievement function; one emphasizing
the group maintenance function; and one consisting of a balanced
presentation emphasizing both functional dimensions. These emphasis
conditions were designed to increase the stress in the experimental
situation for certain leaders identified by their LPC score. The
2 x 3 factorial design was used to test the interactional effect of
LPC and situational stress on self-presentation and leadership
performance. To test certain hypotheses, comparisons of the other
predictor scales (The Behavior Survey and the Self-Other Orientation
Scales) with results derived in the experimental situation were made.
Such comparisons, where possible, consider the total subject pool and/
or specific cells, appropriately identified.
Leadership of the group was rotated among the four members with each
serving as the "platoon leader" for one of the tactical problems presented.
It was felt that such a procedure would increase the intrinsic interest
in the exercise and maintain a high level of involvement by all partici-
pants at all phases. Subjects were required to make forced comparison
performance ratings of their three group peers following the completion
of the problem sequence.
Twelve groups were formed as indicated above and scheduled for a
90 minute period either during the late afternoon or early evening.
Each of three commissioned officers in uniform conducted four groups,
balanced for experimental condition. All sessions were held in the
same classroom in the ROTC Department building.
When the participants arrived, each was given a name tag, instructed
to be seated at one of four separate tables, and asked to remain silent.
After all four participants were seated, the experimenter introduced
himself and explained the purpose and procedures of the exercise.
Cadets were asked to read their copy of the instructions as the
experimenter read them aloud. The instructions began:
You will be participating in a group decision making
exercise involving platoon level tactical problems.
The group will act as if they were a platoon leader
and his three squad leaders. Four situations will be
presented with the group's leadership rotating so that
each of you will be the leader once. The leader will
have full decision authority in selecting the group's
solution to a problem. He may, however, use any method
he wishes to analyze the situation and arrive at that
The instructions included one of three emphasis manipulation
statements prepared with the intent of challenging the participants
with a concentration on one of the two basic leadership functional areas
or with both equally weighted. The text of the task emphasis statement
The Army's leadership manual FM 22-100 states that:
'Our Army has been and must always be mission-oriented.
Thus, our ultimate objective-our primary leadership
goal-must continue to be mission accomplishment.' In
solving the situations to come, mission accomplishment-
that is, finding the best possible solution to the
problem, must be the leader's primary concern.
Following this, it was necessary to justify the making of structured
self-presentations by participants. Borrowing liberally from Schlenker
(in press), the following explanation was given:
In a project of this sort, it is very important that the
persons participating get the feeling that they are members
of a real group, and not simply individuals interacting with
strangers. We will be simulating the decision making structure
of a military unit. In an actual unit, you would know each
other's strengths and weaknesses, but here we have only a
short time to complete the exercise. Therefore, we must
facilitate the exchange of information about each other.
Doing this has been shown to foster a feeling of "groupness"
and we hope that will happen here as well. Your packet
includes a sheet entitled "Information Exchange Worksheet"
and I would like each-of you to read the instructions and
fill out every item on the sheet. Do this as honestly as
you can because when it comes your turn to be the platoon
leader,.you will read your responses to the others. These
items will give both a fairly complete picture of you and
provide a common ground for information exchange. You are
making out the sheets now and will be reading them so that
those who go later in the exercise will not be unduly
influenced by those leading first. Please don't feel you
have to be excessively modest. It is best to describe
The "Information Exchange Worksheet" contained 33 personal descriptive
terms. Items were designed to permit participating individuals to
rate themselves on a wide range of task and interpersonal "traits."
Subjects were instructed to rate how representative each attribute is
of them using a five point scale from highly representative (1) to
completely unrepresentative (5).
When all subjects had completed the worksheet, these were collected
and the group asked to move to a single circular table. Prior to taking
seats, subjects drew sequence numbers to determine the order of leader-
ship appointment. The first platoon leader was instructed to sit in
the position with a place marker bearing a second lieutenant's gold
bar insignia on it. The other group members each selected one of the
other places marked with a squad leader's sergeants stripes. The
platoon leader was given his self-presentation form which he was
instructed to read to the group. When this was completed, the experi-
menter presented a copy of the first tactical problem to each member
of the group and an answer sheet to the platoon leader. He retrieved
the platoon leader's worksheet and stated "You have five minutes to
reach a decision. You may select one of the three courses of action
provided with the problem or propose one of your own design. There
are no absolutely right or wrong answers. No extra time will be
allowed. I will notify you when one minute remains." The experimenter
then left the room.
At the end of the first five minutes, the situation forms and the
solution were collected. Group leadership was then rotated with the
new leader occupying the "leader's chair" (always placed so the leader
faced away from the windows) and other members changing seats also.
The new leader was given his worksheet to read and the process continued
as before. All four situations were completed in the same manner.
Following the fourth situation, subjects were asked to return to
the individual tables. No performance feedback was given at this time.
The post-exercise measures were issued. Part One consisted of peer
ratings, asking for comparative rankings of the other three members on
leadership, followership, overall contribution to the group, and the
degree their self-description matched their actual behavior during the
discussions. Part Two contained a series of items including ratings
of group interpersonal atmosphere, satisfaction with one's own performance
as a leader, feeling of personal responsibility and an open-ended state-
ment of the cadet's conceptualization of the "essence" of leadership.
This final question was intended as a manipulation check on the
emphasis feedback. All but the last item used a 15 point equal interval
scale with end points labeled.
When these measures had been completed, cadets were informed only
of the group's overall performance based on a quick review of solutions.
Cadets were asked not to discuss the procedure or the situations with
others. They were informed that individualized performance feedback on
all the scales and peer evaluations would be provided to each of them
in writing in the near future and that a detailed debriefing would be
conducted after completion of the project.
Predictor Scale Relationships
The comparison of LPC scores with responses on the Self-Other
Orientation Scale produced some support for the Hypothesis One. Overall,
a low but reliable inverse relationship between LPC score and self-
esteem score was observed, r (46) = -.331, p <.025. This is interpreted
as some support for the prediction that high self-esteem would be more
characteristic of low LPC persons. Using a median split for self-esteem
scores, a 2 x 2 distribution matrix was formed with high/low LPC categories.
Assuming an equal distribution, cell frequencies were analyzed by a chi
square test, X2 (1) = 3.64, p <.06. The distribution pattern provides
support for the predicted inverse relationship, with high LPC/low self-
esteem and low LPC/high self-esteem cells predominating.
Predictions of an inverse relationship between LPC score and majority/
minority identificatio- scores (Majority Identification = high score on
items) were also confirmed, r (46) = -.246, p<.05. However, measures
of Social Interest and Personal Power were not found to be reliably related
to LPC scores, r (46) = -.061, N.S. and r (46) = .180, N.S.
No support was found for the corollary to Hypothesis One that sug-
gested a direct relationship between LPC score and measures of marginality,
self-centrality and inclusion. Correlations for these items with LPC were
.13, .06, and .03.
LPC/Behavioral Survey Relationship
Based on Hypothesis Two, it was predicted that those scoring low
on the LPC scale would be more likely to describe themselves as high
on both the task and interpersonal relations dimensions but emphasize
the interpersonal relations dimension while high LPC leaders would rate
themselves more moderately on both but emphasize the task dimension.
One-way analysis of variance comparing high and low LPC leaders on each
dimension separately did not indicate any reliable differences in the
responses, F (1,46) = 1.674, p<.203 for task and F (1,46) = .713, p< .403
for interpersonal relations. However, task dimension scores were found
to correlate reliably with LPC scores, r (46) = .304, p <.025. This is
interpreted as a tendency by high LPC leaders to see themselves as more
active on the task dimension as expected. The relationship between LPC
and the interpersonal relations scores was not significant, r (46) = .176,
The prediction that low LPC leaders would describe themselves as highly
active on both dimensions while high LPC leaders would rate themselves more
moderately received little support. Scores for almost all subjects were
high. Plotting the actual scores on the standard task-maintenance grid,
it was found that all but five subjects were firmly in the upper right
or high task/high maintenance quadrant. Using the approximate means for
all subjects on the two dimensions (36 and 36) a new grid center point
(or 5, 5 point) was superimposed on the data. This converted grid pro-
vides for a more meaningful examination of these data. High LPC leaders
were expected to describe themselves primarily as "compromise" (5,5) or
"total task" group leaders. On the converted grid (Figure 1), high LPC
0 5 10 Is 20 25 30 35
Note.--Triangles = High LPC leaders.
Circles = Low LPC leaders.
Figure 1. Johnson and Johnson's Behavior Survey Grid.
individuals did predominantly fall in the compromise area (n = 9) but
also had reasonable representation in the inactive and highly active
areas (5 and 6 subjects, respectively).
Low LPC leaders were expected to describe themselves as "ideal"
(highly active or 10,10) or as relationship oriented (1,10). In fact,
they frequented the balanced, highly active (10 and 8), with the high
relationship (1,10) quadrant represented by a single individual.
Assuming a uniform distribution, a chi square test performed on the
overall distribution was significant (X2 (1) = 11.324, p<.001), with
the balanced quadrants predominating. On the modified grid 22 subjects
saw themselves as highly active on both dimensions (12 low LPC and 10
high LPC) and 14 rating themselves as less active but still balanced
(7 high and 7 low LPC). A one-way analysis of variance comparing overall
ratings of high and low LPC leaders on the Behavior Survey was performed.
No reliable difference was found, F (1,46) = 1.360, N.S. It must be
concluded that the results provide only limited support for Hypothesis Two.
Self-Esteem/Behavior Survey Relationship
The prediction of a positive relationship between self-esteem and total
self-evaluation on the Behavior Survey was not supported, r (46) = -.051,
N.S. Likewise, neither dimension revealed a reliable relationship, r (46) =
-.062, N. S. for task and r (46) = -.21, N.S. for interpersonal relations.
The second prediction derived from Hypothesis Three was also not supported.
Of the 13 subjects whose dimension scores differed by 5 points or more
(range 48 to 17) 7 were scored as high self-esteem and 6 low self-esteem
by median split of all subjects. Self-esteem scores for this sub-group
ranged from 10 to 23 out of a possible score range of 4 to 24. It should
be noted that the overall mean self-esteem score was 18.29 for the
entire subject pool. This figure corresponds to a single measure mean
of 4.57 in a range of 1 to 6. This grand mean is somewhat higher than
the average self-esteem score of 3.733 reported by Ziller (1973a) for
a group of 298 teachers, supervisors, and principles. The subjects
in the present project appear to be higher in self-esteem as a group
than others previously tested.
Verification of the Contingency Model Situational Dimensions
Analysis of the post-exercise questions concerning the favorability
of interpersonal relations and perceived leader power revealed nearly
universal ratings of favorable relations in groups and strong position
power in the role of platoon leader. As the task was designed to be
unstructured, this places the experimental situation in Octant III of
the Contingency Model. The mean rating on the question of quality of
the interpersonal relations in the group was 12.56 out of a possible
maximum of 15. Average rating of perceived position power and decision
authority was 11.85. In both cases the standard deviation was less
than 3 points.
Verification of the Situational Stress Manipulation
The impact of the experimental manipulation of various aspects of
the leader's responsibility or functions was checked by an open response
question. Subjects were asked to describe in a few words the "fundamental
principle of military leadership." The responses were evaluated by two
officers not otherwise connected with the project. Their review
indicated that a majority of subjects in each experimental condition
responded in accordance with the emphasis manipulation they received.
Some subjects in all conditions responded with comments on both
dimensions of leadership function. This type of response, however,
was most characteristic of the "Both" emphasis condition subjects.
It is concluded that the manipulation was a success at least to the
degree that subjects acknowledged the differences among experimental
Prediction of Self-Presentation
As a result of the lack of major differentiations among subjects
on the Behavior Survey, adequate assessment of the fourth hypothesis
cannot reasonably be made. Partial testing was completed using those
subjects whose total score on the Behavior Survey was 80 or above
with balanced dimensions (n = 12) and those whose dimension scores
differed by five points or more (n = 13). Not all cells in the
experimental design are represented in these subgroups, therefore,
any evaluation of structured self-presentations would be inconclusive.
However, peer ratings of behavior on the jbb can serve to indicate
how these claimants of high activity and disproportionate behavior
are perceived by others.
Overall means for peer ratings of task activities and interpersonal
relations (IPR) were 10.75 and 10.04, respectively. For the subgroup
of high balanced claimants (n = 12) these means were 10.69 for task
and 10.78 for IPR, neither of which differs reliably from the overall
means. Similarly, average scores for those in low stress (n = 9) and
high stress situations (n = 3) did not differ reliably. For the sub-
group of unbalanced claimants, mean peer ratings were 11.01 and 9.94
for task and IPR, respectively. Once again these do not differ reliably
from overall means. Of this group, seven claimed to be more task active
and six to be more IPR active. Peer ratings indicate 10 of the 13 as
emphasizing task behavior over IPR activity although differences in
most cases are not extreme. Similarly, only two of the highly active
balanced claimants were rated as performing equally on both dimensions.
In general, there is little support for Hypothesis Four.
Contingency Model and Prediction of Self-Presentation
In order to permit analysis of the emphasis placed on different
trait dimension by the individual under certain conditions, the self-
descriptive data from the Information Exchange Worksheet was subjected
to two series of factor analysis procedures using Biomedical and
Statistical Packages for the Social Sciences programs which include
orthogonal rotation subprograms. Against expectations, two clear
factors related to the task and interpersonal relations dimensions
could not be readily identified. A three factor model was tested but
again attributes semantically related to task behaviors were mixed with
those related to interpersonal relations. Given the limited number of
subjects and the relatively large number of items, factor analysis
may have been an inoperable procedure in this study. Therefore,
Hypothesis Five and the predictions of differential use of task and
interpersonal relations terms in self-presentations cannot be assessed
in the manner intended. Had each scale item been evaluated and placed
into one of two categories prior to the use of the scale, total
dimension socres of these predetermined combinations might have been used.
However, this was not done and a test of the Hypothesis must await
The corollary to Hypothesis Five, however, can be assessed by use
of total self-presentation scores. In effect, the 33 characteristics
were treated as equally weighted elements in the leader's description
of his social self. The scores of all semantically negative terms
were reversed prior to compiling total scores. As the rating scale
used a value of one (1) for highly representative and all items were
computed as positive traits, the lower the score the more representative,
positive, and favorable the self-presentation was. The range of the total
scores for the 33 characteristics was 42 to 115 of a possible range of
33 to 165.
No reliable interaction effect within the 2 x 3 matrix (LPC by
emphasis) was found for total self-presentation scores. As expected, a
reliable main effect of LPC classification was found, F (1,46) = 4.861,
p .05 (Table 4). As a group, low LPC leaders described themselves more
favorably than did high LPC leaders. No reliable main effect of the
emphasis manipulation was found, F (2,42) = 1.236, N.S.
Certain cell comparisons were made based on predictions that assumed
an interaction between LPC and situational challenge or stress. As
predicted, low LPC leaders in the "Both" condition (assumed to be low
stress and compatible with the leader's motivational hierarchy) did
present themselves most favorably while high LPC leaders in the same
condition presented themselves much less favorably, t (46) = 1.84,
p .05. However, the expected effect of a low stress situation on the
high LPC leader (interpersonal relations emphasis condition) did not
produce a relatively favorable self-presentation when compared to the
low LPC leaders receiving the same emphasis manipulation. It appears
LPC Designation Emphasis Condition (Manipulation)
Task IPR Both Row Means
LOW LPC 65.79 62.31 60.15 62.75
HIGH LPC 71.21 73.81 70.74 71.92
Note.--Lower scores indicate more favorable (representative) self-
presentation (All negative terms reversed prior to computing
that personal rather than situational determinants were predominant
in the production of self-presentations in this study.
Evaluation of Leader Behavior
No frequency counts of actual behavior in the group setting were
made in this study. The only eivdence for in-role behavioral emphasis
available for analysis are the subjective ratings by peers made as the
part of the post-experimental evaluation. No interaction or main effects
were found for ratings of either task-oriented behaviors or for inter-
personal relations behaviors. Therefore, as perceived by others, leaders
across all experimental conditions behaved in a similar manner.
Given that leaders were generally perceived as behaving similarly,
Hypothesis Six gains in significance. Univeriate analysis of variance
revealed a highly reliable interaction of LPC and emphasis manipulation
for rating of predictability of the self-presentation for later in-role
behavior, F (2,42) = 4.889, p<.Ol. Neither main effect was significant.
In the "Both" condition, low LPC leaders were seen as more accurately
presenting themselves than were high LPC leaders, t (46) = 2.652, p<.01.
In the task condition, high and low LPC leaders did not differ reliably
in their degree of predictability. However, in the IPR condition, high
and low LPC leaders did tend to differ, t (46) = 1.63, p< 10 (two-tailed
test) with high LPC subjects seen as marginally more predictable. Given
these results and those previously reported for differences in total
self-presentation scores, the "Both" condition emerges as the situation
of major interest (Table 5).
Certain within row comparisons are also of interest. Low LPC
leaders in the task condition were seen as less predictable than low
COMPARISON OF LEADERSHIP PERFORMANCE RATINGS
LPC Designation Emphasis Condition (Manipulation)
Task IPR Both
LOW LPC 10.14 10.83 11.66
HIGH LPC 10.79 11.76 9.86
LOW LPC 6.88 8.88 11.00
HIGH LPC 8.88 10.63 12.25
LPC leaders in the "Both" condition, t (46) = 1.93, p<.05. These
situations were both assumed to be low stress conditions for the low
LPC leader and while their self-presentations did not differ reliably,
the perception of their behavior in the role of group leader led others
to see them as differentially predictable. This difference may be
attributed to the expectations fostered among the group members by the
emphasis manipulation, but, except for the evidence provided by the
manipulation check question, no direct assessment of expectations was
"Both" condition high LPC leaders were seen as significantly less
predictive of their in-role behavior than were IPR condition high LPC
leaders, t (46) = 2.739, p <.01. High LPC leaders in the task and "Both"
conditions did not differ in their degree of predictability.
Predictions made based on Hypothesis Six were generally confirmed
but with an unexpected reversal. Based on the predictability scores,
leaders in the low LPC/"Both" condition and the high LPC/IPR condition
(rated as highly predictive) should have received the highest average
leadership ratings while the high LPC/"Both" condition leaders (rated
as the least predictable) should have received the lowest average lead-
ership ratings. In fact, raw scores indicate that the high LPC/"Both"
condition leaders received the highest average leadership rating, M =
12.25 (15 point scale) followed by the low LPC/"Both" condition leaders,
M = 11.0, and the high LPC/IPR condition leaders, M = 10.63. These
scores do not reliably differ from each other but do differ from the
average ratings of some of those leaders in other cells seen as
moderately predictable. Specifically, the low LPC/Task condition
leaders were rated lower (M = 6.88) than the low LPC/"Both" condition
leaders, t (46) = 2.68, p 4.01, as suggested by their differences in
predictability of self-presentation. High LPC/Task condition leaders
while not reliably more predictable than high LPC, were rated as less
effective than high LPC/"Both" condition leaders, t (46) = 2.621, p <.05
Given the lack of differences reported for the behavior of leaders
as rated by peers and the radical differences in the degree of favorability
between the self-presentations of low LPC/"Both" condition leaders and
high LPC leaders in the IPR and "Both" conditions, the comparison of
predictability and effectiveness scores in these cells is particularly
informative. All these leaders were rated as highly effective compared to
peers in other conditions, however, their relative predictability varied.
High LPC/IPR condition leaders and low LPC/"Both" condition leaders were
seen as highly predictive but the high LPC/"Both" condition leaders were
seen as highly unpredictive of their in-role behavior. These results
support the conclusion that there may be at least two ways of establishing
initial credibility and fostering one's acceptability as a task group leader.
In summary, being either highly predictable in one's self-presentation or
highly unpredictable may foster higher evaluations of leadership effective-
ness. Since the high LPC leaders' self-presentations in the IPR and "Both"
conditions were equally modest, the differences in predictability must be
attributed to either actual behavior differences not reported by peers or
to different expectations generated by the emphasis condition manipulations.
Unfortunately, no data is available to resolve this question.
Some support was found for Hypothesis Seven. A comparison of
self-esteem scores and overall self-presentation scores revealed a
weak negative relationship, r (46) = -.223, p <.10. Since low self-
presentation scores were more favorable, the prediction of a weak but
positive relationship between self-esteem and self-presentation is
Predictor Scale Relationships
The discovery of small but reliable correlations between the measures
of self-esteem and minority identification and LPC scores is important in
that it provides additional information on the construct measured by the
LPC scale and on its interpretation. High self-esteem and identification
with the minority seem to characterize low LPC persons. The very fact
that an individual scores low on the LPC scale is an indication that he
or she is willing to critically evaluate the work performance of others.
This may require a secure self-concept or at least a high level of self-
esteem and a sense of identification with a select group of high performers.
Some evidence in support of this interpretation comes from a study
by Fishbein, Landy and Hatch (1969). The authors asked individuals pre-
viously classified as high or low LPC's to describe, in their own words,
people with whom they had difficulty working on a common task. Frequency
counts of the adjectives used were obtained. The high LPC person tended
to see this least preferred co-worker as a competitor for prominence and
recognition (leadership) within the group, citing them as intelligent,
strong willed, self-motivated, assertive, bossy, and a "know-it-all."
The low LPC persons in this survey tended to see their least preferred
co-worker as unintelligent, incompetent, sloppy, unreliable, careless,
and unpleasant. To the low LPC person, such an individual represents a
threat to task efficiency and accomplishment, rather than a direct
threat to self. These results suggest that there are differences in
the selection, perception and evaluation of the object being considered
when an individual takes the LPC scale.
The high LPC person is, in effect, rating a competitor, a challenger
or rival forthe leader's prestige and esteem. This person then con-
stitutes a threat to the high LPC person's secondary motives. However,
viewing one's least preferred co-worker as a competitor may require the
leader to rate him/her favorably on many of the dimensions of the LPC
scale in order to maintain one's own self-concept. If, in addition, one
is lower in self-esteem such threats may be all the more salient to the
high LPC leader.
The low LPC leader is presumably rating a disliked individual seen as
an "opposite," an inferior, and a threat to the task at hand and therefore
to the group as a whole. Thus, the low LPC leader can, with minimal self-
concept constraint, rate the least preferred co-worker low on every dimension.
The low LPC leader's self-esteem is challenged by task failure, and dero-
gation of a source of possible task failure is possibly a defense mechanism
that helps maintain high self-esteem.
It is difficult to judge the merits of the Behavior Survey Scale for
predicting how a leader will attempt to establish his credibility and
suitability for leadership from this study. As indicated by the generally
high and balanced ratings, those who aspire to military leadership perceive
the competing and yet mutually supporting nature of the task and main-
tenance functions. This scale is probably best suited as a training device
rather than as a predictor of later behavior.
The Prediction of Self-Presentation
Two scales do appear to have some value as predictors of an individual's
self-presentation in a group setting. Self-esteem was found to be marginally
related to the favorability of one's self-presentation. Similarly, the main
effect of LPC score on the favorability of self-presentation supports the
use of this scale in studies of impression management. Given the experi-
mental design, the results have been presented primarily with respect to
LPC score. However, the negative correlation between LPC and self-esteem
scores suggests that the differences in favorability of self-presentation
might be attributed to the effects of self-esteem as well. Studies by
Adessa (1974), Schneider and Turkat (1975) and others suggest a positive
relationship between self-esteem and the favorability of one's self-
presentation, at least under certain conditions. For example, Schneider
and Turkat (1975) report that defensive high. self-esteem subjects presented
themselves more favorably after failure than after success. Subjects seen
as "genuine" high self-esteem individuals responded in a similar manner
but to a lesser degree. These results, however, can help explain the
relationships found in the present study only if one is willing to assume
that high LPC subjects, as a group, expected to succeed and low LPC subjects
expected to fail in the task-group situation. This writer is not willing
to make this assumption, since high rather than low self-esteem subjects
should logically expect to succeed. However, it is reasonable to infer
that the two measures a e tapping some common dimension of human variability
and, therefore, may well have a joint application in the prediction of
Self-Presentational Predictability and Effectiveness Ratings
The most interesting finding of this study is the apparent curvilinear
relationship between the predictability of one's self-presentation for later
behavior and ratings of leadership effectiveness. For certain leaders, a
highly favorable self-presentation reflecting confidence, competence or
high expectations for oneself, when followed by behavior that matches the
claims one has made, will result in high marks in playing the role. In
contrast, a modest self-presentation that is far exceeded by actual behavior
may have set the stage for attributions of extra effort on the part of the
role occupant. Credibility does not suffer if one exceeds the expectations
aroused in those who must evaluate you. Low and high LPC leaders in the
"Both" condition appear to have used these two very different approaches
to the assumption of leadership.
Given certain conditions (i.e., emphasis on interpersonal relations)
some leaders may present themselves as modest or average performers, then
behave in a manner that is consistent with their claim or line. This
veridical self-presentation may be especially effective for the leader who
is primarily motivated by good interpersonal relations (the high LPC
leader). Modest claims and moderate behavior in the group may well be
effective in securing favorable interpersonal relations and subsequent
attributions of effectiveness, as it was for the high LPC leaders in the
IPR condition of this study. For the task-motivated leader (low LPC),
claims of high competence may be the first step in insuring control over
the group's task activities. In the ambiguous environment surrounding a
newly appointed leader, self-presentations may be directed towards the
securing of one's primary goals. Further study is needed to confirm these
Evaluation of the Research Design
The situational self-presentation device, the Information Exchange
Worksheet, was inadequate for its purpose. An insufficient number of
task-specific characteristics were included. Words that should logically
have been incorporated include: decisive, enthusiastic, objective or
rational, directive, task oriented (also people oriented for contrast),
high achiever, and skillful. Even with such items, a clear differentiation
between task and interpersonal relations dimensions may be difficult. In
all probability, an abbreviated form of the Behavior Survey Scale with its
clear descriptions or an open-ended written self-description would have
been more effective both in differentiating among subjects and permitting
later analysis. In addition, some form of behavioral frequency counts
should probably have been taken rather than relying solely on subjective
peer ratings of behavioral emphasis on-the-job. Bales (1950) Interaction
Process Analysis Technique would be preferable.
The actual group environment may have contained forces and pressures
that significantly moderated the effect of the emphasis manipulation on
in-group behavior. For these cadets, the lack of experience with military
tactics and doctrine seem to have fostered a cooperative and friendly
atmosphere in these peer groups such as is found only in the most cohesive
of naturally occurring military groups. No member by formal training or
rank had a legitimate claim to expertise nor were status differences
apparent. No member posed a threat to another based on rank or experience.
As a result, all members were potential sources of mutual support and
opinion that could be accepted openly. Reciprocation may well have been
the most powerful but implicit norm of these groups.
The rotation of leadership may have created different challenges
to the establishment of credibility for the first leader as compared
to those that followed. The first leader was forced to operate in an
environment devoid of norms. His approach may well have become the
criterion against which subsequent leaders were evaluated. By the end
of the first five minute period, the group had established some opera-
tional norms that those to follow in the leadership role may have felt
compelled to recognize. Having observed the first leader and the group's
reaction to him subsequent leaders could make modifications in their
In this regard, an analysis of those "first to lead" was made.
Purely by chance (since cadets picked their own sequence numbers), six
high and six low LPC leaders led first, two in each cell of the experimental
design. Peer leadership ratings, followership ratings, and ratings of the
predictability of self-presentation were compared. Mean predictability
ratings did not differ (M = 10.28 and M = 10.78 for low and high,
respectively). Followership ratings tended to differ (M = 10.15 and M =
11.47), t (22) = 1.33, p .10 with high LPC cadets rated more favorably.
Leadership ratings, however, differed radically. Compared to a grand
mean for all subjects of 9.75, low LPC "first leaders" received an average
rating of 9.0; high LPC first leaders were rated a surprising 11.63!
Needless to say, these means differ reliably, t (22) = 2.18, p<.05
(two-tailed test). It appears that being first to lead was detrimental
to low LPC leaders but beneficial to high LPC leaders! It may well be
that the high LPC leader's interpersonal relations motivation and modest
self-presentation were more suited to the early group environment than
the low LPC leader's task orientation and firm claims to proficiency.
The short time period allowed for each problem coupled with other
situational constraints may have severely limited the opportunities for
interpersonal relations activities not connected directly with accomp-
lishing the task (i.e., such as prevently arguments or other disruptive
actions). A better evaluation of the impact of the self-presentational
form read to the group by each leader may have been gained if a period
of open discussion had been provided (for example, a break between the
second and third sessions) so that subjects could interact in an
environment not constrained by the task demands.
Finally, it must be concluded that given the very real demands of
the decision making situation in which each leader found himself, the
situational stress manipulation may have been poorly designed for the
task it was intended to perform. The major results of this study can
be derived from the "Both" condition for it is the natural and ever-
present one. Rarely can the leader escape the competing demands of the
mission and those that must accomplish it. The results in this "natural"
condition indicate that equally successful leaders can approach the
problem of establishing credibility in different ways.
In summary, the results provide some support for the predicted
inverse relationship between LPC scores and self-esteem scores as
measured in this study but little support for the other hypothesized
relationships between the predictor scales. Hypotehsis Four could not
be adequately tested but the generally favorable responses on the
Leadership Behavior Survey may indicate that this group of aspiring
military leaders recognizes the role ideal represented by Johnson and
Johnson's 10,10 leader. Only the corollary to Hypothesis Five could be
tested. Low LPC leaders presented themselves more favorably overall.
In the "Both" condition, differing self-presentations were seen as highly
predictive or high non-predictive of in-role behavior by low and high
LPC leaders, respectively. However, these leaders were rated equally
effective by their peers. The relationship between self-esteem and
self-presentation found in this study may confound the LPC/self-
presentation relationship, but this potential three-way interaction
requires more study.
This research project was conducted to analyze the self-presentation/
impression management activities of individuals in the initial stages of
role-making. Three questions were asked at the outset: (1) Is there a
way to predict a leader's self-presentational emphasis? The answer to
this must be a qualified yes. The LPC scale does appear to be a potential
predictor although additional research is needed. (2) Do situational as
well as personal variables effect self-presentation? Again, the answer
must be a qualified yes. Self-esteem and one's motivational hierarchy
do appear to be significant personal variables, while situational stress
also appears to have an effect on self-presentation as part of a causal
interaction. (3) Do the leader's early efforts at defining his/her role
effect later evaluations of performance on-the-job? Once again, a tentative
yes is appropriate. Self-presentations that were either solidly confirmed
by later behavior and those radically at odds with in-role behavior, appear
to be related to evaluations of effectiveness. However, the degree of
causality in this relationship cannot be assessed from this research.
Further research into the role making process, and the influence
of skillful impression management on the assumption of leadership, will
be of value not only to the student of the leadership process, but also
to the practitioner of leadership in formal organizations.
LEADERSHIP BEHAVIOR SURVEY SCALE
When you are a member of a group, what is your leadership behavior
like? In what ways do you try to influence other group members toward
accomplishing the group's goals? The purpose of the survey below is to
get a description of your behavior in groups. Circle the letter to the
left that most appropriately describes your likely behavior--(A) always,
(F) frequently, (0) occasionally, (S) seldom, or (N) never--in connection
with the given statement. Each of the items below describes aspects of
leadership behavior; respond to each one according to the way in which
you would be most likely to act if you were part of a problem-solving
When I am a member of a problem-solving group....
A:F:O:S:N 1. I offer facts, give my opinions and ideas, provide
suggestions and relevant information to help the
A:F:O:S:N 2. I warmly encourage all members of the group to parti-
cipate, giving them recognition for their contributions,
demonstrating receptivity and openness to their ideas,
and generally being friendly and responsive to them.
A:F:O:S:N 3. I ask for facts, information, opinions, ideas, and
feelings from other group members to help the group
A:F:O:S:N 4. I try to persuade members to analyze constructively their
differences in opinions and ideas, searching for common
elements in conflicting or opposing ideas or proposals,
trying to reconcile disagreements.
A:F:O:S:N 5. I propose goals and tasks in order to start action
within the group.
A:F:O:S:N 6. I try to relieve group tension and increase the enjoyment
of group members by joking, suggesting breaks, and pro-
posing fun approaches to group work.
A:F:O:S:N 7. I give direction to the group by developing plans on how
to proceed with group work and by focusing members'
attention on the tasks to be done.
A:F:O:S:N 8. I help communication among group members by showing good
communication skills and by making sure that what each
member says is understood by all.
A:F:O:S:N 9. I pull together related ideas or suggestions made by
group members and restate and summarize the major
points discussed by the group.