Professional Documents
Culture Documents
August 2004
UMI Number: 3140603
Copyright 2004 by
Stanley, Angelia Denise
All rights reserved.
________________________________________________________
Abstract
This dissertation has explored the relationship between leadership styles and
different conflict management styles exist and can be assessed using the
Thomas-Kilmann Conflict Mode Instrument (TKI). The literature implies, but does
not explicitly state, that a relationship may exist between leadership styles and
conflict management styles. Data were collected from 99 leaders who completed
the MLQ and the TKI. Each leader received a score on Transactional Leadership,
Leadership. This study used the scores on the leadership styles as measures of
each leader exhibits. This study used TKI scores as measures: of the extent to
between the set of leadership styles (as assessed by the MLQ) and the set of
conflict management styles (as assessed by the TKI). In all tests the data fail to
reject the null hypothesis that the set of MLQ scores is not related to the set of
TKI scores. Although none of the bivariate correlations between the three MLQ
iv
scores and the five TKI scores are statistically significant, some correlations
among the MLQ variables are statistically significant. For example, Laissez-faire
correlation is significantly different from zero. Generally speaking, this means that
those with high (low) scores on Laissez-faire Leadership have low (high) scores
needed to explore the implications of the current study and to corroborate any
implied relationship that may exist between leadership styles and conflict
management styles.
v
Dedication
Acknowledgements
Thanks to Dr. Barry A. Love for his help with my research method and
statistical analysis; to Dr. Felicia Mitchell for her research assistance, editing, and
for listening to me, believing in me, and not giving up on me; to Dr. Jim Duchamp
for retrieving numerous articles and books for me; to Patty Greany, Jane
Caldwell, and David Baber for assistance in obtaining books and articles for my
research via interlibrary loan; to Lisa Eskridge for scanning documents and
encouraging me; to Dr. Chris Fielitz for helping me stay focused; and to Ann
employed at Emory & Henry College. Rick passed away before we could share
Table of Contents
Table 1 ................................................................................................................. 6
Table 2 ............................................................................................................... 45
Table 3 ............................................................................................................... 46
Table 4 ............................................................................................................... 47
Table 5 ............................................................................................................... 47
Table 6 ............................................................................................................... 53
Table 7 ............................................................................................................... 54
Table 8 ............................................................................................................... 54
Table 9 ............................................................................................................... 55
Table 10 ............................................................................................................. 57
Table 11 ............................................................................................................. 57
Table 12 ............................................................................................................. 57
Table 13 ............................................................................................................. 58
Table 14 ............................................................................................................. 59
Table 15 ............................................................................................................. 59
Table 16 ............................................................................................................. 59
Table 17 ............................................................................................................. 59
1
Chapter I – Introduction
organizations (Darling & Walker, 2001; Thomas, 1992). Zigarelli (2002) reports
83 percent of leaders agree that “it’s their job to resolve conflict among their
employees quickly” (p. 4). Dubrin (2004) states leaders “devote about 20 percent
of their time to dealing with conflict” (p. 386) and managers “spend roughly nine
was the first most frequently mentioned topic and conflict ranked as the fifth most
frequently mentioned topic (p. 585). Thus, it is no surprise that leadership and
in research studies.
conflict management style and asserted that “insight into the way individuals
of conflict and thus a useful tool for improved management” (p. 3). Marion used
the Myers-Briggs Type Indicator (MBTI) to assess personality type and the
management style. Marion concluded “the findings from this study suggest that
Kilmann and Thomas (1975) and Mills, Robey, and Smith (1985) that also found
Blanchard (1969a, 1969b, 1974, 1982,1996) while the two conflict style
change” (p. 125). Holt (1986) also examined situational leadership and conflict
management style. Holt also used the LEAD-Self Instrument, but Holt used the
TKI instead of the two conflict management style instruments used by Romero.
leadership styles and conflict management techniques in the sample studied” (p.
73).
Marion’s study (1995) used the MBTI and the TKI while Romero’s (1983)
study used the LEAD-Self Instrument and two other conflict style instruments.
Holt (1986) used the LEAD-Self Instrument and the TKI. No study has specifically
responses to the Multifactor Leadership Questionnaire (MLQ) and the five conflict
The literature (Bass, 1985; Bass & Avolio, 1994; Burns, 1978; Hartog,
Muijen, & Koopman,1997; Yammarino, Spangler, & Bass, 1993) implies, but
does not test or expressly state, that different leadership styles (transactional,
"when one person takes the initiative in making contact with others for the
leadership implies that leaders will enter into bargained-for exchanges with their
Since transactional leaders want to enter into an exchange with their followers,
transactional leaders are not likely to avoid demands by their followers nor
automatically give in and accommodate the demands of their followers. Thus, the
followers, encouraging the followers to become leaders. Bass and Avolio (1994)
state:
4
Transformational leaders will work together with their followers, reminding the
followers of the importance of the team and the organization. Thus, the theory of
distinguish between these active forms of leadership and the “extremely passive
laissez-faire leadership” (p. 21), noting that the laissez-faire leader “is inactive,
rather than reactive or proactive” (p. 21). Laissez-faire leaders “avoid decision
making and supervisory responsibility” (p. 21). Thus, the theory of laissez-faire
leadership styles and conflict management styles. Leadership styles are those
(MLQ). Specifically, these are: (a) leaders’ scores on transactional leadership, (b)
leadership.
the TKI. These are: (a) leaders’ scores on the competing style of conflict
Based on his or her responses to the TKI, each leader received a score ranging
Hypothesis
Table 1 shows the expected relationships (‘+’ for positive, ‘-‘ for negative,
Table 1
MLQ and leaders’ scores on the five conflict management styles (competing,
TKI. Specifically,
H01: The set of MLQ scores is not related to the set of TKI scores.
In order to test this hypothesis, two research instruments (the MLQ and
the TKI) were administered to leaders. The MLQ provided a score for each
leadership. The TKI provided a score for each leader on each of the following five
and accommodating. The data (scores from the MLQ and the TKI) was entered
into SPSS. All three of the MLQ variables (the scores on transactional
faire leadership) were used as well as the five TKI variables (competing,
correlation analysis was used to examine the relationship between the scores on
Scope
completed both questionnaires, the questionnaires were scored; and each leader
in the sample received a score on each of three leadership styles and each of the
five conflict management styles. Specifically, from his or her responses to the
From his or her responses to the TKI, each leader received a score
ranging from 0 to 12 for each of the five conflict management styles (competing,
There are two specific factors that may have influenced the results of this
study. First, the sample size was 99 leaders. Therefore, the results may not be
generalized to all leaders without additional research efforts. The second factor
that may influence the results of this study is the self-reporting nature of the two
chose responses that do not accurately reflect their behavior. This limitation is
not unique to my study, but applies to all research studies that use instruments of
a self-reporting nature.
9
This chapter reviews previous research related to the four areas that are
significant to this study. First, to offer a background for the current study’s focus
Darling and Walker (2001) assert that leaders inevitably face conflict,
different values and situations as the creators of conflict. Darling and Walker
unit appear to be incompatible” (p. 230). They advocate “when such conflict is
organizational benefits will accrue” (p. 230). In their research regarding effective
conflict management, Darling and Walker emphasize that leaders should view
10
Furthermore, they assert that leaders have “a dominant behavior style that is
reflected in how that individual works, interacts, and communicates with others”
(p. 232). If Darling and Walker are correct, then a relationship may exist between
Zigarelli (2002) notes that 83 percent of leaders agree that “it’s their job to
resolve conflict among their employees quickly” (p. 4). His research also states
that 55% of these leaders report they “strongly agree” with the following
statement: “It is a high priority for me to serve the needs of my employees” (p. 4).
people” has been around a long time before Zigarelli’s study. In the 1960s,
Robert Blake and Jane Mouton (1964, 1978) developed “The Managerial Grid.”
The Grid’s horizontal axis represents a “concern for production” while the vertical
axis represents a “concern for people.” Blake and Mouton’s Grid “provides a
or cooperate with one another” (1964, p. 15). Blake and Mouton (1985) assert
that “conflict can be either disruptive and destructive or creative and constructive,
depending on how it is handled” (p. 3). This dissertation explores whether or not
Dubrin (2004) states that, leaders “devote about 20 percent of their time to
dealing with conflict” (p. 386). Dubrin notes that a study by Accountemps reveals
personality clashes” (p. 386). In addition to resolving conflict, leaders have many
other important tasks. Bennis and Nanus (1997) conducted a study consisting of
ninety leaders (p. 18) to obtain a better understanding of leadership: “the most
studied and least understood topic of any in the social sciences” (p. 19). Prior to
whose footprints are everywhere but who is nowhere to be seen” (p. 19). Thus,
including many CEOs from Fortune’s top-200 list (p. 23). The median age of the
leaders in their sample was 56 and the average income (not including “perks”) of
those leaders was $400,000 (p. 23). Their research method included
without effective leadership” (p. 19). They assert that leaders are to effectively
achieve their goals. These goals are developed from the organization’s mission
statement/purpose. Leaders not only see where the organization currently is with
regard to its mission statement and goals, but they also have a plan to help the
organization achieve its goals and its purpose. This is known as vision. Bennis
into action” (p. 29). They proclaim, “management of attention through vision is
12
individuals in the world, and results get attention” (p. 26). After approximately 200
vision, a dreamless society, and this will result, at best, in the maintenance
of the status quo or, at worst, in the disintegration of our society because
Ford (1991), after working closely for thirty-one years with renowned
leader and evangelist Billy Graham, states, “vision is the very stuff of
leadership—the ability to see in a way that compels others to pay attention” (p.
99). He explains that other tasks of a leader include, but are not limited to,
notes that “in recent years conflict resolution has become a major field of study”
(p. 253). He describes Jesus of Nazareth as a model leader and notes that Jesus
dealt “creatively and flexibly with various conflict situations” (p. 254). Ford
emphasizes that today’s leaders must also know how to deal with conflict, and
specifically, how to handle “divisions within modern society, how to tone down
the rhetoric, how to search for acceptable compromises, how to help opposing
parties understand how the other perceives the problems, and how to open the
research clearly affirms the existence and the significance of different conflict
resolution styles. Ford does not explain why some leaders have a need to
dominate and why others have a need to please. This dissertation builds upon
Ford’s research. If Ford is correct in his assertion that some leaders have a need
to dominate their followers while others have a need to please their followers,
then why do these differences exist? This leads one to wonder whether or not
expands upon Ford’s research and examines the relationship, if any, between
that these faculty members send him their syllabi for their courses in
order to produce a list of topics covered in these courses. Rahim reported that
leadership was the most frequently mentioned topic and conflict ranked as the
fifth most frequently mentioned topic (p. 585). Thus, it is no surprise that
dissertation examines the first and fifth most frequently mentioned topics that
programs.
conflict management style. She used the Myers-Briggs Type Indicator (MBTI) to
assess personality type and the Thomas-Kilmann Conflict Mode Instrument (TKI)
Carl Jung and centering on the two functions of perception and judgment” (p. 15).
Marion explains that the MBTI “consists of four scales each with two polar
dimensions” (p. 15). The first scale is attitudes and is measured by extraversion
Marion asserts “insight into the way individuals manage conflict provides a
framework for greater understanding of the dynamics of conflict and thus a useful
tool for improved management” (p. 3). Marion’s assertion provides support for
this research.
conflict management style. As previously stated, she used the TKI to assess
responses” (p. 1) and participants select the response that best characterizes
community colleges. She ran bivariate correlations for each of the personality
dimensions and each of the conflict management styles. Marion concluded “the
findings from this study suggest that there is indeed a relationship between
individual personality type and conflict management style” (p. 9). This
dissertation emerges from the literature, most notably Marion’s finding that there
Kilmann and Thomas (1975) and Mills et al., (1985) also found
correlations between some elements of the MBTI and the TKI. Kilmann and
Behavioral Science for Management) using the MBTI and various conflict-
handling instruments, including the TKI. After analyzing their data, Kilmann and
Thomas conclude, “the results suggest that the Jungian functions related to
(p. 971). Marion (1995) states that her study “confirms most of the Kilmann-
16
Thomas findings with those scoring high on feeling tending to be less assertive
and more cooperative in conflict management” (p. 137). She also concludes,
“there is, in the conflict process, a central role for the individual with his/her
perceptions values, beliefs, and attitudes” (p. 22). Marion emphasizes “that a
and styles” (p. 22-23). If Marion is correct, then this dissertation may provide data
dissertation differs from Marion’s study because the current study explores
measured by the TKI). Their results supported the research of Kilmann and
complete the MBTI and the TKI. After data analysis, the authors conclude, “in
evaluating our hypotheses, we find support for the hypotheses that Thinking and
Feeling are associated with the distributive dimension of conflict handling” (p.
1141). Walton and McKersie (1965) explain that the five conflict-handling styles
and cooperating” (p. 1136) while an integrative dimension “assumes that both
situational leadership and conflict management styles. She gave the Lead-Self
Instrument (Hersey & Blanchard, 1969a, 1969b, 1974, 1982, 1996) and two
conflict style as situations change” (p. 125). This dissertation differs from
leadership styles (as measured by the MLQ) and conflict management styles (as
measured by the TKI). Romero neither used the MLQ nor the TKI.
the relationship between leadership styles and conflict management styles. Holt
used the Lead-Self Instrument and the TKI in her study of 156 middle managers
in various hospitals in North Dakota. After data analysis, Holt concluded “there
conflict management techniques in the sample studied” (p. 73). This dissertation
differs from Holt’s study because this dissertation uses the MLQ to assess
leadership styles while Holt used the LEAD-Self Instrument to assess leadership
styles.
(measured by leaders’ responses to the TKI). The literature (Bass, 1985; Bass &
Avolio, 1994; Burns, 1978; Hartog et al., 1997; Yammarino et al., 1993) implies,
but does not test or expressly state, that different leadership styles (transactional,
MLQ and leaders’ scores on the five conflict management styles (competing,
TKI. In order to answer this question the following null hypothesis was tested:
H01: The set of MLQ scores is not related to the set of TKI scores.
Leadership
that “leadership over human beings is exercised when persons with certain
the motives of followers” (p. 18). Burns explains that leadership is different than
19
power, noting that “to control things—tools, mineral resources, money, energy—
is an act of power, not leadership, for things have no motives. Power wielders
may treat people as things. Leaders may not” (p. 18). He also makes the
assertion that “all leaders are actual or potential power holders, but not all power
many of the men and women in power, but leadership rarely rises to the
full need for it. The fundamental crisis underlying mediocrity is intellectual.
If we know all too much about our leaders, we know far too little about
the modern age and hence we cannot agree even on the standards by
where, and why? How do leaders lead followers without being wholly led
styles.
Transactional Leadership
purely as an exchange occurring "when one person takes the initiative in making
contact with others for the purpose of an exchange of valued things" (p. 19). After
this transaction is complete, the relationship ceases to exist. Burns further states,
"A leadership act took place, but it was not one that binds the leader and follower
The theory of transactional leadership implies that leaders will enter into
an exchange with their followers, transactional leaders are not likely to avoid
their scores on avoiding and leaders’ scores on transactional leadership and their
scores on accommodating.
as those who bargain. The theory of transactional leadership also implies that
transactional leaders do not give in, but instead bargain or negotiate. Thus, it is
22
accommodating).
Transformational Leadership
leadership is present, followers respect and trust the leader. This motivates
followers and encourages the followers to become leaders. Bass also identified
Bass, Avolio, and their colleagues continued to refine this theory and they
Avolio, 1990, 1995). The MLQ intercorrelates the four transformational factors to
leadership. Schmidt (1993) notes that Bass and Avolio used the intercorrelation
of the transformational factors in the MLQ “to discover that if leaders were only
followers, encouraging the followers to become leaders. Bass and Avolio (1994)
state:
Transformational leaders will work together with their followers, reminding the
followers of the importance of the team and the organization. The theory of
leaders work together with their followers. If this is true, then it is reasonable to
leaders desire to work with others and collaborate to resolve conflict, as opposed
Laissez-faire Leadership
occurring in their organizations and acting to solve problems. Hartog et al. (1997)
distinguish between these active forms of leadership and the “extremely passive
laissez-faire leadership” (p. 21), noting that the laissez-faire leader “is inactive,
rather than reactive or proactive” (p. 21). Laissez-faire leaders “avoid decision
making and supervisory responsibility” (p. 21). Since the theory of laissez-faire
leadership implies that laissez-faire leaders are inactive and passive, as opposed
and their scores on avoiding and a negative relationship between leaders’ scores
Blake and Mouton (1964) developed an extremely useful tool called “The
styles. The Grid’s horizontal axis represents a “concern for production” while the
vertical axis represents a “concern for people.” Blake and Mouton’s Grid
assert that “one of the advantages of this classification scheme is that the five
extent to which the individual attempts to satisfy his or her own concerns” (p. 7).
satisfy the other person’s concerns” (p. 7). These two dimensions, represented
by the vertical and horizontal axes, are used to illustrate the five conflict-handling
(2002) as follows:
does not immediately pursue his or her own concerns or those of the other
27
person. He or she does not address the conflict. Avoiding might take the
position. (p. 8)
28
leadership styles that can be measured by the MLQ. Furthermore, the literature
unequivocally reveals that different conflict management styles exist and can be
assessed using the TKI. The literature pertaining to leadership styles and conflict
management styles implies, but does not explicitly state, that a relationship may
Chapter 3 – Method
clearly exist and can be measured by the MLQ. Furthermore, the literature
unequivocally reveals that different conflict management styles exist and can be
assessed using the TKI. The literature pertaining to leadership styles and conflict
management styles implies, but does not explicitly state, that a relationship may
This chapter presents the research method for this study. First, the
MLQ and the TKI) is discussed. Fourth, the sample population and frame is
Hypothesis
Table 1 shows the expected relationships (‘+’ for positive, ‘-‘ for negative,
leadership theory asserts that transactional leaders do not give in, but instead
leaders work together with their followers. If this is true, then is there a positive
that laissez-faire leaders will score high on avoiding and low on collaborating.
tested:
31
H01: The set of MLQ scores is not related to the set of TKI scores.
and laissez-faire) as measured by the MLQ and leaders’ scores on the five
expected relationships posited by the literature (and shown in Table 1), is tested
in this dissertation.
Data Collection
the questionnaires were scored; and each leader in the sample received scores
on each of the three leadership styles and scores on each of the five conflict
management styles.
the TKI. These are: (a) leaders’ scores on the competing style of conflict
Based on his or her responses to the TKI, each leader received a score ranging
The MLQ (Bass & Avolio, 1995, 2000) was chosen for this study because
faire leadership. The MLQ (Form 5X) contains 45 questions and allows
During the years 1997 to 2000, “the latest version of the MLQ, Form 5X, has
been used in nearly 200 research programs, doctoral dissertations and masters
theses around the globe” (Bass & Avolio, 2000, p. 2). Kirnan and Snyder (1995),
in their evaluation of the instrument, state that the MLQ "is designed to be used
at all levels of leadership" (p. 651). Kirnan and Snyder also conclude: "the MLQ
exception); and the nonleadership factor, laissez-faire. Bessai (1995) states, "one
management styles. The MLQ has been used extensively in various research
studies by corporations and individuals. Bessai (1995) states that the instrument
has "good construct validity, adequate reliability, and a strong research base" (p.
Conflict management style was assessed using the TKI. The TKI scores
any, between leadership styles and conflict management styles. The TKI
(Thomas & Kilmann, 1974, 2002) “is designed to assess an individual’s behavior
leader chooses the statement in each pair that best characterizes his/her
behavior.
The TKI was chosen for this study because it is “used extensively in
and development” (Volkema & Bergmann, 1995, p. 6). Thomas and Kilmann
(1978) emphasize that the TKI is “specifically designed to minimize the effect of
social-desirability bias” (p. 1141). There is also immense support for external
validity and reliability of the instrument (Kilmann & Thomas, 1977; Kabanoff,
1987).
Leaders for this study were chosen primarily through referral by alumni
from Emory & Henry College. Various Emory & Henry alumni throughout the
United States were contacted and asked to identify several leaders in their
city/town who would be willing to participate in this study. Using alumni to provide
sample. A total of 120 leaders agreed to participate in the study and thus
instruments and were thus not eligible for inclusion in the statistical analysis.
Analysis
styles and conflict management styles. Leadership styles are those measured by
36
leadership. This study did not assign leaders to one of the three leadership
styles. Instead, this study used the scores on the leadership styles as measures
laissez-faire) is a categorical variable, the same leader exhibits all three styles to
a different degree.
The MLQ provided scores on: TR, TF, and LF. TR is “scores on”
leadership (ranging from 0 to 16) on the four MLQ questions for laissez-faire
leadership.
did not assign leaders to a specific conflict management style. Instead, this study
used scores as measures of the extent to which a leader is competing, the extent
the extent to which a leader is avoiding, and the extent to which a leader is
instruments (the MLQ and the TKI) were administered to leaders. The MLQ
leadership, and laissez-faire leadership. The TKI provided a score for each
MLQ and the TKI were entered into SPSS. The MLQ variables (the scores on
that is used to analyze the relationship between two sets of variables. Hair,
state “in situations with multiple dependent and independent variables, canonical
correlation is the most appropriate and powerful multivariate technique” (p. 444).
Thompson (1980) also discusses the power of canonical correlation and the
“theoretically rich” (p. 17) results that can be obtained from this procedure. He
emphatically declares:
39
The purpose of this dissertation was to explore the relationship (if any)
leadership styles and conflict management styles were obtained from a sample
determine whether to reject or fail to reject the null hypothesis that the set of
Chapter 4 – Results
leaders’ responses to the TKI). This chapter discusses data collection and then
presents the results found upon examination of the leaders’ responses to the
MLQ and the TKI. The results include general observations of the data gathered
Data Collection
For this study, a leader was defined as a person who guides or instructs
others. Leaders were chosen for this study primarily through referral by alumni
from Emory & Henry College. Various Emory & Henry alumni throughout the
United States were contacted and asked to identify several leaders in their
city/town who would be willing to participate in this study. Using alumni to provide
41
sample.
permission to reproduce up to 150 copies of the MLQ) was purchased from Mind
Garden and 120 copies of the Thomas-Kilmann Conflict Mode Instrument were
purchased from CPP. Coding of the leadership instruments (the MLQs) and the
conflict management style instruments (the TKIs) was essential not only to
maintain confidentiality, but also to ensure that leaders’ scores on the leadership
style instruments. The instruments were numbered from 1 to 120 to ensure that
pertaining to the type of organization in which the leader served, the leader’s
gender, the number of months the leader has served in a leadership position, the
number of people in the organization, and the number of people who report to the
42
leader. The Demographic Questionnaires (and the envelopes that contained the
research instruments for the participants in the study) were also numbered.
A total of 120 packets were prepared for the sample. Each envelope contained a
letter to each leader (see Appendix), briefly explaining the data collection
TKI. These 120 packets were then distributed to leaders. A total of 105 leaders
returned their packets. Of the 105 packets returned, the data collected from 99
respondents were usable for statistical analysis. Six of the respondents left
several questions on the MLQ and/or the TKI unanswered and therefore,
analysis, state that sample size should include “at least 10 observations per
variable to avoid ‘overfitting’ the data” (p. 447). This study contains eight
variables. Thus, this study’s sample size of 99 leaders is adequate for canonical
correlation analysis.
Upon receipt of the packets, the MLQs and the TKIs were scored. Each
leader who returned a completed packet received scores on each of the three
leadership styles and scores on each of the five conflict management styles. The
data (for each of the 99 leaders) were entered into an Excel spreadsheet and
then imported into a software package (SPSS) for statistical analysis. These data
The data collected from the leadership instrument (MLQ) and the conflict
management style instrument (TKI) were imported into SPSS for statistical
styles and conflict management styles. The leadership styles were measured by
leaders’ scores on the TKI (specifically, leaders’ scores on the five conflict
accommodating).
Table 2 presents the minimum and maximum scores on the MLQ and TKI
for leaders participating in this study. The mean and standard deviation are also
shown. It is important to note that for the MLQ leaders could receive a score for
scores could range from 0 to 48. Therefore, the average score was only slightly
above half of the maximum score available, indicating that, on average, the
44
leaders in this sample did not score high on Transactional Leadership. The mean
Leadership was 2.38, indicating that, on average, the leaders in this sample
The results of this study indicate that, on average, Compromising was the
first most frequently used conflict management style, and on average, Avoiding
was the second most frequently used conflict management style. Holt’s (1986)
management style (p. 76). Holt reported that Collaborating was the second most
frequently used conflict management style in her study (p. 77). The results of my
study indicate that, on average, Avoiding was the second most frequently used
conflict management style and, on average, Collaborating was the third most
Table 2 also indicates that some leaders scored the maximum possible on
also indicate that some leaders rely heavily on Compromising and Avoiding.
45
Table 2
MLQ Scores
TKI Scores
Tables 3 and 4 present the frequency distributions of the MLQ and TKI
scale from 0 to 48, symmetrical about a midpoint of 24. In contrast, the scores for
to 80. Almost all of the Laissez-faire scores were less than or equal to six on a
scale from 0 to 16. A comparison of the TKI frequencies in Table 4 with those for
MLQ scores in Table 3 shows that all five of the TKI scores were more evenly
Table 3
Table 4
0 2.0 2.0 0.0 0.0 0.0 0.0 0.0 0.0 6.1 6.1
2 12.1 14.1 4.0 4.0 0.0 0.0 5.1 5.1 27.3 33.4
4 27.3 41.4 12.1 16.1 15.2 15.2 18.2 23.3 27.3 60.7
6 29.3 70.7 29.3 45.4 21.2 36.4 27.3 50.6 15.2 75.9
8 13.1 83.8 30.3 75.7 35.4 71.8 30.3 80.9 13.1 89.0
10 16.2 100.0 20.2 95.9 24.2 96.0 18.1 99.0 9.0 98.0
12 0.0 4.0 99.9 4.0 100.0 1.0 100.0 2.0 100.0
Table 5
MLQ Scores
TR 1.000
TF .110 1.000
LF .359** -255* 1.000
TKI Scores
Competing .101 .045 -.061 1.000
Collaborating -.114 .086 -.032 -.003 1.000
Compromising -.120 .157 -.084 -.328** -.137 1.000
Avoiding -.008 -.154 .017 -.337** -.379** -.357** 1.000
Accommodating .099 -.133 .150 -.576** -.477** -.031 .134 1.000
Table 5 shows that none of the bivariate correlations between the three
MLQ scores and the five TKI scores are statistically significant; however, Table 5
does show that some correlations among the MLQ variables are statistically
significant. For example, LF and TR are positively correlated and the correlation
general, those with high (low) scores on Laissez-faire Leadership have high (low)
Generally speaking, this means that those with high (low) scores on Laissez-faire
The data presented in Table 5 also show that 6 correlations among the
and Competing is highly significantly different from zero. The correlation between
and Accommodating and Collaborating are also highly significantly different from
zero.
MLQ scores and the five TKI scores are statistically significant. These results are
transactional leaders will enter into “bargained for” exchanges, it would seem
and Compromising while correlating negatively with Avoiding. Two of these three
positive and the correlation between Transactional Leadership and Avoiding was
will work together with their followers. The theory also implies that
passive, not active, and will avoid making decisions. The theory thus implies a
and Avoiding. Once again, however, the correlations were not statistically
To see if the set of three MLQ scores is related to the set of five TKI
scores, a canonical correlation was performed. The objectives of the analysis are
as follows:
In this analysis two canonical variates—one for the MLQ scores and one
for the TKI scores—are defined as linear combinations of their respective scores.
For example, the three MLQ scores imply a canonical variate MLQ1, where
For the first canonical root, the α and β coefficients are selected to give the
several tests and analyses were performed. For multicollinearity Table 5 shows
that there is one significant pairwise correlation among the MLQ scores and
some significant pairwise correlations among the TKI scores; however, in all
cases the correlations are below .80. The highest--between Accommodating and
problem if the “correlations of r > .80” (p. 45). All of the correlations are below
.80.
Table 6 gives the results of the Shapiro-Wilks tests for normality of the
MLQ and TKI scores. We fail to reject normality at the .05 significance level in all
cases but that of scores for the Laissez-faire leadership style and the Competing
highly skewed) does not decrease the correlation with other variables. (p.
448)
52
Since the correlations in both of these cases are very high and transforming
Laissez-faire would require using logs that are not defined for almost twenty
percent (as shown for LF scores of zero in the frequency distribution of MLQ
Scores presented in Table 3) of the sample, these variables were left unchanged.
Tabachnick and Fidell (1996) support this decision and explain the reasoning
and for failures of normality, linearity, and homoscedasticity, they are not
from the variables that are in it and transformed variables are sometimes
To test for linearity between the TKI scores and the MLQ scores, between
the canonical variates and their respective set of variables, and between the
scatterplots between each TKI score and each MLQ score. There are no signs of
Table 6
Variable r p
MLQ Scores
TR .9958 0.87255
TF .9984 0.97984
LF .9713 0.00001
TKI Scores
Figures 3 and 5 give the scatterplots between the canonical variates and
observations in the rightward third of the plot. (Thus, all of the LF scatterplots
between the canonical variates and their opposite set of variables. Again, none of
completeness, Figure 7 plots the relationships between the canonical variates for
there are three MLQ scores. Results of significance tests for the canonical
correlations between the set of five TKI scores and the three MLQ scores are
shown in Tables 7 and 8. The first canonical correlation is only .298 and in all
tests the data fail to reject the null hypothesis that all three canonical correlations
are zero. Furthermore, the F-tests in Table 8 confirm the statistical insignificance
relationship between the three MLQ scores and the five TKI scores.
Table 7
Table 8
Approximate
Statistic Value F Statistic p
Table 9 gives the results of a redundancy analysis for the three canonical
composites (canonical variates), even though these linear composites may not
(Hair et al., 1998, p. 451). In effect, the index weights the canonical correlation
with the correlation between the variate and its set of variables by multiplying the
Table 9
Redundancy Analysis
the squared correlations between the variate and its set of variables. A high
independent canonical variate) and a high amount of shared variance (i.e., a high
percentage of the variance in the variate explained by its set of variables) (Hair
The redundancy indices in Table 9 are very low for the first TKI variate
(.022) and the first MLQ variate (0.030). In both cases, the low indices result from
the low canonical R2 and the low shared variances (.247 for the TKI variate and
.341 for the MLQ variate). The redundancy results for the second and third
through 13. The raw canonical coefficients in Table 10 can be used to write
equations for each of the three canonical variates implied from the five TKI
scores. For example, the five TKI scores imply a canonical variate TKI1 for the
Furthermore, the data in Table 12 can be used to write equations for the three
canonical variates implied from the three MLQ scores. The following equation is
Table 10
Table 11
Table 12
Table 13
Tables 11 and 13 give the standardized coefficients for TKI scores and
Competing and Accommodating were the most important for their first canonical
variate and scores on Compromising were the least important. Table 13 shows
important for their first canonical variate and scores on Laissez-faire Leadership
optimize the canonical correlation, the canonical loading and cross-loadings are
Table 14
Table 15
Table 16
Table 17
The correlations between canonical variate TKI1 and the five TKI variables
scores are more highly correlated with the MLQ variate than the Laissez-faire
leadership scores are. Recall that the relatively low loadings shown in Tables 14
the TKI scores and the MLQ variates (in Table 16) and low correlations between
the MLQ scores and the TKI variates (in Table 17). These correlations imply that
very low percentages of the variance in the variables are explained by the
opposite variate. For example, from the information in the first row and column of
Table 16, only about 0.2 percent (.04825 squared) of the variance in the scores
The last row and first column of the same table shows that only 2.8 percent
TKI variate shows that the variate only explains about 4 percent of the variation
in that leadership style. The small correlations between the original variables and
their opposite canonical variate confirm the insignificant relationships among the
61
three pairs of canonical variates shown in Tables 7 and 8 and contribute to the
relationship exists between the set of leadership styles (as assessed by the
MLQ) and the set of conflict management styles (as assessed by the TKI).
Results of significance tests for the canonical correlations between the set of five
TKI scores and the three MLQ scores are shown in Tables 7 and 8. The first
canonical correlation is only .298 and in all tests the data fail to reject the null
hypothesis that all three canonical correlations are zero. The F-tests in Table 8
sample offers no evidence of a relationship between the three MLQ scores and
Chapter 5 – Discussion
upon conflict management styles. The study’s research design explains the data
gathering process. The results section presents the data in a meaningful format
and explains how canonical correlation was used to analyze the relationship
between the MLQ scores and the TKI scores. This study clearly demonstrates
Table 5 shows that none of the bivariate correlations between the three
MLQ scores and the five TKI scores are statistically significant; however, Table 5
does show that some correlations among the MLQ variables are statistically
significant. For example, LF and TR are positively correlated and the correlation
general, those with high (low) scores on Laissez-faire Leadership have high (low)
Generally speaking, this means that those with high (low) scores on Laissez-faire
63
found in this study confirm the results found by Bass and Avolio (2000) in their
examination of the intercorrelations among MLQ scores, using data from 2,154
leaders.
The results of this study also confirm some of Holt’s (1986) findings. Holt
used the LEAD-Self Instrument and the TKI to examine leadership styles and
conflict management techniques in the sample studied” (p. 73). The LEAD-Self
behavior and classifies leadership styles as: participating, selling, delegating, and
telling. This study differed from Holt’s in two specific ways: (a) this study used the
MLQ (not the LEAD-Self) to assess leadership styles and (b) this sample was not
While some of the results in this study confirm some of Holt’s (1986)
findings, other results of this study differ significantly from Holt’s previous
management style in her study. Collaborating was second, Competing was third,
Avoiding was fourth, and Accommodating was fifth. The current study’s results
64
used conflict management style; however, the rankings for second, third, fourth,
and fifth most frequently used conflict management styles all differ from Holt’s
study. The current study found that Avoiding was the second most frequently
fourth, and Competing was fifth in the sample studied. These findings may be the
result of the inclusion of leaders from an extensive vocational range or the use of
The results of this study show support for six of expected bivariate
shows the expected relationships (‘+’ for positive, ‘-‘ for negative, and ‘0’ for no
laissez-faire leadership styles and the five conflict management styles. Table 5
were negatively correlated. Although the literature implies that leadership styles
may be related to conflict management styles, this study was not able to
extremely complex.
Observations/Limitations
leadership styles and conflict management styles. Although there may be many
reasons that explain why a statistically significant relationship was not found in
the sample studied, four specific reasons include (a) the self-reporting nature of
the research instruments, (b) the possibility that leaders may behave differently
(i.e., depart from their normal leadership style) in conflict situations, (c) the
organizational culture/climate that the leader is currently working in, and (d) the
possibility that the leadership theories and literatures may inadequately describe
possible that leaders’ responses on the research instruments may have reflected
how they perceive themselves behaving or how they would like to behave, rather
than their actual behavior. It is also possible that leaders may depart from their
normal leadership style when faced with conflict. A leader may have answered
the questions on the MLQ with regard to their leadership behavior in situations
It is possible that leadership style may not be the predominant factor that affects
describe the traits and behaviors of leaders. For example, Burns (1978)
and their scores on the Compromising style of conflict management. The results
of this study do not show a positive correlation; instead, the results show a
Implications of Findings
style indicates that when resolving conflict, the leader is moderately concerned
with his or her own needs and moderately concerned with the needs of others.
Collaborating style is used when the leader is extremely assertive and extremely
cooperative. The Avoiding style is used when the leader is extremely unassertive
67
and extremely uncooperative. Finally, the Accommodating style is used when the
Tables 3 and 4 provide frequency distributions for leadership styles and conflict
management styles. This study found Compromising (see Table 2) to be the first
partial win-win situation for both parties; however, Collaboration may provide
Table 2 also shows that Avoiding is the second most frequently used
conflict management style. This may reflect a leader’s desire to obtain more
information before making a decision. This could also indicate that a leader
prefers to avoid conflict. It is possible that leaders in this study use the Avoiding
style too frequently and the Collaborating style too infrequently. Use of the
for both parties. Due to the high scores on Avoiding, it is evident that leaders do
not realize the inherent dangers of overuse of this conflict management style.
Leaders who scored in the high range on Avoiding should consider the possibility
that their organizations may be suffering as a result of their own lack of input
(Thomas & Kilmann, 2002, p. 14). Leaders who scored in the high range on
Avoiding should also ask: “Are decisions on important issues sometimes made
68
by default” (p. 14)? Organizations should provide leaders with training programs
management styles. Future studies using the MLQ could also use the
Leadership, but one could also examine the individual scores for the five
Stimulation, and (e) Individualized Consideration (Bass & Avolio, 1995, 2000).
69
may help leaders more efficiently and more effectively resolve conflict.
70
References
Bass, B. M., & Avolio, B. J. (Eds.). (1994). Introduction. In B.M. Bass, & B.J.
Bennis, W., & Nanus, B. (1997). Leaders: The strategies for taking charge. New
of Mental Measurements.
Blake, R. R., & Mouton, J. S. (1964). The managerial grid. Houston: Gulf.
Blake, R. R., & Mouton, J. S. (1978). The new managerial grid (2nd ed.).
Houston: Gulf.
71
Blake, R. R., & Mouton, J. S. (1985). The managerial grid III (3rd ed.). Houston:
Gulf.
Darling, J. R., & Walker, W. E. (2001). Effective conflict management: use of the
22(5), 230-242.
Hair, J. F., Jr.. Anderson, R. E., Tatham, R. L., & Black, W. C. (1998).
Mulitvariate data analysis (5th ed.). Upper Saddle River, NJ: Prentice Hall.
Hersey, P., & Blanchard, K. H. (1969a). Life cycle theory of leadership. Training
Hersey, P., & Blanchard, K. H. (1974). So you want to know your leadership
Holt, T.L. (1986). The relationship between leadership styles and conflict
1606.
Mills, J., Robey, D., & Smith, L. (1985). Conflict-handling and personality
1135-1143.
Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.).
Psychologists Press.
81-102.
Synergy.
76
Appendix
Demographic Questionnaire
June 2, 2004
Dear
I need your help in identifying leaders for my study. Please select leaders in your
workplace, church, or community. For purposes of this study, a leader is a
person who guides or instructs others.
Enclosed are several small white envelopes. Each of these small white
envelopes contains research instruments that the leaders will complete. Please
give each leader one white envelope and ask them to complete the research
instruments, seal the envelope, and return the envelope to you. I have also
included instructions for the leaders in their envelope.
If leaders do not return these to you in a few days, please be sure to follow up
with the leader to ensure his/her completion of the research instruments. The
leaders should return the white envelopes to you by Friday, June 11, 2004. After
you have received the completed instruments in the small white sealed
envelopes, please put all of the small white envelopes into the large green and
white envelope and return to me. Please mail the large green and white
envelope to me as soon as possible, preferably no later than Monday, June 14,
2004.
Sincerely,
June 2, 2004
Dear leader,
Please note that all of your responses are confidential. While instruments and
envelopes are coded with numbers, these codes will only be used to link the
multiple instruments being administered.
Sincerely,
Competing 10
10 20 30 40
TR
10
Competing
40 50 60 70 80
TF
10
Competing
0 5 10
LF
12
Collaborating
10 20 30 40
TR
12
Collaborating
40 50 60 70 80
TF
12
Collaborating
0 5 10
LF
Figure 2. (Continued)
88
12
11
10
9
Compromising
8
7
6
5
4
3
10 20 30 40
TR
12
11
10
9
Compromising
8
7
6
5
4
3
40 50 60 70 80
TF
12
11
10
9
Compromising
8
7
6
5
4
3
0 5 10
LF
Figure 2. (Continued)
89
12
Avoiding
10 20 30 40
TR
12
Avoiding
40 50 60 70 80
TF
12
Avoiding
0 5 10
LF
Figure 2. (Continued)
90
10
Accommodating
10 20 30 40
TR
10
Accommodating
40 50 60 70 80
TF
10
Accommodating
0 5 10
LF
Figure 2. (Continued)
91
-1
-2
MLQ1
-3
-4
-5
10 20 30 40
TR
-1
-2
MLQ1
-3
-4
-5
40 50 60 70 80
TF
-1
-2
MLQ1
-3
-4
-5
0 5 10
LF
-1
-2
MLQ1
-3
-4
-5
0 5 10
Competing
-1
-2
MLQ1
-3
-4
-5
2 7 12
Collaborating
-1
-2
MLQ1
-3
-4
-5
3 4 5 6 7 8 9 10 11 12
Compromising
-1
-2
MLQ1
-3
-4
-5
2 7 12
Avoiding
-1
-2
MLQ1
-3
-4
-5
0 5 10
Accommodating
Figure 4. (Continued)
94
20.5
19.5
18.5
TKI1
17.5
16.5
15.5
0 5 10
Competing
20.5
19.5
18.5
TKI1
17.5
16.5
15.5
2 7 12
Collaborating
20.5
19.5
18.5
TKI1
17.5
16.5
15.5
3 4 5 6 7 8 9 10 11 12
Compromising
20.5
19.5
18.5
TKI1
17.5
16.5
15.5
2 7 12
Avoiding
20.5
19.5
18.5
TKI1
17.5
16.5
15.5
0 5 10
Accommodating
Figure 5. (Continued)
96
20.5
19.5
18.5
TKI1
17.5
16.5
15.5
10 20 30 40
TR
20.5
19.5
18.5
TKI1
17.5
16.5
15.5
40 50 60 70 80
TF
20.5
19.5
18.5
TKI1
17.5
16.5
15.5
0 5 10
LF
20.5
19.5
18.5
TKI1
17.5
16.5
15.5
-5 -4 -3 -2 -1 0
MLQ1
87
86
85
TKI2
84
83
82
81
-5 -4 -3 -2 -1 0
MLQ2
-33
-34
-35
TKI3
-36
-37
5 6 7 8 9 10
MLQ3