You are on page 1of 12

Personality and Situational

Effects on Leader Behavior^


STEPHEN G. GREEN
University of Washington
DELBERT M. NEBEKER
United States Navy
M. ALAN BONI
University of Washington

Tiiis paper examines Fiedler's proposed relationship


between leader beitavior and two variables related to tire
Contingency Model. Tlie evidence strongly suggests tiiat
under stress in a threatening situation, leaders witii iiigii
Least Prefered Coworker (LPC) scores are more interpersonally oriented witiie low LPC leaders are more task
oriented.
Although Fiedler's contingency model of leadership effectiveness (1967)
has proven to be particularly successful in delineating the conditions under
which leaders are relatively effective, it has experienced considerable difficulty in providing a behavioral explanation for these differences. This
theory posits a contingent relationship between leadership performance and
a leadership style measure referred to as the esteem for the Least Preferred
Coworker (LPC). This measure results from leaders rating their least preferred coworker on a set of bipolar adjectives scales such as friendly
unfriendly, and efficientinefficient. Depending on whether the leader
describes his LPC in relatively positive terms or in a negative fashion, he
is classified as either a high LPC leader or a low LPC leader, respectively.
The empirical model postulates that low LPC leaders perform best in either
very favorable or very unfavorable situations; the high LPC leader performs
better in situations which are intermediate in favorability. For Fiedler's
purpose, situational favorability is classified by using three dichotomous
dimensions:
Stephen G. Green is a doctoral student at the University of Washington, Seattle, Washington.
Delbert M. Nebeker is a psychologist with the Navy Personnel Research and Development
Center, San Diego, California.
M. Alan Boni is a doctoral student at the University of Washington, Seattle, Washington.
1 This paper is based on research performed under ARPA Order 454, Contract N0001467-A-0103-0013 with the Advanced Research Projects Agency, United States Navy (Fred
E. Fiedler, principal investigator) and Contract NR 177-472, N00014, 67-A-0103-0012 with
the Office of Naval Research, Department of the Navy (Fred E. Fiedler, principal investigator).
184

1976

Volume 19, Number 2

185

(a) The degree to which there are positive feelings in the group (group
atmosphere, GA);
6 F ve F
(b) The amount of structure in the task (task structure, TS); and,
(c) The leader's power as a function of his position (position power, PP).
These three elements combine to form eight octants, with octant one
(positive GA; high TS; high PP) being most favorable and octant eight
(poor GA; low TS; low PP) least favorable for the leader.
The interpretation of what leader behavior results in these differences in
effectiveness is not so clear. Originally, Fiedler (1961) viewed LPC as a
leadership trait or style measure, with high LPC leaders being interpersonally oriented and low LPC leaders task oriented. However, when
support for this interpretation could not be found empirically, Fiedler began
to look for an alternative explanation (Bass, Fiedler & Krueger 1964Bishop, 1964).
In an attempt to account for inconsistent data, Fiedler came to a motivational interpretation of LPC (1973). He argued that LPC is a motivational
index which corresponds to a leader's hierarchy of goals. This hierarchy is
arranged with the leader having both primary and secondary goals, the
satisfaction of the primary goals allowing the leader to pursue his secondary
goals. For the high LPC leader, interpersonal relations are primary, and
prommence and self-enhancement (gained through task-relevant behavior)
are secondary. On the other hand, for the low LPC leader, task accomplishment is a primary goal and good interpersonal relations a secondary goal.
A reanalysis by Fiedler (1972) of a number of studies yielded some support
for this interpretation of the LPC leader's behavior (e.g., Fiedler, Meuwese
& Oonk, 1961; Hawley, 1969). In favorable situations, the high LPC
leaders attended to the task and low LPC leaders behaved in a relationship
onented manner. Conversely, when the situation was unfavorable, the high
LPC leaders became interpersonally oriented and the low LPC leaders were
task oriented.
Fiedler suggests that the factor which moderates the behavior of the
leader is the threat inherent in the situation and the concomitant stress to
the leader (1972). Intuitively, this notion is appealing and somewhat
compelling. It would seem reasonable that the situations which threaten the
leader and stress him would cause him initially to seek the primary goals in
his motivational scheme. Implicit in this conception is the assumption that
under low threat, nonstressful conditions the leader is more likely to pursue
his secondary goals. Thus the empirical support presented by Fiedler (1972)
is interpretable if we can assume the favorable situations were less threatening than the unfavorable. This conclusion seems warranted by Fiedler's
analysis; however, it is severely obscured by the diverse independent
variables and dependent measures addressed in his review.
The present study was designed to test Fiedler's hypotheses concerning
the motivational hierarchy and specifically explore the notion that the
threat involved in situations of differing favorability is the significant under-

186

Academy of Management Journal

June

lying dimension interacting with the leader's motivational goals. In addition,


since the paucity of empirical evidence on the behavioral correlates of LPC
may be the result of unstandardized observations, this study provided the
opportunity to observe the behavior of high and low LPC subjects under
standardized and controlled conditions.
The hypotheses tested in this study are founded upon Fiedler's interpretation that there is an interaction between the personality of the leader
(measured by LPC) and the situation in which he acts. In general, it is
believed that in a situation which is threatening to the leader (and thus
a high stress condition), the high LPC leader will be interpersonally
oriented and the low LPC leader task oriented. On the other hand, in situations of low threat (therefore, nonstressful), the high LPC leader will be
task oriented and the low LPC leader interpersonally oriented. In effect, the
two types of leaders will pursue their primary goals in stressful situations
and their secondary goals in nonstressful situations. The major hypotheses
are:
Hypothesis 1In conditions of low stress, iow LPC leaders will be more
interpersonally oriented and less task oriented than high
LPC leaders.
Hypothesis 2In conditions of high stress, high LPC leaders will be
more interpersonally oriented and less task oriented than
iow LPC leaders.
Hypothesis 3Across all situations, high LPC leaders will be more
interpersonally oriented and less task oriented than iow
LPC leaders.
METHOD
Subjects

Sixty male subjects from the University of Washington undergraduate


psychology subject pool were administered a battery of personality tests
including the LPC scale. Their mean LPC score (16-item version) was
62.1, which did not differ significantly from normative findings (Posthuma,
1970). Eighteen subjects scored in the upper third of the distribution and
were chosen as high LPC subjects. Fifteen subjects scored in the lower
third of the distribution and were chosen as low LPC subjects (a total of
33 subjects). This selected sample of subjects then participated in a laboratory experiment. Participation was encouraged by extra class credit.
Stress Manipulation

Upon arrival at the laboratory, each group of subjects was given one of
two sets of verbal instructions. One set of instructions, designed to threaten
the leader (high stress), told the group that they were participating in
leadership research and that the test they were about to take was a good
indicant of leadership ability. Also, the subjects were informed that "positive

1976

Voiume 19, Number 2

187

correlations between scores on this test and such constructs as I.Q. or need
achievement" had been shown in current studies. Finally, they were informed that a high score on their test could result in their returning for
another study for which they would be paid. The other instructions, devised
to minimize threat to the leader (low stress), informed the group that they
were participating in leadership research and that the test they were about
to take was merely in its developmental stages. The subjects were asked to
relax, as this was only an exploratory test and there was no correct or
incorrect response. Thus, an attempt was made to induce two conditions of
stress within a leadership situation by manipulating the threat of the test
atmosphere.
The Task

Each subject listened to a tape-recorded committee meeting as if he were


in the meeting as its chairman. The meeting was divided into scenes, each
scene dealing with a somewhat different problem relevant to the context of
the meeting. As a scene was terminated, a cue for the subject-chairman to
respond to the discussion was given. Each subject was given a short time to
respond as if he were actually speaking to the committee. At the end of this
allotted time period, the group discussion would fade back into the next
scene. The subject's responses were taped as the major dependent variable
providing seven samples over a variety of different issues and problems
within a single controlled situation. (To strengthen the stress manipulation,
the high stress group was given only 20 seconds to reply while the low stress
Ss were given 40 seconds to respond. Nevertheless, Ss almost always had
sufficient time to reply.)
The script was developed to present the subjects with a variety of situations in which they could support or reject individual or group ideas, deal
with an argument, deal with a lack of information, and act or not act in a
\yarm and friendly manner. On the basis of Fiedler's categorization of situations, it was anticipated that the committee meeting itself would be an
unfavorable situation. The group atmosphere would be moderate to low,
the task unstructured and the leader's power position low. This methodology
seemed genuinely to engage the subjects, as concern for the group's problems often was evident in their taped responses; thus, it was not uncommon
for the subjects to use the names of committee members in their responses
to the group.
Measures

In addition to the taped responses, other measures were taken. After the
verbal instructions and before performing the task, subjects were given a
short questionnaire and asked to report their present feelings. This instrument was designed to assess the perceived stress as an aid in evaluating
the validity of the manipulation. After the task, each subject was given
a two-part questionnaire. In part one, subjects described their own behavior

188

Academy of Management Journal

June

as committee chairman on an adapted Leader Behavior Description Questionnaire (Halpin & Winer, 1957); in part two, they answered several
questions to ascertain suspicion and the stressfulness of the situation.
A rating system was employed to classify the dependent measurethe
Ss' taped responsesas either interpersonal or task oriented. The classification and ratings were based on five interpersonal and six task orientation
criteria which had been derived from the major LBDO items defining
consideration and initiation of structure. The following items served as
the operational definition for the dependent variables:
Interpersonal orientation (Io)
1. He looks out personally for individual group members.
2. He consults the group on any actions.
3. He backs members in their actions or ideas.
4. He is friendly and approachable.
5. He tries to put group members at ease when talking with them.
Tasii orientation (To)
1. He expresses his attitudes, opinions and/or ideas to the group.
2. He rules with an iron hand.
3. He emphasizes the use of uniform procedures.
4. He criticizes poor work or ideas.
5. He makes certain his part (as the chairman) is clearly understood by
the members.
6. He lets group members know what is expected of them.
A five-point scale from very true to not true was used for each of the 11
LBDO items, with four independent judges' ratings being averaged for
each subject. The totals for the averaged ratings of the interpersonal and
task items were summed across the scenes. This process provided a single
measure of each dimension for each subject. The internal reliability (Nunnally, 1967) of the interpersonal items as rated by the judges was .88 and
.74 for the task items. The interjudge reliability obtained an average intercorrelation of .79.
RESULTS
Validation of Manipulation

In order to evaluate the stress manipulation, a chi-square test was


generated by classifying the subjects into high and low manipulated stress
groups and into stressed and not stressed groups based on their response
to the posttest questionnaire. A chi-square of 2.19 (n.s.) was obtained,
demonstrating the manipulation was not adequate. In fact, the majority of
the subjects reported themselves as experiencing stress, with the average
self report falling in the upper half of an anxiety scale. Apparently, the
total situation was so stressful that a ceiling had been reached. These selfreported indications of stress taken in conjunction with the extremely

Voiume 19, Number 2

189

ambiguous nature of the committee task, the virtually powerless position of


the chairman and the committee's low group atmosphere (X = 49 00)
created an extremely unfavorable situation for the leader. Therefore the
groups were combined and treated as essentially facing the same unfavorable situation as leaders, making it impossible to test hypothesis 1; thus the
total motivation hierarchy cannot be examined with this study. Nevertheless
due to the small amount of empirical evidence available on the behaviorai
correlates of LPC in standardized situations, it was felt that combining the
data would answer valuable questions. An understanding of what leader
behaviors are associated with LPC types in this unfavorable situation would
help test part of Fiedler's interpretation as well as generally demonstrate
the behaviorai implications of LPC.
Analysis of the Leader's Overall Behavior
Using the obtained dependent measures, Ntests were performed between
the high and low LPC leader groups, in that the Ss were now considered as
all facing an unfavorable situation. Significant differences were found, with
the high LPC leaders being more interpersonally oriented (t 2.3S; p <
.02) and the low LPC leaders more task oriented (t = 2.20; p < .05).
These results provide support for hypothesis 2.
TABLE 1
Mean Interpersonal and Task
Orientation Scores
High and Low LPC leaders"
Leaders
High LPC

Low LPC

Interpersonal
Orientation
X = 118.75
X

SD
CT._= 11.28
...o
n ;= (15)

^3J^ == m
90
109.90
SD =
7.75
n = (13)

Task
Orientation
X == 148.45
SD
.^ =
,7.61
, .
n = (15)
^
= ^V5:O3
jc
155.03
SD =
8.16
n = (13)

" Due to equipment failure, five subjects'


responses to several scenes were lost. Thus,
they were excluded from this overall analysis.

TABLE 2
Correlation Between Self Ratings and
Judges RaHngs
of Interpersonal and Task Orientations
Tw^Z
Interpersonal

}7T~
Task

s^,^^^^^.

( - 28^
("-28)

. _ ,ox
(n_28)

SeKk''

-J^t*

- ^ .

Ratings

<IO
< .001

In an attempt to interpret the main effect of LPC, a correlation matrix


for LPC and the two dependent measures (Io; To) was created. This matrix
revealed that LPC correlated positively with interpersonal ratings (r = .42;
p < .02) and negatively with task ratings (r = - . 4 0 ; p < .02). Also, the
dependent measures displayed a negative relationship between themselves
(' ~ ~ ' 7 2 ; p < .001). To clarify these relationships, the task orientation
rating scale was reversed (i.e., 5 became 1, 4 became 2, etc.) and combined
with the interpersonal orientation scale. This transformation provided a
continuum which ran from a high tasklow interpersonai orientation (low
scores) to a iow taskhigh interpersonal orientation (high scores). Cor-

190

Academy of Management Journal

June

relating LPC with the combined Io/To continuum yielded a significant


positive relationship (r = .44; p < .01; n = 28).
Comparison of Self and Judges Ratings
In the posttask questionnaire, the subjects rated their behavior on an
adapted LBDQ that contained the same items on which the judges rated
their responses. Interpersonal (Io) and task (To) scores were computed for
the self ratings in the same manner as described for the judges' ratings.
Correlations of the self ratings with judges' ratings on these two dimensions
provided strong support for the notion that leaders perceived themselves
as engaging in the same behaviors the judges perceived (Table 2 ) .
The Io and To scales for self rating were also combined into a single
continuum as was previously done for the judges. When the self-rated Io/To
continuum was correlated with the judges' Io/To continuum it yielded
strong, significant agreement (r = .72; p < .001, n = 28) between judges
and subjects.
Due to a weaker agreement between subjects and judges on the interpersonal dimension, the correlations between LPC and self ratings were
smaller but still supportive: LPC correlated with self ratings of interpersonal
orientation .01 (n.s.) and with self ratings of task orientation .36
(p < .02). The correlation between LPC and the self-rated Io/To continuum was .30 (p < .05; n = 33).
Discriminators of Leaders
A correlational analysis was performed on the separate LBDQ items
and the seven scenes, correlating LPC with each item's ratings and with
the ratings for each scene. This analysis revealed some differences among
judges' and self ratings as to which items discriminated between the high
and low LPC leaders. However, certain LBDQ items were obviously quite
potent in differentiating the two types of leaders. The high LPC leader
both saw himself and was seen as "looking out" for the group members; the
low LPC leader both saw himself and was seen as expressing his own
attitudes, emphasizing uniform procedures, and making certain his position
as chairman was understood (Table 3).
When a similar analysis was performed to see if some scenes better
discriminate between leaders, two scenes conspicuously provided situations
where the two types of leaders took divergent courses of action. In one
scene (scene four), the committee decides to request some information
from a faculty meeting. Qne member is challenged to accept this assignment
since he has not been contributing, and he resists. In another scene (scene
seven), the agreed upon quitting time is reached and there is some confusion
as to whether the group should quit or continue until the issue at hand is
decided. In the discussion, the chairman is accused of not directing the
committee's actions. Both of these scenes elicited striking differences in the
subjects' behaviors; the high LPC leaders were significantly more inter-

Voiume 19, Number 2

191

TABLE 3
Correlations Between LPC and LBDQ Items
for Judges and Self Rating
Judges
(n = 28)
Consideration
Consideration
11. He looks out personally for individual group members.
2. He consults the group on any actions.
3. He backs members in their actions or ideas.
4. He is friendly and approachable.
5. He tries to put group members at ease when talking with them.
Initiation of Struclure
1. He expresses his attitudes, opinions and/or ideas to the erouD
2. He rules with an iron hand.
3. He emphasizes the use of uniform procedures
4. He criticizes poor work or ideas.
5. He makes certain his part (as the chairman) is clearly
understood by the members.
6. He lets group members know what is expected of them.
*p<.10

42**
'28*
25
'43**
.38**

Self
(n = 33)
39**
19
04
_'J4
!l 1

30*
_'35*
_'47*
_'O9

40*
_07
_'27*
QO'

_ 32*
01

_ 3g
15

** p < .05
***

personally oriented and the low LPC leaders were significantly more task
oriented (Table 4). A comparison of these two scenes with the others
presented indicates that leaders are most likely to display behavior consistent with their LPC when the situation does not suggest either consideration or initiation of structure but provides greater ambiguity about what is
appropriate behavior.
TABLE 4
Correlations Between LPC and Judges Ratings
of Interpersonal Orientation and Task Orientation for Scenes
Scene 4

Scene 7
= 33)

* p < .05
** p < .01

DISCUSSION
.
conditions of threat, leaders are apt to respond
in different manners depending upon their personality. There is also strong
evidence that LPC as a personality measure is capable of tapping these
differences. In an unfavorable situation, the high LPC leader reliably
engaged m considerate behavior in his attempts to cope. His primary
concern was the maintenance of interpersonal relationships within the
group; consistently, he manifested interpersonal orientations rather than task
relevant orientations. Conversely, the low LPC leader tended to initiate
structure when confronting group related problems in a threatening situa-

192

Academy of Management Journal

June

tion. To some extent, this structuring appears to be composed of both


directive behaviors and attempts to clarify the leader's role. In both cases,
however, the object of the low LPC leader's emphasis was the task; direction
was provided toward task completion and clarity was sought in terms of
the role of task leader. The results strongly support the notion that the
low LPC leader perceived task orientation as appropriate in dealing with
group problems. These results are notable when one considers the previous
difficulty in finding behavioral differences between high and low LPC
leaders.
This finding is enhanced by the fact that the leaders not only were seen
by observers in the described manner but also to some extent perceived
their own behavior in the same light. The correlations obtained between
judges' ratings and self ratings greatly strengthen the validity of the conclusion that high LPC leaders are motivated to be interpersonally oriented
and low LPC leaders are motivated to be task oriented in an unfavorable
situation. This motivational interpretation has been supported by similar
results obtained employing various tasks and sources of threat (e.g., Larson
& Rowland, 1973; Mitchell, 1969), but this study represents the first time
the stress of the situation has been examined in conjunction with situationai
favorability as defined by Fiedler.
A post hoc analysis of the situations (i.e., scenes) which best discriminated between high and low LPC leaders yielded several possible hypotheses.
First, it appears that some situations and their concomitant problems are
so structured that situationai demands, to a great extent, shape the leader's
behavior. If a situation and problem are so presented that the course of
action obviously calls for a task orientation, the leader will attend to task
concerns regardless of personality. For example, at one point the committee
agrees that it lacks information on a topic and one member suggests the
chairman obtain it. This particular set of circumstances does not allow for
much variance of behavior; the leader essentially has only a task problem
and is thus task oriented. Second, situations which show personality influence on behavior seemingly provide the leader the opportunity to attend
to either task or interpersonal concerns. Where the choice is genuine, and
one dimension does not so dominate the other, personality is more likely
to serve as a true moderator for behavior.
The concept of a leader having a consistent behavioral style in dealing
with subordinates appears questionable. Obviously, one must consider the
situation in which the leader is acting as well as his motivations to predict
leader behavior. This view offers a plausible explanation for the negligible
effects of training upon leaders' performance. As Campbell et al. (1967)
point out in their comprehensive review of leadership training, individual
differences and the influence of situationai variables are largely ignored by
traditional training programs. In light of the strong interaction found
between personality (LPC) and situationai variables, it is evident that

1976

Volume 19, Number 2

193

ignoring these considerations can only attenuate, if not eliminate, the effects
of training.
The highly significant negative relationship between the consideration
and initiation of structure items calls into question the orthogonality of
these two behaviors. Obviously, because of the limited interaction afforded
the leaders and the reduced number of LBDQ items used, these results
cannot be as easily generalized as one would like. Nevertheless, even
though these behaviors intuitively and conceptually appear orthogonal,
they may be perceived in some situations as being mutually exclusive. That
is, most leaders may not feel it appropriate, or even possible, in certain
situations to use these behavior patterns independently. In this instance,
it may be that LPC as a unidimensional construct is capable of tapping
this behavior.
Several implications for future research derive from the present study.
In retrospect, the type of stress manipulation attempted here may have
been inappropriate. It now seems that the stress, to be meaningful, should
come from the task or from the situation itself, not from an exterior source.
Also, future research will have to consider both favorable and unfavorable
situations in order to study leader behavior in an adquate manner.
Finally, the methodology of this study can serve as the basis for fruitful
research in the area of personality and situational interactions. As so aptly
pointed out by Vroom and Yetton (1973), one of the continuing problems
confronting leadership research is the need to partition individual differences
and situational variables and to develop standardized situations in which
to study leader behavior. The present methodology allows the experimenter
much needed control over the situation, permitting one to vary situational
variables independently and provide a standardized situation to leaders.
REFERENCES
1. Bass, A. R., F. E. Fiedler, and S. Krueger. Personaiity Coneiales of Assumed Simiiariiy
and Related Scores (Urbana, 111.: Group Effectiveness Research Laboratory University
of Illinois, 1964).
2. Bishop, D. W. Relations Between Tasks and Interpersonal Success and Group Member
Adjustment (M. A. thesis, University of Illinois, Urbana, 1964).
3. Campbell, J. P., M. D. Dunnette, E. E. Lawler, and K. E. Weick, Jr. Managerial Behavior, Performance and Effectiveness (New York: McGraw-Hill, 1967).
4. Fiedler, F. E. "Leadership and Leadership Effectiveness Traits: A Reconceptualization
of the Leadership Trait Problem," in Petrullo and Bass (Eds.), Leadership and Inter,
personal Behavior (New York: Holt Rinehart, 1961).
5. Fiedler, F. E. A Theory of Leadership Effectiveness (New York: McGraw-Hill, 1967).
6. Fiedler, F. E. "Personality, Motivational Systems, and Behavior of High and Low LPC
Persons," Human Relations, Vol. 25 (1972), 391-412.
7. Fiedler, F. E. "Personality and Situational Determinants of Leader Behavior," in E. A
Fleishman and J. G. Hunt (Eds.), Current Deveiopments in the Study of Leadership
(Carbondale, III.: Southern Illinois University Press, 1973).
8. Fiedler, F. E., W. Meuwese, and S. Oonk. "Performance of Laboratory Tasks Requiring
Group Creativity,"/Jc/aP.yj'c/io/ogica, Vol. 18(1961), 100-119.
9. Halpin, A. W., and B. J. Winer. "A Factorial Study of the Leader Behavior Descriptions, m R. M. Stogdill and A. E. Coons (Eds.), Leader Behavior: Its Description and
Measurement (Columbus: The Ohio State University, Bureau of Business Research
Monograph No. 88, 1957).

194

Academy of Management Journal

June

10. Hawley. D. E. A Study of tlie Relationship Between the Leader Behavior and Attitudes
of Eiementary School Principals (M. Ed. thesis, University of Saskatchewan, Saskatoon,
1969).
11. Larson, L. L., and K. M. Rowland. "Leadership Style, Stress and Behavior in Task
Performance," Organizationai Behavior and Human Performance, Vol. 9 (1973),
407-420.
12. Mitchell, T. R. Leader Complexity, Leadersiup Style and Group Performance (Doctoral
dissertation University of Illinois, 1969).
13. Nunnally, J. C. Psychometric Theory (New York: McGraw Hill, 1967).
14. Posthuma, A. Normative Data on the Least-preferred Co-worker and Group Atmosphere Questionnaires, Technical Report 70-8 (Seattle, Wash.: Organizational Research,
1970).
15. Vroom, V. H., and P. W. Yetton. Leadersiiip Behavior on Standardized Cases, Technical Report No. 3 (New Haven: Administrative Sciences, Yale University, 1973).

You might also like