Citation
The effects of personality and charismatic leadership on creative performance as mediated by creative self-efficacy

Material Information

Title:
The effects of personality and charismatic leadership on creative performance as mediated by creative self-efficacy
Creator:
Strickland, Susan Lara
Publication Date:
Language:
English
Physical Description:
97 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Personality ( lcsh )
Charisma (Personality trait) ( lcsh )
Performance ( lcsh )
Leadership ( lcsh )
Self-efficacy ( lcsh )
Performance standards ( lcsh )
Charisma (Personality trait) ( fast )
Leadership ( fast )
Performance ( fast )
Performance standards ( fast )
Personality ( fast )
Self-efficacy ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 91-97).
General Note:
Department of Psychology
Statement of Responsibility:
by Susan Lara Strickland.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
57708143 ( OCLC )
ocm57708143
Classification:
LD1190.L645 2004m S77 ( lcc )

Full Text
THE EFFECTS OF PERSONALITY AND CHARISMATIC LEADERSHIP
ON CREATIVE PERFORMANCE
AS MEDIATED BY CREATIVE SELF-EFFICACY
by
Susan Lara Strickland
B.A., Pomona College, 1995
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Masters of Arts
Psychology
2004


This thesis for the Master of Arts
degree by
Susan Lara Strickland
has been approved by
Date


Strickland, Susan Lara (M.A., Psychology)
The Effects of Personality and Charismatic Leadership on Creative Performance as
Mediated by Creative Self-Efficacy
Thesis directed by Assistant Professor Annette J. Towler
ABSTRACT
An organizational field study was conducted to test the effect of personality
and charismatic leadership on creative job performance. The combination of
charismatic leadership, high openness to experience, and low conscientiousness
was predicted to be related to the highest creative job performance. Creative self-
efficacy was hypothesized to mediate this relationship. Data was collected from
170 employee-supervisor pairs. In partial support of the hypothesized model of
creativity, openness to experience was positively related to creative performance,
and this relationship was mediated by creative self-efficacy. There was an
interaction between charismatic leadership and openness to experience such that
employees with high openness had high creative performance regardless of leader
charisma while employees with low openness had higher creative performance
with a charismatic leader than a non-charismatic leader. Creative self-efficacy was
related to both charismatic leadership and creative performance providing indirect
evidence that charismatic leadership may be related to creative performance.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
Signed
m
Annette J. Towler


ACKNOWLEDGEMENT
I would like to thank my advisor, Annette Towler, for her support and guidance. I
would like to thank my committee members, Mike Cook and Joy Berrenberg, for
giving their time and suggestions. I would like to thank Herman Aguinis for his
valuable advice about statistical methods. And last but certainly not least, I would
like to express my great thanks to Melissa Farin for giving so many hours of her
time to code materials for the study.


CONTENTS
Figures..........................................................viii
Tables...........................................................ix
CHAPTER
1. INTRODUCTION.................................................1
Models of Creativity.....................................3
Personality............................................ 6
Openness to Experience...........................8
Conscientiousness............................... 9
Conscientiousness, Openness, and
Creativity-Related Job Performance..............11
Leadership..............................................16
Person-Environment Fit..........................17
Leader-Member Exchange..........................18
Leader-Induced Goal Orientation, Problem
Orientation, and Self-Efficacy..................19
Charismatic Leadership..........................21
The Interaction of Personality and Charismatic Leadership.28
Motivation,
30


Creative Self-Efficacy........................32
Self-Efficacy and Personality.................33
Self-Efficacy and Charismatic Leadership......41
Creative Self-Efficacy as a Mediator..........45
2. METHOD....................................................47
Research Setting, Participants, and Procedure.........47
Measures..............................................49
3. RESULTS.................................................. 58
Personality, Charismatic Leadership and
Creative Performance..................................62
Interaction between Leadership and Personality........64
Creative Self-Efficacy................................65
4. DISCUSSION................................................70
Limitations...........................................75
Directions for Future Research........................78
Implications for Organizations........................78
APPENDIX
A. PERSONALITY MEASURE.......................................81
B. CHARISMATIC LEADERSHIP MEASURE............................83
C. CREATIVE SELF-EFFICACY MEASURE............................84
vi


D. DEMOGRAPHIC QUESTIONS.....................85
E. SUPERVISOR CREATIVITY RATING FORM.........86
F. COVER LETTER TO EMPLOYEES.................87
G. COVER LETTER TO SUPERVISORS...............89
REFERENCES.........................................91
vii


FIGURES
Figure
1.1 Hypothesized Model of Creativity......................................2
3.1 Interaction between Openness to Experience and Charismatic
Leadership on Creative Performance....................................65
viii


TABLES
Table
2.1 Sample Characteristics: Job Category.......................................54
2.2 Sample Characteristics: Level of Education.................................55
2.3 Sample Characteristics: Tenure with the Company............................55
2.4 Sample Characteristics: Length of Time with Current Supervisor.............56
2.5 Sample Characteristics: Age................................................56
2.6 Sample Characteristics: Gender.............................................57
2.7 Sample Characteristics: Race...............................................57
3.1 Means, Standard Deviations, and Intercorrelations of Variables..............59
3.2 Regression of Creative Performance on Openness
and Charismatic Leadership.................................................63
3.3 Regression of Creative Performance on Openness
Controlling for Creative Self-Efficacy.....................................68
IX


CHAPTER 1
INTRODUCTION
As we move towards a knowledge-based economy with increased
competition through globalization and rapid changes in technology, innovation
becomes ever more vital to organizational success (Drucker, 1992). Innovation
has been defined as the successful implementation of creative ideas (Amabile,
1988, p.126). As such, creativity in organizations is getting increasing attention in
the research spotlight. Creativity is the production of something that is both novel
and useful (e.g., Amabile, 1983; Feist, 1998; Guilford, 1950), and researchers have
been in near perfect agreement about the basic definition for decades (Feist, 1998).
Thus, in order to be considered creative, a product or idea must be original and
must be appropriate to the current situation or have some kind of social value.
Social value might be aesthetic, practical, or intellectual. The usefulness criterion
is necessary to differentiate creativity from schizophrenic or bizarre products or
ideas (Feist, 1998).
The present study will propose and test a model of the effects of
charismatic leadership and the personality factors of openness and
conscientiousness on creative performance via their influence on creative self-
efficacy. First, after an explanation of interactionist approaches to creativity, I will
l


review the literature linking individual differences in personality to creativity.
Second, I will discuss the relationship between charismatic leadership and creative
performance. Third, I will argue that creative self-efficacy mediates the joint
effects of personality and charismatic leadership on creative performance.
Figure 1.1
Hypothesized Model of Creativity
The present study will extend existing research by studying the joint effects
of individual differences in personality and charismatic leadership on creative
performance. Previous research in creativity has focused on individual differences
(e.g., Feist, 1998) or on charismatic leadership (e.g., Bono & Judge, 2003) but has
not investigated the joint effects of these variables. Additionally, this study will
2


will contribute to the research by asking why individual differences and
charismatic leadership have an effect on creative performance. I propose that
creative self-efficacy acts as a mediator.
There is a relatively long history of creativity research in individual
differences dating back to the 1950s, predominantly conducted by personality
psychologists (reviewed by Feist, 1998). In the most recent two decades, the focus
has shifted to establishing empirical support for more comprehensive models of
creativity. In the 1980s, Teresa Amabile developed a social psychological
approach to creativity (Amabile, 1983). In a similar vein, in the 1990s, Ford
(1996) and Woodman, Sawyer and Griffin (1993) proposed that creative behavior
is influenced by the individual, the context, and social factors. An increasing
number of studies provide empirical support for these interactionist models.
Models of Creativity
Amabiles componential working model was one of the first
interactionist models of creativity (1983, p. 66). Unlike earlier perspectives on
creativity, interactionist models do not look at any single factor in isolation, but
instead view creativity as a result of both individual and situational factors.
Amabile outlines many assumptions behind her model. First, creative ability
occurs along a continuum and can vary even for a particular person at different
3


times or in different situations or domains. Second, certain individuals may be
best suited for creativity within particular domains such as poetry or psychology
(Feldman, 1980). Third, while individual differences play a role, they do not
explain all differences in creative production. As evidence of this, creativity can
be enhanced (Stein, 1974) and formal education is often a necessary prerequisite
(Feldman, 1980). Fourth, individual differences in cognitive ability, education,
personality and talent are related to creativity, but do not seem sufficient for
creativity. In other words, a person who has high cognitive ability, is well
educated, has a creative personality may or may not produce creative works. And
lastly, creative problems are often solved after conscious work on the problem has
stopped, in a moment of sudden revelation (e.g., Poincare, 1924).
Amabiles componential framework includes three main areas: domain-
relevant skills, creativity-relevant skills and task motivation (1983). All three
components are proposed to interact with each other and the level of creativity will
be a function of the three components. The central role of task motivation
differentiates Amabiles model from the focus on individual differences in
isolation from social and environmental factors that dominated creativity research
prior to the early 1980s. In particular, intrinsic motivation is expected to enhance
creativity and extrinsic motivation is expected to decrease creativity. In support of
the importance of task motivation, Amabile cites research showing that reward,
4


choice, and evaluation expectation affect creativity (Amabile, 1979; Amabile,
Goldfarb, & Brackfield, 1982; Glucksberg, 1962; Glover & Gary, 1976; Amabile,
Goldberg, & Capotosto, 1982; Amabile & Gitomer, 1982). Much social
psychological research has shown that reward undermines intrinsic motivation,
and this phenomenon is usually explained in terms of over-justification, a
cognitive process whereby people interpret their own behavior as being a function
of the external reward (e.g., Kruglanski, Alon, & Lewis, 1972). However,
Amabile argues that cognitive processes alone do not explain all of the research
findings and affective processes may play a key role (e.g., Amabile, DeJong, &
Lepper, 1976). Additionally, reward can only undermine intrinsic interest when
there is a high level of intrinsic interest (Amabile, 1983).
Another major interactionist model of creativity was developed by
Woodman, Sawyer, and Griffin in 1993. This organizational creativity model is an
extension of an earlier model of individual creativity (Woodman and Schoenfeldt,
1989). Woodman et al. (1993) propose that individual, group, organization, and
environmental factors are related to creativity. Individual creativity is a function
of antecedent conditions, cognitive style, personality, knowledge, intrinsic
motivation, social influences, and contextual influences from both the organization
and the external environment. Group creativity is a function of individual
creativity, group composition, group characteristics, group processes, and
5


contextual influences. Organizational creativity is a function of group creativity
and contextual influences.
Ford developed a comprehensive theory of creative action with the purpose
of integrating creativity and innovation research in 1996. Creativity research has
primarily been concerned with individuals and groups, whereas innovation
research has focused on organizations. According to Fords theory, problem
orientation, motivation, and knowledge-ability factors determine creative or
habitual action. A problem-finding orientation with many alternative schemata
facilitates creative action. Motivation is comprised of goals (for creativity,
independence, and achievement), beliefs about rewards for creativity, confidence in
creative ability, and emotions (interest, anger, and pleasure rather than anxiety and
boredom). Knowledge and ability are comprised of domain-related knowledge
(i.e., broad expertise), behavioral abilities (i.e., communication skills, social
networking skills), and creative-thinking ability (i.e., divergent thinking, ability to
associate).
Personality
Empirical research has shown a consistent relationship between personality
factors and creative performance in both qualitative and quantitative reviews
(Barron & Harrington, 1981; Feist, 1998). Many different personality traits have
6


been associated with creativity, including independence, tolerance for ambiguity,
and emotional variability (Barron & Harrington, 1981). Feists (1998) meta-
analysis aggregated studies employing a variety of different personality measures,
(e.g., 16PF, Adjective Check List, Eysenck Personality Questionnaire (EPQ),
California Psychological Inventory (CPI)) by mapping them to the Five-Factor
Model. Feist also compared differences between measures and narrower
personality dimensions than the Big Five and concluded:
A consistent and clear portrait of the creative personality... has
emerged: Creative people are more autonomous, introverted, open
to new experiences, norm-doubting, self-confident, self-accepting,
driven, ambitious, dominant, hostile, and impulsive. (Feist, 1998,
p. 299)
Of the Big Five, openness to experience and conscientiousness were the
strongest differentiators between scientists and non-scientists, between more
creative and less creative scientists, and between artists and non-artists (Feist,
1998). The effect sizes (Cohens d) for openness were small to medium. The
effect sizes for conscientiousness were generally medium to large. Smaller effects
were found for neuroticism, extroversion, and agreeableness in the meta-analysis.
Furthermore, the prior research findings regarding the effects of neuroticism,
extroversion, and agreeableness on creativity are inconsistent (e.g. Dollinger &
Clancy, 1993).
7


Openness to Experience
Individuals high on openness to experience tend to be imaginative, to be
independent thinkers, and to prefer variety and depth of experience (McCrae,
1996). Openness is abroad construct and encompasses vivid fantasy, artistic
sensitivity, depth of feeling, behavioral flexibility, intellectual curiosity, and
unconventional attitudes. (McCrae, 1996, p. 323). Open individuals have a
reoccurring need to examine and broaden their experience (McCrae, 1996). They
have a greater penchant for uncertainty and complexity than their closed
counterparts. Closed individuals, those low on openness, prefer familiarity,
simplicity, and closure. The quality of openness is thought to come from
differences in mental structure and functioning. McCrae argues that openness is
the personality dimension with the biggest impact on a variety of social behaviors
including political ideology, marriage and family, friendships, authoritarianism,
and attitudes about change.
Openness to experience may play a role in the development of adult
identity (Tesch & Cameron, 1987). Tesch and Cameron studied identity
development and openness in a sample of young, college-educated adults ages 18
to 30. Interviews were used to assess identity in the areas of occupation, religion,
politics, and sex role. Raters then scored the taped interviews in order to measure
8


commitment, current and past exploration of alternatives, and involvement (time,
energy, and emotion). Individuals high on openness to experience were more
likely to have explored their identity in the past or at the time of the study.
Individuals high on openness also expressed less intense commitment to their
present identity. Thus, openness may be associated with greater identity flexibility
and greater likelihood of considering a life change.
Conscientiousness
Alongside openness to experience, conscientiousness is the other
dimension of the Five- Factor Model with a strong and relatively consistent
relationship to creativity (Feist, 1998). Unlike openness to experience,
conscientiousness tends to be negatively related to creativity (Feist, 1998).
Conscientiousness is considered to encompass general motivational
predispositions (e.g., Martocchio & Judge, 1997). Individuals high on
conscientiousness tend to control their impulses and follow established norms
(Costa & McCrae, 1992). They are generally cautious, responsible, organized,
persevering, and achievement-striving (Costa & McCrae, 1992). Because
conscientious individuals tend to follow the established ways of doing things, they
may be less likely to generate novel ideas or solutions that are outside of the norm.
The cautious nature of conscientious individuals may lead them to avoid the risk of
9


creativity, that is, the risk of trying something that may not work. Also, because
conscientious individuals control their impulses in general, this could lead them to
control creative impulses if they see such impulses as counter-productive.
In order to provide a complete picture of conscientiousness it is important
to note that, in spite of its negative relationship to creativity, conscientiousness has
the strongest and most consistent positive relationship to overall job performance
of the Big Five (Barrick & Mount, 1991). The correlation (r) between
conscientiousness and performance was found to be .31 and job type was not
found to affect the strength of this relationship (Mount & Barrick, 1995).
Furthermore, conscientiousness is essentially unrelated to cognitive ability (r =
.02, Ackerman & Heggestad, 1997). It is worth noting that while artists are quite
low on conscientiousness compared to the general population, such is not the case
with scientists (Feist, 1998). Scientists are more conscientious than non-scientists
by about half a standard deviation, but more creative scientists are less
conscientious than less creative scientists. Based on these findings, it seems likely
that some degree of conscientiousness may be necessary for success in demanding
jobs that require originality. At the extreme, the negative pole of
conscientiousness is described as psychopathic deviant (Feist, 1998), hardly a
desirable quality in employees.
10


Because there is limited research investigating conscientiousness and
openness specifically on creative job performance, in the next section I will review
research that investigates the role openness and conscientiousness on broader types
of job performance. In particular, job performance under changing task conditions
(Le Pine, Colquitt, and Erez, 2000) requires adaptive and divergent thinking, and
these are considered to be creativity skills.
Conscientiousness, Openness, and
Creativity-Related Job Performance
A study by Le Pine, Colquitt and Erez (2000) sheds some further light on
the relationship between conscientiousness, openness to experience, and job
performance. They conducted a laboratory study to investigate whether
conscientiousness, openness to experience, and cognitive ability could predict
decision-making performance under changing task contexts. Subjects were
undergraduates enrolled in a management course. In exchange for their
participation, they received course credit and the opportunity to compete for prizes
awarded to the best three performers. Subjects completed measures of cognitive
ability, conscientiousness and openness during the first week of class and prior to
the laboratory task. The task was a three-hour computer simulation requiring risk
assessment of unidentified aircraft. Subjects received training in assessing nine
characteristics of the aircraft and were given three different risk-assessment rules,
11


named Alpha, Bravo, and Charlie. The rules each used certain airplane
characteristics to determine the magnitude of threat. However, subjects were not
told which rule to apply. They were told that the rules might lead to different
decisions and that the rules would not necessarily be equally useful for arriving at
the correct decision. During the simulation, subjects assessed the threat and
decided a course of action for each aircraft on a 7-point range of aggressiveness.
Immediately after each response, subjects were given feedback consisting of the
absolute value of the difference between their response and the ideal response.
During a set of training trials, the airplane characteristics were constructed in such
a way that using any of the three rules would yield the same correct threat
response. The researchers chose this design in order to prevent subjects from
learning that one of the rules was better than another prior to beginning the real
trials. Once the real test was underway, applying different rules would often lead
to different decisions. During the first 25 trials, applying Rule Charlie would
correctly predict the threat. During the second 25 trials, Rule Bravo would yield
the correct decision. During the last 25 trials, Rule Alpha would yield the
correct decision. Le Pine and his colleagues found that only cognitive ability
predicted performance during the first 25 trials, prior to the change in context, (i.e.,
the change in which rule needed to be applied). Openness to experience and
cognitive ability were positively associated with performance after the change in
12


context, (i.e., after the first 25 trials). Conscientiousness was negatively related to
performance after the change in context. Additionally, performance during the
first 25 trials was only moderately correlated with performance during the second
25 trials and third 25 trials (r = .32 and .44 respectively). Performance during the
second 25 trials and third 25 trials was highly correlated (r = .72). Based on these
findings, Le Pine and his colleagues suggest that there is something inherently
different about performance under changing conditions than performance under
static conditions. Openness to experience appears to facilitate adaptation whereas
conscientiousness seems to inhibit adaptation.
Recent research has investigated the interaction between personality and
context on creative performance. George and Zhou (2001) looked at specific
situational characteristics and the Big Five personality factors. Employees in a
variety of office jobs filled out measures of openness to experience,
conscientiousness, feedback valence (the extent to which supervisor feedback was
positive or negative), supervisor monitoring, the nature of job tasks, and several
work environment factors. The employees were not aware the measures were for a
study on creativity. On a separate questionnaire, supervisors rated employee
creative performance. George and Zhou (2001) found that openness to experience,
positive feedback, and task type predicted high creativity in a three-way
interaction. A particular combination of all three variables was associated with
13


maximum creativity in their organizational sample. The creative individual
needed to be high on openness to experience, supervisor feedback needed to be
perceived as positive, and the task needed to have either unclear ends or unclear
means.
Tasks are considered heuristic to the extent that they have unclear ends,
unclear means, and/or multiple means. In the George and Zhou (2001) study, only
unclear ends and unclear means were related to high creativity after controlling for
the other task types. The same task can fall into more than one category, (that is,
unclear ends, unclear means, or multiple means), and so the task types are
correlated with each other. George and Zhou controlled for the other two task
types in their moderated regression analysis before entering the interaction term
for task type, openness to experience, and feedback valence. George and Zhou
found that both tasks with unclear ends or unclear means interacted with openness
to experience and high feedback valence to account for additional variance in
creative performance. In contrast, multiple means combined with openness to
experience and high feedback valence did not account for additional variance in
creative performance. George and Zhou conclude that personality alone is not
sufficient for high creativity in an organizational context: openness to experience
may indicate creative tendencies, but it must be encouraged by the situation in
order to be maximally expressed.
14


Additionally, George and Zhou (2001) found conscientiousness, close
monitoring and work environment interacted to predict the lowest creativity. In
particular, creativity was lowest when the individual was high on
conscientiousness, the supervisor was perceived to be monitoring work closely,
and the work environment created by peers was negative. The work environment
was considered negative to the extent that communications were inaccurate,
coworkers were unhelpful, and the environment was generally negative. George
and Zhou interpreted their results to mean that highly conscientious employees can
still be creative, but must be encouraged to do so by the environment.
In keeping with the past studies of personality and creativity, it is expected
that:
Hypothesis la: Individuals high on openness to experience will
have higher levels of creative performance.
Hypothesis lb: Individuals high on conscientiousness will have
lower levels of creative performance.
Another pioneering interaction study that used a relatively broad individual
difference measure was Oldham and Cummings (1996). The research was
conducted in two manufacturing facilities. Employees completed the Creative
Personality Scale (CPS; Gough, 1979) along with measures of job complexity,
supervisory style, and intention to quit. Creative behavior was assessed with
supervisor ratings and patent disclosures. Oldham and Cummings found that
15


creative personality, job complexity, and non-controlling supervision were
positively related to creativity outcomes (patents and supervisor ratings).
In summary, individual characteristics are related to creative performance,
particularly openness to experience and conscientiousness (Feist, 1998). Some
studies have shown that a combination of employee personality and environmental
characteristics is related to creative behavior in the workplace (George & Zhou,
2001; Oldham & Cummings, 1996). Many of the environmental characteristics
related to creativity appear to be related either directly or indirectly to leadership
and management styles. Examples of workplace characteristics with an apparent
connection to leadership include helpfulness of coworkers (George & Zhou, 2001)
and non-controlling supervision (Oldham & Cummings, 1996). In the following
sections, I will build a case for the connection between leadership and creative
performance based on existing theory and previous research, with a focus on
charismatic leadership.
Leadership
Many different aspects of leadership have been studied with regard to their
effect on performance. Leader-Member Exchange (LMX) research investigates
the nature of the relationship between a supervisor and employee at the dyad level
(e.g., Tierney, Farmer, & Graen, 1999). Charismatic or transformational
16


leadership research seeks to determine the extent to which followers are motivated
to go above and beyond expectations when supervised by an inspirational leader
(e.g., Towler, 2003). In contrast to LMX models where one supervisor may have a
different relationship with each subordinate, charismatic and transformational
models imply that the leader has a similar effect on many followers or attracts
many individuals who are similarly influenced by the leader (Conger & Kanungo,
1987). Other research has studied the effect of leadership style on goal orientation,
problem orientation, and self-efficacy (Redmond et al., 1993). First, I will review
several studies that investigate the relationship between leadership and creative
performance (Livingstone, Nelson, and Barr, 1997; Tierney, Farmer, & Graen,
1999; Redmond, Mumford, & Teach, 1993). Then, I will focus on charismatic
leadership and will include studies with any kind of task or job performance as the
criterion (e.g., Fuller, Patterson, Hester, & Stringer; Howell & Frost, 1989)
because only one set of studies was found to-date that investigated charismatic
leadership and performance on a creative task (Bono & Judge, 2003).
Person-Environment Fit
The person-environment fit (P-E fit) approach makes predictions based on
the interaction between the individual and the environment (Livingstone, Nelson,
and Barr, 1997). While not strictly a leadership theory, P-E fit has important
r
17


implications for leadership since the workplace environment is largely shaped by
leadership. P-E fit models include two different versions of fit: supply-value and
demand-ability (Livingstone et al., 1997). Supply-value fit is how well individual
needs match the supplies in the environment. Demand-ability fit is how well an
individuals abilities match the demands of the environment.
Livingstone and her colleagues (1997) studied P-E fit on a creativity
dimension in a cross-sectional study in three different organizations. They found
that demand-ability fit with regard to creativity was associated with lower job
strain and higher satisfaction. However, the strongest predictor of strain,
satisfaction, performance and commitment was the work environment by itself
(supply), not an interaction between person and environment. The work
environment for creativity was measured with the Work Environment Inventory
(WEI) and included items such as People are rewarded for creative work in this
organization and In this organization, there is a lively and active flow of ideas
(Amabile & Gryskiewicz, 1989). Thus, the environment for creativity here is a
quite broad construct.
Leader-Member Exchange
Creativity may be influenced by the interaction between supervisor and
employee. Tierney, Farmer, and Graen (1999) found evidence that the styles of
18


supervisors and employees and the quality of the Leader-Member Exchange
(LMX) relationship interacted to predict creativity outcomes in an R&D sample.
LMX explained incremental variance in creativity after controlling for employee
and supervisor traits. Cognitive adaptors were more creative in high LMX dyads,
whereas cognitive innovators had high creativity regardless of the quality of the
relationship with their supervisor. Cognitive adaptors accept the given definition
of problems and generate ideas consistent with established norms whereas
innovators redefine problems and generate ideas outside the established norms
(Kirton, 1976). Tierney et al. (1999) also extended previous empirical evidence on
intrinsic motivation with their finding that creative performance was highest when
both employee and supervisor had high intrinsic motivation.
Leader-Induced Goal Orientation.
Problem Orientation, and Self-Efficacy
Redmond, Mumford, and Teach (1993) investigated the relationship
between leader behaviors and follower creative performance and motivation.
Subjects were undergraduates who had completed at least one marketing course.
In a laboratory study, Redmond and his colleagues manipulated three leadership
variables: self-efficacy, goal orientation, and problem construction. Subjects were
asked to do marketing exercises that entailed writing descriptions of a target
market, a television commercial, and a magazine advertisement for a new product.
19


Trained confederates played the role of leader and followed a predetermined
script. For the self-efficacy manipulation, the leader either stated that the subjects
test scores were in the top 20% (high self-efficacy condition) or that the subjects
test scores were average (low self-efficacy condition). For the goal orientation
manipulation, the subject was told either that the purpose of the task was to help
them understand the approach used to solve marketing problems (learning goal
condition) or that the task v/ould be used to assess their skill in solving marketing
problems (performance goal condition). In the problem construction condition,
subjects were asked to state problems and considerations that might influence
advertising success for this product and to generate their own problem statement
prior to beginning the marketing exercises. In the no-problem construction
condition, this instruction was omitted. In order to measure creative performance,
judges rated the quality and originality of the completed marketing exercises. The
amount of time a subject spent on the task and the words written were also used as
measures of time and effort respectively. Redmond and his colleagues found that
leader-induced self-efficacy and problem construction were positively related to
higher creative performance, while goal orientation was not. The correlations
between quality and originality were about .80, indicating that no meaningful
distinction was made between quality and originality in judging the marketing
exercises.
20


Redmond, Mumford, and Teach (1993) found evidence that leader
behaviors affected motivation (time and effort) as well as performance. Subjects
spent more time on the marketing exercises when leaders gave the problem
construction or when leaders induced self-efficacy. No main effect was found for
goal condition, but an interaction effect was found with goal condition and
problem construction. In the performance goal condition, problem construction
had no effect on time spent, but in the learning goal condition, subjects spent more
time if there was problem construction. The researchers interpret this finding to
suggest that lack of performance concerns may decrease motivation unless the
leader increases involvement and commitment with problem construction
behaviors. Finally, subjects in the high self-efficacy condition wrote more words
that those in the low self-efficacy condition.
Charismatic Leadership
Charismatic leaders use vision or ideological goals, instill confidence,
inspire others, need influence, and are self-confident, dominant, articulate and
unconventional (Conger & Kanungo, 1987). One of the shortcomings of the
charisma literature has been the lack of a consistent definition of charismatic
leadership and the difficulties differentiating it from the related concept of
transformational leadership (Fuller, Patterson, Hester, & Stringer, 1996).
21


According to Bass, charismatic leadership is one component of transformational
leadership (Bass, 1985a, 1985b, 1990). Bass states that transformational
leadership, as measured by the Multifactor Leadership Questionnaire (MLQ),
includes charisma plus individualized consideration, intellectual stimulation, and
inspiration.
Motivation is thought to be an important factor in determining creative
action, (e.g., Amabile, 1983; Zhou & Oldham, 2001), and will be discussed in
more detail later. If motivation affects creative performance, leadership that
increases motivation should increase creativity. Transformational leaders are
believed to increase congruence between individual goals and organizational
goals, autonomy, and motivation of their followers (House, 1995). Individuals
who are independent and self-directed will prefer participative leadership over
directive leadership (e.g., Abdel-Halim, 1981), and, as discussed earlier, creativity
is associated with independence and self-direction in both personality (e.g., Feist,
1998) and situation (e.g., Zhou & Oldham 2001). Therefore, highly creative
persons would be expected to prefer participative leadership and thereby to prefer
charismatic leaders.
In a meta-analysis of 32 studies, Fuller and his colleagues (1996) found
that charismatic leadership was most highly correlated with satisfaction with the
leader, perceived leader effectiveness, and performance. The charisma subscale of
22


the MLQ had a .45 overall mean correlation with performance. This correlation
was moderated by the research design, such that studies with multi-source
performance data had a lower correlation of .34. Studies collecting both
performance and charisma data at the same time using the same instrument had a
higher correlation of .48, (i.e., the MLQ was used to measure both charisma and
outcomes). This finding provides empirical evidence in support of the
methodological superiority of multi-source research designs. It is important to
note that the vast majority of studies in the Fuller et al. meta-analysis were
correlational in design and did not manipulate charisma. The handful of studies
found to-date with experimental manipulations of charisma will be discussed in
more detail (Howell & Frost, 1989; Kirkpatrick & Locke, 1996; Bono & Judge,
2003; Shea & Howell, 1999, for role of self-efficacy as a mediator).
Howell and Frost (1989) tested the proposition that charismatic leadership
would lead to higher task performance in the first known laboratory study to
manipulate charisma. The task was a management in-basket exercise, rather than a
creative performance task. Trained actors played the role of leader. The
leadership manipulation was charismatic, considerate or structuring. In the
charismatic condition, the leader used both content and style associated with
charisma. The content behavior included stating a transcendent goal, high
performance expectations, and confidence in follower ability. The style behavior
23


included using an engaging voice, making eye contact, pacing, and using lively
facial expressions. In the structuring leader condition, factual information was
given in detail with a neutral tone. In the considerate leader condition, the leader
was warm, friendly, and engaged in conversation with the follower. Howell and
Frost found that charismatic leadership led to high task performance, as well as
high adjustment to the task, leader, and group. Howell and Frost did not separately
manipulate any of the various aspects of charisma, such as visionary content.
To further explore the relationship between charisma and performance,
Kirkpatrick and Locke (1996) separately manipulated three different aspects of
charisma: vision, charismatic communication style, and vision implementation
(task cues). A trained actor played the role of the leader in their laboratory study.
In the vision condition, the leader told two stories to convey the vision and the
importance of quality. In the no vision condition, the leader presented factual
information for an equal length of time. In the charismatic communication style
condition, the leader shook hands, made eye contact, used hand gestures, and had
animated facial expressions and a dynamic tone of voice. In the non-charismatic
leadership condition, the leader was friendly and polite but kept a neutral
demeanor toward the task and participants and did not show much variation in
voice tone or facial expressions. In the task cue condition, the leader gave four
suggestions for improving work procedures. The no task cue condition omitted
24


these suggestions. While vision and charismatic communication style are traits
considered to be unique to charismatic leaders, task cues may be used by more
transactional leaders as well.
Kirkpatrick and Lockes subjects were undergraduate students in a business
class who received extra credit for participating in the study. Each subject
participated in one two-hour session conducted in small groups of 4-12
participants. The task involved inserting pages into notebook binders and
simulated work done in printing companies; the task was designed to be routine.
Following a practice trial conducted by the experimenter (to be used as a measure
of initial ability), the leader arrived, introduced the study, and gave instructions.
The subjects then completed two 15-minute binding trials. Each of the three core
components of charismatic leadership had different effects on followers. Vision
was related to more trust, congruence, and inspiration, but had only a marginally
significant effect on task performance. Vision implementation or task cues,
operationalized here as specific suggestions for how to do the task, led to higher
task performance in both quality and quantity. It is important to note that task cues
may be used by non-charismatic leaders as well as charismatic ones; therefore, the
importance of task cues (vision implementation) to follower performance is not
a strong argument by itself for the advantages of charismatic leadership. A
charismatic communication style was not associated with more positive attitudes
25


or task performance. Kirkpatrick and Locke suggest that content may be more
important than style in a business environment. However, as noted by Kirkpatrick
and Locke, both levels of the charismatic communication style manipulation were
perceived as charismatic. Because charisma is fundamentally an attribution made
about a leader (e.g., Willner, 1984), this may be problematic.
Bono and Judge (2003) conducted both a field and laboratory studies to
investigate the relationship between transformational leadership and performance,
hypothesizing that the relationship would be mediated by goal self-concordance.
Self-concordance is defined as the degree to which an individuals genuine values
and interests are expressed in work goals and tasks (Sheldon & Elliot, 1999).
While their field study only measured overall performance and did not specifically
assess creative performance, the laboratory study did include a creative task.
In their field study, Bono and Judge (2003) collected data in nine different
organizations of varying industries. Followers (i.e., subordinates) evaluated their
leader (i.e., supervisor) with the Multifactor Leadership Questionnaire and
completed a measure of goal self-concordance. About two months later, the leader
evaluated follower job performance and attitude. Also at this time, followers
evaluated their job attitude and satisfaction. Bono and Judge found that
transformational leadership was related to self-concordance. Self-concordance
was associated with higher organizational commitment and job satisfaction and
26


acted as a partial mediator between leadership style and job satisfaction and
commitment. No relationship was found between self-concordance and job
performance.
In their laboratory study, Bono & Judge (2003) found that transformational
leadership led to higher levels of creative performance. Subjects watched a video
of an actor whom they believed was a local restaurant owner. When the restaurant
owner was a transformational leader, subjects generated more ideas for names and
slogans for the restaurant. Creative performance was measured strictly by total
number of ideas without any judgment about the quality of the ideas. In Bono et
al.s leadership manipulation, transformational leadership included statements of
vision, values, high expectations, and unconventional practices. To communicate
these ideas, the transformational leader used stories, imagery, and repeated
phrases. Although Bono et al. attempted to control most of the leadership
manipulation through the script, the actor also used facial movements and tone of
voice to express optimism and passion characteristic of transformational
leadership (Bono & Judge, 2003, p. 565). Because the script and non-verbal
behaviors were not separately manipulated, it is unknown whether the increase in
creative performance was caused by the statement content combined with
rhetorical techniques or by the communication style or both.
27


In summary, previous research has shown a consistent, positive
relationship between charismatic leadership and general performance in field
studies (Fuller et al., 1996; Shamir et al., 1993), and partial support in
experimental studies where charisma was manipulated and tasks were not creative
(Howell & Frost, 1989; Kirkpatrick & Locke, 1996; Shea & Howell, 1999 see
self-efficacy section). One set of studies found a positive relationship between the
closely related concept of transformation leadership and performance on a creative
task (Bono & Judge, 2003). Existing theory supports the ideas that leader
charisma will increase follower motivation (House, 1995) and that increased
motivation will bring about higher creative performance (e.g., Amabile, 1983;
Zhou & Oldham, 2001). Thus,
Hypothesis 2\ Charismatic leadership will be associated with
greater creative performance.
The Interaction of Personality and Charismatic Leadership
Past findings indicate that openness to experience positively
predicts creative performance. Multiple factors that support creativity are
expected to interact with each other and enhance creativity beyond both
their contributions alone. Both charismatic leadership and openness to
experience are predicted to have positive relationships with creative
performance. Consequently, I predict that the combination of charismatic
28


leadership and openness to experience components will have the strongest
effect on creative performance, rather than the simple sum of the
components. Research investigating creativity within a social context has
found that a combination of factors results in the highest creativity (e.g.,
George & Zhou). Thus,
Hypothesis 3a: There will be a two-way interaction of charismatic
leadership and high openness to experience on creative
performance, such that the highest creative performance will be
achieved when both charismatic leadership and openness to
experience are high.
Past research indicates that conscientiousness is related to lower creative output
(Feist, 1988). As discussed earlier, there is some evidence that the negative
relationship between conscientiousness and creative performance only holds true
under certain conditions, such as when combined with a negative work
environment and close supervision (George & Zhou, 2001). Conscientiousness is
expected to interact with factors inhibiting creativity such that a combination of
factors has a greater impact than a set of isolated factors. In the present study, the
absence of charismatic leadership is expected to be the condition under which
conscientiousness will inhibit creativity. Thus,
Hypothesis 3b: Conscientiousness and charismatic leadership will
interact to affect creative performance such that individuals who are
high on conscientiousness will exhibit the lowest level of creative
performance when their supervisor in low on charismatic leadership
than for any other combination of these variables.
29


Motivation
Intrinsic motivation is defined as doing something because it is inherently
interesting or enjoyable (Ryan & Deci, 2000). In contrast, extrinsic motivation is
doing something in order to get a tangible outcome, such as a reward (Ryan &
Deci, 2000). Research has shown that intrinsic motivation is positively related to
creativity (e.g., Tierney et al., 1999). Evidence suggests that intrinsic motivation
has a trait component (e.g., Amabile, Hill, Hennessey, & Tighe, 1994) but that it is
susceptible to influence by environmental factors such as reward structure
(Amabile, Conti, Coon, Lazenby, & Herron, 1996). Furthermore, intrinsic
motivation is proposed to mediate the effect of environmental factors on creativity
(e.g., Amabile, 1983). In other words, environmental factors may influence
creativity via their influence on intrinsic motivation which in turn affects
creativity. I will review one study investigating motivation for creativity from the
perspective of Cognitive Evaluation Theory (Zhou & Oldham, 2001). Then, I will
discuss creative self-efficacy (Tierney & Farmer, 2002) which is to be the focus of
the present study. Lastly, I will review several studies linking self-efficacy to
personality (Martocchio & Judge, 1997; Lee & Klein, 2002; Gully, Payne, Koles,
& Whiteman, 2002) and to leadership (Shea & Howell, 1999; Kirkpatrick and
Locke, 1996).
30


According to Cognitive Evaluation Theory, motivation is believed to be
highest when information is high and external control is low (Deci & Ryan, 1980,
1985; Ryan 1982). To provide further evidence of the relationship between
motivation and creativity, Zhou & Oldham (2001) designed a situation with
differing levels of external control and information. Following Cognitive
Evaluation Theory, they hypothesized that high information and low control would
lead to the highest creativity. Subjects were asked to generate novel and useful
solutions to an in-basket exercise. External control was manipulated at three
different levels: subjects were given expectations of self-assessment, expert
assessment, or no assessment. While the level of information was not manipulated
independently, the researchers interpreted both self and other assessment
conditions as high information because they included a statement of the benefits of
assessment for personal creative development. The no assessment condition was
considered to be low information. As expected, subjects who assessed their own
creativity-related skills later exhibited higher levels of creativity, whereas subjects
assessed by others did not.
In addition to studying the level of information and external control, Zhou
and Oldham (2001) administered Goughs 1979 Creative Personality Scale after
subjects completed the in-basket task. Higher scores on the Creative Personality
Scale predicted more creative solutions to the in-basket exercise. Personality and
31


assessment condition interacted such that subjects with a creative personality
expecting to assess their own work displayed the highest creative performance on
the in-basket task. Subjects with low scores on the Creative Personality Scale
performed similarly across self, other and no assessment conditions. Zhou and
Oldham speculate that less creative individuals may have had unsuccessful
experiences with creative tasks and do not appreciate the value of developmental
opportunities or believe in their potential benefit. This study provides evidence of
a complex interaction between intrinsic motivation, context, and individual
differences.
Creative Self-Efficacy
Confidence in ones ability to be creative has been proposed to enhance
creativity and to be a component of motivation (e.g., Ford, 1996; Amabile, 1983).
A recent study by Tierney and Farmer (2002) established creative self-efficacy as
a new construct and demonstrated its relationship to creative performance.
Tierney and Farmer gathered data from two different organizational sources: the
manufacturing division of a large consumer products company and in a high tech
firm. Independent variables included creative self-efficacy, job self-efficacy,
supervisor support, job tenure, and education. The dependent variable was
creative performance. A factor analysis was conducted on a preliminary sample
32


and resulted in a reliable three-item measure of creative self-efficacy. Employees
were asked to consider their work role when responding to general questions, such
as I have confidence in my ability to solve problems creatively, on a 7-point
Likert scale. Reliability was confirmed by the alphas of .83 for the manufacturing
sample and .87 for the operations sample. Job self-efficacy was measured with a
previously established scale (Spreitzer, 1995). Creative performance was
measured through supervisor ratings on a Likert scale. The results indicated that
creative self-efficacy was distinct from job self-efficacy. Creative self-efficacy
was positively related to creative performance, even after controlling for tenure,
education, supervisor support, complexity, and job self-efficacy. Creative self-
efficacy itself was predicted by job self-efficacy, supervisor support, and job
complexity, providing evidence that these contribute to creative self-efficacy. Job
tenure was negatively related to creative self-efficacy. When creative self-efficacy
was high, job self-efficacy was a moderator of creative performance.
Self-Efficacy and Personality
Because creative self-efficacy is a newly measured construct (Tiemey &
Farmer, 2002) and, therefore, a largely unexplored avenue of research, I will draw
upon the self-efficacy literature to build a case for the connection between creative
self-efficacy and personality. Self-efficacy is believed to play a role in the
33


relationship between personality and performance. Self-efficacy is considered to
be a proximal construct, acting in specific tasks and situations (Chen, Casper, &
Cortina, 2001). hi contrast, individual differences such as personality and
cognitive ability are considered to be distal, in other words, general and stable.
Because self-efficacy operates proximally, it has been proposed as a possible
mechanism by which individual differences assert themselves (e.g., Phillips &
Gully, 1997).
Judge and Ilies (2002) conducted a meta-analysis of the correlations
between Big Five personality traits and performance motivation. Performance
motivation was conceptualized in terms of goal-setting, expectancy, and self-
efficacy. However, self-efficacy is considered to be a more proximal influence on
motivation than goal-setting or expectancy. All three are cognitive theories of
motivation seeking to predict on-the-job performance. Data from 65 studies was
included, with a total of 150 correlations. Personality traits from measures not
directly assessing the Big Five were included if they could be associated with one
of the five factors in the judgment of psychologist raters. Based on 14
correlations, conscientiousness had an estimated true correlation of .22 with self-
efficacy. Conscientiousness had similar correlations with goal setting and
expectancy motivation. Based on 3 correlations, openness to experience had an
estimated true correlation of .20 with self-efficacy. Openness to experience had a
34


weaker relationship to goal-setting motivation and was slightly negatively related
to expectancy. Neuroticism was negatively related to all three motivation
constructs, with a predictive strength similar to conscientiousness. Extroversion
had a fairly large correlation with self-efficacy, (estimated true correlation of .33),
but much smaller correlations with goal-setting and expectancy. Agreeableness
had small and inconsistent relationships with the three motivation constructs. The
Big Five traits together had an average multiple correlation of .49 with
performance motivation, thus explaining about 24 percent of the variance.
Chen, Casper, and Cortina (2001) tested the hypotheses that self-efficacy
mediates the relationships between conscientiousness and performance and
between cognitive ability and performance. They also investigated whether task
complexity would moderate these relationships. Chen, Casper, and Cortina
conducted a primary meta-analysis of 33 studies that employed task specific self-
efficacy measures, as opposed to general self-efficacy measures. Then, they
combined their meta-analytic data with data from four earlier meta-analyses to
form two correlation matrices and did a path-analysis on the resulting correlations.
The researchers found that conscientiousness was more highly correlated with self-
efficacy prior to complex tasks than simple ones. The researchers explain this
finding by relating conscientiousness to motivation. Conscientiousness is the
motivational personality component (e.g., Martocchio & Judge, 1997) and
35


motivation may be activated more by difficult tasks than simple ones (e.g., Daniel
& Esser, 1980). Chen and colleagues found evidence that for simple tasks, self-
efficacy partially mediated the relationships between conscientiousness, cognitive
ability, and performance. In contrast, for complex tasks self-efficacy did not seem
to act as a mediator. Self-efficacy was not found to predict performance beyond
that predicted by conscientiousness and cognitive ability alone. Chen and
colleagues proposed that self-efficacy was not found to predict additional
performance due to their focus on field research. The majority of studies
included were field research, but it is not evident what the proportions were of
field versus laboratory studies. This explanation is based on prior research that has
found self-efficacy to be a better predictor of performance on laboratory tasks than
in organizational settings (Stajkovic & Luthans, 1998). Additionally, larger effects
have been found for self-efficacy in laboratory settings where it is often
manipulated (Cortina & DeShon, 1998).
Martocchio and Judge (1997) conducted research to determine whether
self-efficacy and self-deception mediated the relationship between
conscientiousness and learning. Self-deception is a sincere but positively biased
self-view. Subjects were clerical and administrative university employees in a
Windows 3.1 software training course. At the beginning of the first training
session, subjects completed measures of cognitive ability, computer knowledge,
36


Windows 3.1 knowledge, conscientiousness, self-efficacy, and self-deception. The
training was two sessions long and included both lecture and practice. At the end
of the second training session, subjects completed measures of self-efficacy and
learning. Conscientiousness was positively related to both self-deception and self-
efficacy. Self-efficacy was then found to mediate the relationship between
conscientiousness and learning, such that individuals with high self-efficacy
learned more. Self-deception also mediated the relationship between
conscientiousness and learning, but such that individuals engaging in high self-
deception learned less. For self-efficacy, the path of their structural equation
model went from conscientiousness to pre-training self-efficacy to post-training
self-efficacy to learning. For self-deception, the path was from conscientiousness
to self-deception to learning. Additional links directly from conscientiousness to
post-training self-efficacy or from conscientiousness to learning did not improve
the model fit. In surprising contrast to the positive relationship between
conscientiousness and job performance (Barrick & Mount, 1991), Martocchio and
Judge found a negative zero-order correlation between conscientiousness and
learning (r = -.27). In a case where dual mediators are working in opposite
directions, the strength of the zero-order correlation is determined by the relative
effects of the mediators (Lee & Klein, 2002).
37


Lee and Klein (2002) sought to understand the circumstances affecting the
zero-order correlation between conscientiousness and learning and the relative
impacts of self-efficacy and self-deception as mediators. They extended upon
Martocchio and Judge (1997) by testing changes in the relative strengths of self-
efficacy and self-deception over time. Lee and Klein used an introductory
international business course as their training. Subjects were students enrolled in
the course who volunteered to participate in exchange for extra credit. The course
was 10 weeks long and was primarily lecture and discussion. Subjects completed
measures of conscientiousness, self-deception, and initial self-efficacy during
Week 2 of the course. Midterm exams on Week 4 and Week 8 were the learning
measures. Additionally, on Week 5, following feedback from the first midterm
exam, subjects completed a second measure of self-efficacy. Unlike the
Martocchio and Judge (1997) study, neither cognitive ability nor pre-training
knowledge was assessed or controlled for. Lee and Klein found that
conscientiousness was positively related to self-efficacy early in training (Week 2),
but not after the first midterm (Week 5). Conscientiousness was also positively
related to self-deception. All of the correlations between self-efficacy (early and
later) and learning (first and second midterm grades) were positive. Self-deception
was negatively related to learning at midterm 1, but had no relationship to the
second midterm grade. In spite of these correlations, self-efficacy and self-
38


deception were not found to mediate the relationship between conscientiousness
and learning when their indirect effects were measured in the structural equation
model. The importance of time in determining the proportion of influence from
self-efficacy versus self-deception was supported in two different ways. First,
there was no zero-order correlation between conscientiousness and early learning,
but conscientiousness was positively related to later learning. Second, Lee and
Klein tested the impact of time in their structural equation model by comparing
versions of the model with and without a path between self-deception and later
learning (midterm 2). Adding a path between self-deception and later learning did
not improve the model fit which suggests that self-deception did not have
independent impact on later learning. The Lee and Klein study failed to replicate
Martocchio and Judges (1997) finding that self-efficacy and self-deception were
dual mediators of the relationship between conscientiousness and learning. Lee
and Klein point out that this could be due to the difference in statistical tests that
were used: Martocchio and Judge (1997) did not include a direct test of mediation
using structural equation modeling to measure the indirect effects.
Gully, Payne, Koles, and Whiteman (2002) investigated the relationship
between individual differences and error treatment in training on post-training self-
efficacy and performance. Subjects were undergraduate psychology students.
Prior to the training, subjects completed measures of conscientiousness and
39


openness to experience. SAT scores were used as a measure of cognitive ability.
The researchers randomly placed subjects into three different training conditions:
error-encouragement, error-avoidance or control (no direct mention of errors).
The training session included instructions, time to read the manual, a practice trial,
and a longer training trial. The task was a computer simulated radar and decision-
making task. At the end of the training, subjects completed measures of self-
efficacy and declarative knowledge.
Gully and his colleagues found main effects for cognitive ability and
openness to experience on all three training outcomes, such that individuals high
on either of these characteristics had higher self-efficacy, task performance, and
declarative knowledge. Of greater interest to their hypotheses, several interactions
were found between individual differences, error treatment during training, and
outcomes. Individuals with high cognitive ability had higher self-efficacy and task
performance when encouraged to make errors. Individuals with high
conscientiousness had lower self-efficacy when encouraged to make errors, but no
difference was found between the control and error-avoidance groups. Individuals
high on openness to experience had higher self-efficacy, declarative knowledge
and task performance when encouraged to make errors during training. These
findings show how both individual traits and environmental characteristics may
interact to influence self-efficacy and performance. While the present study did
40


not directly test the hypothesis that self-efficacy mediated the relationship between
individual traits and error training, the pattern of relationships found for openness
and cognitive ability would be consistent with self-efficacy acting as a mediator.
In summary, these studies collectively paint a complex picture of the
relationship between conscientiousness, openness to experience, self-efficacy and
performance, such that the relationship seems to be influenced by characteristics of
the environment and task. One study found that self-efficacy and self-deception
mediated the relationship between conscientiousness and performance during
training (Martocchio & Judge, 1997) but another study found that they did not
(Lee & Klein, 2002). On simple tasks but not complex tasks, self-efficacy
mediated the relationship of conscientiousness and cognitive ability on job
performance (Chen et al., 2001). Conscientious individuals had lower self-
efficacy when encouraged to make errors during training, while open individuals
had higher self-efficacy, knowledge, and task performance when encouraged to
make errors (Gully et al., 2002)
Self-Efficacy and Charismatic Leadership
Shamir, House and Arthur (1993) posit that self-efficacy is the mechanism
by which charismatic leaders affect followers. Charismatic leaders are proposed to
raise their followers perceptions of self-efficacy which in turns brings about
41


extraordinary performance. Shamir and his colleagues also hypothesize that
charismatic leaders will be most effective when performance goals are difficult to
quantify, when the means to achieve goals cannot be readily delineated, and when
extrinsic rewards cannot be coupled with the performance of individuals.
Shea and Howell (1999) tested the proposition that the relationship
between charismatic leadership and performance was mediated by self-efficacy.
They manipulated 2 different variables: leader charisma (charismatic or non-
charismatic) and task feedback (internal, external, or none). Subjects were
graduate management students, and they were randomly assigned to one of the six
cells. Subjects were given the task of building electrical wiring harnesses and each
attended one two-hour session with a small group. The sessions were led by a
trained and experienced actor portraying either a charismatic or non-charismatic
leader. Charismatic leader behavior included vision statements, statements of
confidence in subject ability, facial expressiveness, animated tone of voice, and
direct eye contact. After a few minutes of introduction during which the leader
used the set script, the subjects watched an instructional video. Subject completed
a self-efficacy measure and then had 15 minutes to build a harness. At 15 minutes,
subjects received feedback. In the internal feedback condition, subjects were
given the equipment to evaluate their own harnesses. In the external feedback
condition, technical specialists inspected the harnesses in another room and then
42


subjects received feedback. In the no feedback condition, subjects received no
feedback. After the feedback or no feedback, the procedure was repeated for a
total of four trials: self-efficacy assessment, 15 minutes to build a harness, and
feedback. Subject performance was then assessed based on harness performance
and how well the harness conformed to specification. Shea and Howell did not
find a main effect for leadership on performance, so they could not test for self-
efficacy as a mediator. However, they did find a significant interaction between
leadership and feedback on performance and some evidence that the effect of this
interaction on performance may have been mediated by self-efficacy. In the no
feedback condition, individuals with a charismatic leader out-performed
individuals with a non-charismatic leader. In order to test for self-efficacy as a
mediator, Shea and Howell added self-efficacy as a covariate in their repeated
measures MANCOVA. When controlling for self-efficacy, the interaction
between leadership and feedback on performance became non-significant,
suggesting that self-efficacy was mediating the relationship. Shea and Howell
note that a path-analysis could not be performed because of the four different trials
and sets of data. Even though Shea and Howells results did not support their
hypothesis that leader charisma would be positively related to performance, they
indicate that their findings as consistent with other studies that have found an
interaction between charismatic leadership and situational variables (e.g., Howell
43


& Frost, 1989, for group productivity norms). In other words, whether or not
charismatic leadership is superior to non-charismatic leadership may very much
depend on the situation.
In Kirkpatrick and Lockes (1996) study discussed earlier, self-efficacy
was not found to mediate the relationship between charismatic leadership and
performance. In contrast to the Shea and Howell (1999) study, Kirkpatrick and
Locke did find a main effect for leader attributes on performance, albeit a weak
one. Aspects of charismatic leadership, (vision statements and task cues or vision
implementation) were positively related to performance. Kirkpatrick and Locke
found correlations between vision statements, goals pertaining to quality, self-
efficacy pertaining to quality, and performance quality. This pattern of
correlations suggests a two-part casual sequence, so Kirkpatrick and Locke then
performed an exploratory path model. Their results suggest that leader vision
statements may be related to performance quality to the extent that the vision
statements affect goals and self-efficacy. Kirkpatrick and Locke state that this
finding should be considered preliminary until confirmed by other studies.
In summary, existing theory proposes that self-efficacy mediates the
relationship between charismatic leadership and performance (Shamir et al., 1993)
but so far there has been only limited empirical support. Shea & Howell (1999)
found that self-efficacy mediated the interaction between charismatic leadership
44


and feedback on task performance. In the absence of task feedback, followers of
charismatic leaders had higher task performance and this was mediated by self-
efficacy. Kirkpatrick and Locke (1989) did not find that self-efficacy was a
mediator, but they did find preliminary evidence that leader vision may be
associated with higher performance quality to the extent that leader vision
influences followers to set higher goals and have higher self-efficacy.
Creative Self-Efficacy as a Mediator
As discussed, empirical studies have provided some evidence to support
the theoretical propositions that self-efficacy mediates between personality and
performance and between charismatic leadership and performance. To return to
our focus on creative tasks, creative self-efficacy is thought to be a component of
motivation for creative tasks (Ford, 1996), and has been shown to predict creative
performance on the job (Tierney & Farmer, 2002). Motivation may be a mediator
between the context and creative performance (Zhou & Oldham, 2001). Taken
together, the theories and research findings reviewed suggest that creative self-
efficacy may likewise function as a mediator between context and creative
performance. Thus,
Hypothesis 4a \ The effects of openness to experience,
conscientiousness, and charismatic leadership on creative
performance will be mediated by creative self-efficacy.
45


Based on the leadership literature, leaders are believed to affect follower beliefs
about their abilities (Conger & Kanungo, 1987; Shamir et al., 1993). Individual
differences in creative ability are expected to be associated with creative self-
efficacy.
Hypothesis 4b\ Creative self-efficacy will be highest with
charismatic leadership and high openness to experience.
Hypothesis 4c\ Creative self-efficacy will be lowest with non-
charismatic leadership and high conscientiousness.
46


CHAPTER 2
METHOD
Research Setting. Participants, and Procedure
Participants in the study were employees at a manufacturing company with
two business locations in the United States. The company was small-to-medium
size, with fewer than two thousand employees worldwide. Employees located in
the United States were recruited to participate from a variety of functional areas:
for example, finance, customer service, and information technology. Employees
in human resources were excluded from the study.
Study materials were sent to 629 employees and 76 supervisors.
Supervisors were included and counted in the 629 employees, so long as they had
an assigned supervisor themselves at the time of the study. Thus, most supervisors
received both an employee packet and a supervisor packet. The 76 supervisors
received a total of 629 forms, one creativity rating form for each employee. Of the
employee questionnaires, 217 were returned or 34% of the total. Of the supervisor
forms, 467 were returned or 74% of the total. In order to use the data in the
present study, a match was needed between the employee data and the supervisor
data for that employee. There were 170 such matches, equal to 27% of the total
number of employees.
47


Of the 170 matches, three cases were excluded from analysis due to
unusual responses. In one case, there were contradictory responses give to
personality items. For example, the respondent indicated that quiet, shy, and
extraverted were all very accurate self descriptions. In two other cases, the
respondents marked all personality items with either Moderately inaccurate or
Neither accurate nor inaccurate.
Employees received a packet containing a questionnaire and cover letter.
The questionnaire included leadership, personality, and creative self-efficacy
measures, along with demographic items. The questionnaire was in the form of a
booklet comprised of two sheets of 8 ^ by 11 inch paper folded in half. The cover
letter stated that this is a research study investigating the link between leadership
style and on-the-job creativity and requested each employees participation. In
order to increase participants comfort with the questionnaires, no names were
attached to the completed questionnaires and an envelope was included to mail the
completed questionnaire to a university psychology department. Participants were
told that their information would be held in strictest confidence by the
researchers and that their responses would not be reported with any personally
identifying information. Only aggregated responses would be reported to the
company.
48


Employees who were supervisors also received a second packet with a
supervisor cover letter and creativity rating forms. The creativity rating forms
asked supervisors to complete individual ratings for each of their direct reports.
An envelope was provided to return the completed forms to a university
psychology department.
Due to the need to match responses from employees and supervisors, the
questionnaires were coded with a three-digit number unique to a group of
employees with the same supervisor followed by a two-digit number unique to
each employee in the group. On the employee questionnaires, the code appeared
on the back cover of the booklet. The creativity rating forms included the name of
the employee to be rated, with the instruction to supervisors to cut off the
employee name before returning the form. The employee code was written about
two inches below the name and remained on the completed forms. An
undergraduate psychology student generated the codes based on a list of
employees and supervisors provided by the human resources department, put
names and codes on the forms, and inserted them into envelopes.
Measures
Openness to experience. The Saucier measure was used to assess openness
to experience (Saucier, 1994). Respondents were asked how much different
adjectives describe them on a nine-point Likert scale ranging from extremely
49


inaccurate to extremely accurate. The Saucier measure consists of forty total
items, with eight items for each of the Big Five personality factors. Each scale
includes adjective markers of both the positive and negative poles of each factor.
Examples of adjectives related positively to openness are intellectual and
complex. An example of a negative adjective for openness is uncreative. The
present study had an alpha of .70 for the openness to experience scale.
The Saucier measure is a shortened version of Goldbergs adjective list
(Saucier, 1994). For each of the five scales, the shorter list has lower alphas,
higher mean inter-item correlations, and smaller correlations between the five
personality factors (Saucier, 1994). For the self-rating sample, the openness scale
had a Cronbach's alpha of .78 on the Saucier version compared to .85 on Goldberg
(Saucier, 1994). The mean inter-item correlation was .32 on the shorter Saucier
scale and .26 on the longer Goldberg scale.
Conscientiousness. The Saucier measure was also used to assess
conscientiousness (Saucier, 1994). Examples ofpositive adjectives for
conscientiousness are efficient and organized. An example of a negative
adjective is sloppy. The present study had an alpha of .78 for the
conscientiousness scale indicating good reliability.
The Saucier measure of conscientiousness was validated against an
accepted Big Five measure (Saucier, 1994). For the self-rating sample, the
50


conscientiousness scale had a Cronbach's alpha of .83 on the Saucier version
compared to .90 on Goldberg (Saucier, 1994). The mean inter-item correlation was
.38 on the shorter Saucier scale and .30 on the longer Goldberg scale.
Charismatic leadership. The charisma scale of the Multifactor Leadership
Questionnaire (Bass & Avolio, 1985,1995) was used to measure charismatic
leadership behavior. This scale has 11 items on a five-point Likert scale that
assess inspirational motivation, attributed charisma, and idealized influence (Bass
& Avolio, 1985,1995). Employees will be asked to answer the questions with
regard to their current supervisor. The items were worded to be in the present
tense, and examples include Displays a strong sense of power and purpose and
Acts in ways to build your trust. The present study found high reliability
indicated by an alpha of .92.
Creative self-efficacy. Tierney and Farmers (2002) measure of creative
self-efficacy was used. The measure consists of three items on a 7-point Likert
scale. A sample item is I have confidence in my ability to solve problems
creatively. The Tierney and Farmer creative self-efficacy scale had good
reliability with alphas of .83 and .87 in their two samples (Tierney & Farmer,
2002). At the request of Pam Tierney, an additional item was added to the original
scale for a total of four items (personal communication, November 2003). The
present study found an alpha of .82 for the four-item scale.
51


Creative behavior. Supervisory ratings were used to measure creative
behavior. Items were adapted from Scott and Bruces 1994 measure. Creative
behavior in organizations has frequently been measured with supervisory ratings
(e.g., Oldham & Cummings, 1996; George & Zhou, 2001). Supervisor ratings
have been found to be correlated with objective measures of creativity (e.g.,
Tierney, Farmer, & Graen, 1999). Tierney and her colleagues found correlations
of .29 and .28 between supervisory ratings and two different objective measures:
invention disclosure forms and research reports (1999). Scott and Bruce (1994)
found a .33 correlation between supervisor ratings on their innovative behavior
measure and an objective measure of creativity: invention disclosures.
The supervisory rating of employee creativity consist of six items on a
five-point Likert scale from very inaccurate to very accurate. Sample items
included Comes up with creative ideas related to work and Develops sufficient
plans for implementing new ideas. The inter-item reliability coefficient alpha
was .92 for the six-item scale used in the present study, indicating very high
reliability. In the Scott and Bruce (1994) sample, their six-item innovative
behavior measure had an inter-item reliability (Cronbachs alpha) of .89.
Importance of creativity. The supervisory creativity rating form also
included two items designed to measure the importance of creativity for each job
position: Creativity is important in this job position and Creativity is a valuable
52


asset in this job position. These items used the same five-point Likert scale as the
creative behavior items, with responses ranging from very inaccurate to very
accurate. These two items had very high inter-item reliability with an alpha of
.97.
Demographics. Items included the following categorical questions: gender
(male or female), age (18-30, 31-40,41-50, over 50), ethnicity (African/American,
Asian, Latino/Hispanic, Caucasian, other), level of education (Some high school,
High school diploma, Some college or technical school, 4-year degree, Graduate
degree), length of time with the company (less than 3 months, 3 to 6 months, 6
months to 1 year, 1 year to 3 years, 3 to 5 years, 5 to 10 years, more than 10 years),
length of time with current supervisor (less than 3 months, 3 to 6 months, 6
months to 1 year, 1 year to 3 years, 3 to 5 years, 5 to 10 years, more than 10 years),
and job category (Sales, Customer Service/Order Administration, Manufacturing,
Finance/InformationTechnologies/Marketing/Art, Other). The job categories
were chosen to be broad enough that large numbers of employees would fall into
each category in order to conceal the identity of those responding. If more than
one response was marked for the current level of education, the highest level was
used. Several respondents marked both high school diploma and some college
or technical school.
53


The sample was 70 percent female and 83 percent Caucasian. Thirty-one
percent worked primarily in manufacturing. Forty percent had some college or
technical school as their highest educational level. Thirty-four percent had been
with the company for more than ten years, and 40 percent had been with their
current supervisor for one to three years. See Tables 2.1 through 2.7 for the
complete demographic responses.
Table 2.1
Sample Characteristics: Job Category
Job Category Percent
Sales 13.5%
Customer Service/Order Administration 17.6%
Manufacturing 31.2%
Finance/Information T echnologies/Marketing/Art 20.0%
Other 17.1%
No response 0.6%
54


Table 2.2
Sample Characteristics: Level of Education
Education Percent
Some high school 3.5%
High school diploma 29.4%
Some college or technical school 40.0%
4-year degree 19.4%
Graduate degree 7.1%
No response 0.6%
Table 2.3
Sample Characteristics: Tenure with the Company
Tenure with company Percent
Less than 3 months 0.6%
3 to 6 months 2.9%
6 months to 1 year 6.5%
1 year to 3 years 20.0%
3 to 5 years 14.1%
5 to 10 years 21.8%
More than 10 years 34.1%
No response 0.0%
55


Table 2.4
Sample Characteristics: Length of Time with Current Supervisor
Time with current supervisor Percent
Less than 3 months 9.4%
3 to 6 months 10.6%
6 months to 1 year 13.5%
1 year to 3 years 40.0%
3 to 5 years 11.8%
5 to 10 years 9.4%
More than 10 years 4.7%
No response 0.6%
Table 2.5
Sample Characteristics: Age
Age Percent
18-30 20.6%
31-40 22.4%
41-50 30.6%
Over 50 26.5%
No response 0.0%
56


Table 2.6
Sample Characteristics: Gender
Gender Percent
Male 30.0%
Female 69.4%
No response 0.6%
Table 2.7
Sample Characteristics: Race
Race Percent
African American 5.3%
Asian 3.5%
Latino/Hispanic 3.5%
Caucasian 81.2%
Other 4.7%
No response 1.8%
57


CHAPTER 3
RESULTS
Means, standard deviations, and zero-order correlations can be found in
Table 3.1. A series of hierarchical regressions was conducted to test the
hypotheses. Demographic characteristics and the importance of creativity for the
job position were used as control variables.
Demographic responses for level of education, tenure with the company,
length of time with the same supervisor and age were in sequential categories, so
these were treated as scale variables. The categories were consolidated in order to
get approximately even numbers of responses in each category. Level of
education was coded into three values representing (1) high school diploma or less,
(2) some college or technical school, and (3) 4-year degree or more. Tenure with
the company was coded into five values representing (1) less than one year, (2)
one to three years, (3) three to five years, (4) five to ten years, and (5) more than
ten years. Length of time with the same supervisor was coded into three values:
(1) less than one year, (2) one to three years, and (3) more than three years.
Since the gender, race, and job category were categorical, they were
converted into dichotomous variables with a value of 0 or 1 in order to be used in
58


Table 3.1
Means. Standard Deviations, and Intercorrelations of Variables
Variable Mean SD Creative performance Importance of creativity Charisma Openness Conscientiousness
Creative performance 3.31 0.90 -
Importance of creativity 3.72 1.16 0.57*** -
Charisma 2.71 0.89 0.10 0.17* -
Openness 3.73 0.55 0.14* 0.02 0.14* -
Conscientiousness 4.14 0.55 0.02 -0.06 0.02 0.05 -
Creative self-efficacy 5.50 0.86 0.34*** 0.23** 0.21** 0.54*** 0.12
* p < .05
**p<.01
***p<.001


the regression analysis. Gender was coded as a single variable with a value of 1
for male and 0 for female. Due to the small number of responses in non-white race
categories, race was coded as a single variable with a value of 1 for white and 0 for
non-white. Job category was dummy-coded in order to create sets of variables
with the value of zero or one. There were five possible responses for job category.
Job category was converted into four new dichotomous variables representing
Sales, Customer Service/Order Administration, Manufacturing, and
Finance/Information Technologies/Marketing/Art. For the job category
selected, the value would be one. For all other job categories, the value would be
zero. In order to avoid adding redundant information and multicolinearity, one
less variable than the number of possible responses was created. Therefore, no
variable was created for the fifth choice Other for job category. If no response
was given, the set of dummy variables would all reflect a missing value for that
respondent.
Each of the dependent variables (creative performance and creative self-
efficacy) was regressed on the control variables in order to determine which
control variables were related to the dependent variables. Control variables with a
B significance less than .20 were retained for the regression analyses testing the
hypotheses. The relatively high cut-off of .20 was chosen because there was some
correlation between the control variables, and the B significance values decreased
60


as variables were removed. This way, only control variables that were somewhat
related to the dependent variable were retained. For creative performance, the
control variables retained were tenure with company, length of time with the same
supervisor, race, and importance of creativity for the job position. For creative
self-efficacy, the control variables retained were job category, education, gender,
and importance of creativity for the job position.
Mean values were used for charismatic leadership, conscientiousness,
openness, creative self-efficacy, and creative performance scales. Therefore, as
long as at least one response was received for each scale, it would not be
considered missing. Missing values were then excluded listwise in each of the
regression analyses. Thus, in order to be included, at least one response was
required for charismatic leadership, conscientiousness or openness, creative self-
efficacy, and creative performance but responses to all demographic questions
were required.
When testing for interactions between continuous variables, centering the
variables on the mean before computing an interaction term reduces correlation
with the component variables and can facilitate interpretation (Aiken & West,
1991). Thus, in order to test for interactions between charismatic leadership and
personality factors, each of the variables was converted to new centered variable
with a mean equal to zero by subtracting the mean from each individual value.
61


Then, the centered personality and leadership variables were multiplied together to
create interaction terms: charisma X openness and charisma X conscientiousness.
Personality. Charismatic Leadership
and Creative Performance
Hypothesis la: Individuals high on openness to experience will
have higher levels of creative performance.
A hierarchical regression analysis was conducted with controls entered in
step 1 and openness to experience entered alone in step 2. As predicted, there was
a main effect of openness to experience on creative performance (AR2 = .02, p <
.05, one-tailed).
Hypothesis lb: Individuals high on conscientiousness will have
lower levels of creative performance.
A hierarchical regression analysis was conducted with controls entered in
step 1 and openness to experience entered alone in step 2. Contrary to predictions,
no main effect was found for conscientiousness on creative performance (AR =
.00, n.s.).
Hypothesis 2: Charismatic leadership will be associated with
greater creative performance.
A hierarchical regression analysis was conducted with controls entered in
step 1 and charismatic leadership entered alone in step 2. No main effect was
found for charismatic leadership on creative performance (AR2 = .00, n.s.).
62


Table 3.2
Regression of Creative Performance on
Openness and Charismatic Leadership
Hierarchical Step Independent variable Standardized regression coefficient (P) R2
1 Demographics Importance of creativity Total R2 AR2 0.51 ** 0.36 ** 0.36 **
2 Openness to experience Total R2 AR2 0.12 1 0.38 ** 0.02
3 Charismatic leadership Total R2 AR2 0.01 0.38 ** 0.02
3 Openness X Charisma Total R2 AR2 1 O U> * 0.40 ** 0.02 *
Dependent variable: Creative performance (supervisor rating)
N= 162
1 p < .05, one tailed
* p < .05
** p < .01
63


Interaction between Leadership and Personality
Hypothesis 3a\ There will be a two-way interaction of charismatic
leadership and high openness to experience on creative
performance, such that the highest creative performance will be
achieved when both charismatic leadership and openness to
experience are high.
As expected, an interaction was found between openness to experience and
charisma on creative performance (AR2 = .02, p < .05). However, the nature of the
interaction was not as expected. Individuals who were high in openness to
experience had the highest creativity regardless of whether their supervisor was
charismatic, whereas individuals who were low in openness to experience had
higher creativity with a charismatic supervisor than a non-charismatic supervisor
(P = -.13, t = -1.98, p < .05). See Figure 3.1 for an interaction graph.
Hypothesis 3b\ Conscientiousness and charismatic
leadership will interact to affect creative performance such that
individuals who are high on conscientiousness will exhibit the
lowest level of creative performance when their supervisor in low
on charismatic leadership than for any other combination of these
variables.
Contrary to expectation, no interaction was found between charismatic
leadership and conscientiousness on creative performance (AR2 = .00, n.s.).
64


Figure 3.1
Interaction between Openness to Experience
and Charismatic Leadership on Creative Performance
Openness to Experience (Centered)
Creative Self-Efficacy
Hypothesis 4a\ The effects of openness to experience,
conscientiousness, and charismatic leadership on creative
performance will be mediated by creative self-efficacy.
The procedures delineated in Baron and Kenny (1986) were used to test
creative self-efficacy as a mediator. First, the mediator should be regressed on the
independent variable in order to determine if the independent variable affects the
65


mediator. In separate regressions for each of the independent variables with
control variables entered in the first step, creative self-efficacy was predicted by
openness to experience (AR = .23, p < .01), by conscientiousness (AR = .03, p <
.05), and by charismatic leadership (AR2 = .03, p < .05). Thus, the first condition
is satisfied.
Second, the dependent variable should be regressed on the independent
variable in order to establish a relationship between the dependent variable and the
independent variable (Baron & Kenny, 1986). As was seen above, creative job
performance was related to openness to experience (AR2 = .02, p < .05, one-tailed)
but not to charismatic leadership (AR = .00, n.s.) or conscientiousness (AR = .00,
n.s.). Thus Baron and Kennys second condition for mediation was satisfied for
openness to experience but not for charismatic leadership or conscientiousness.
Third, the dependent variable should be regressed on both the mediator and
the independent variable in order to determine the mediator is related to the
dependent variable and if controlling for the mediator reduces the relationship
between the dependent variable and the independent variable. Because openness
to experience is the only independent variable to have passed the first two criteria
of mediation, only openness to experience will be examined further. In a
hierarchical regression with creative performance as the dependent variable,
demographic variables and importance of creativity were entered in step 1,
66


creative self-efficacy was entered in step 2, and openness to experience was
entered in step 3. See Table 3.3. The relationship between openness to experience
and creative performance was no longer significant after controlling for creative
self-efficacy (AR2 = .00, n.s.). Thus, creative self-efficacy mediates the
relationship between openness to experience and creative performance according
to the Baron and Kenny (1986) criteria.
Because an interaction was found between openness to experience and
charisma on creative performance, creative self-efficacy might also mediate this
relationship. A regression was conducted to determine if a relationship existed
between creative self-efficacy and the interaction between openness and charisma
after controlling for demographics, importance of creativity, openness, and
charisma in a hierarchical regression. Creative self-efficacy was not predicted by
the interaction between charisma and openness (AR2 = .00, n.s.). Therefore,
creative self-efficacy does not mediate between this interaction and creative
performance.
67


Table 3.3
Regression of Creative Performance on Openness
Controlling for Creative Self-Efficacy
Hierarchical Step Independent variable Standardized regression coefficient (3) R2
1 Demographics Importance of creativity Total R2 0.51 ** 0.37 **
AR2 0.37 **
2 Creative self-efficacy Total R2 0.20 ** 0.41 **
AR2 0.04 **
3 Openness to experience Total R2 0.01 0.41 **
AR2 0.00
Dependent variable: Creative performance (supervisor rating)
N= 159
* p <.05
** p <.01
68


Hypothesis 4b: Creative self-efficacy will be highest with
charismatic leadership and high openness to experience.
Contrary to expectations, no interaction was found between charisma and
openness on creative self-efficacy (AR = .00, n.s.). However, as stated above,
there was a main effect for openness to experience on creative self-efficacy (AR
y
.23, p < .01) and charismatic leadership on creative self-efficacy (AR = .03, p <
.05).
Hypothesis 4c\ Creative self-efficacy will be lowest with non-
charismatic leadership and high conscientiousness.
Contrary to predictions, no interaction was found between charisma and
conscientiousness on creative self-efficacy (AR = .01, p = .11). However, as
stated above, there was a main effect for conscientiousness on creative self-
efficacy (AR2 = .03, p < .05) and charismatic leadership on self-efficacy (AR2 =
.03, p < .05). Higher conscientiousness was associated with higher creative self-
efficacy (P = .16, p < .05).
69


CHAPTER 4
DISCUSSION
The aim of the present study was to propose and test an interactionist
model of creativity. According to the hypothesized model, creative performance
would be a function of both individual personality factors and leadership style. In
particular, openness to experience, conscientious, and charismatic leadership were
tested as predictors of creative performance in an organizational setting. Creative
self-efficacy was proposed and tested as a mediator of the hypothesized
relationship between personality and leadership on creative performance.
The proposed model was partially supported. Employee personality
appears to influence creativity in the workplace. In particular, employees with
higher openness to experience were judged to have higher levels of creative
performance by their immediate supervisors. This relationship between openness
to experience and creative performance was found to be mediated by creative self-
efficacy.
The control variables (demographics and importance of creativity for the
job position) explained 36 percent of the variance in creative performance.
Openness to experience was able to predict creative performance beyond that
predicted by such factors as education and tenure with the company. Because the
70


sample represented a very wide variety of jobs from manufacturing to art,
controlling for the importance of creativity adds to the relevance of the study. In
fact, the results showed that valuing creativity influenced creative performance.
The organization in this study was not a typical R & D organization and many
employees held positions not typically associated with creative performance ; the
results suggest that when creativity is valued in an organization this can lead to
higher levels of creative performance.. Including a wide range of jobs in this
study increases the generalizability of these findings. These findings suggest that
creative performance exists in different types of organizations and jobs and are not
just limited to positions typically associated with creative performance.
No relationship was found between conscientiousness and creative
performance. Perhaps, this was because the environment at this organization
supported creativity. George and Zhou (2001) found that conscientiousness was
only negatively related to creative performance when supervisors were controlling
and coworkers were unhelpful. The results of the present study combined with
previous research suggest that conscientiousness may not be detrimental to
creative performance under certain organizational conditions. Another possible
explanation for the lack of relationship between conscientiousness and creativity is
the halo effect. Conscientious employees may be viewed in such generally
71


favorable light that supervisors give them higher ratings for creativity because they
have performed well on tasks not involving creativity.
Contrary to expectations, conscientiousness was positively related to
creative self-efficacy. One reasons for this is that highly conscientious individuals
tend to be highly motivated to perform and have higher general self-efficacy (e.g.,
Martocchio & Judge, 1997). Furthermore, conscientiousness individuals are more
likely to engage in self-deception (Martocchio & Judge, 1997). Thus,
conscientious individuals may have an inflated view of their own creative abilities.
Perhaps under some circumstances, the higher self-efficacy in conscientious
individuals counteracts any negative effect from conscientiousness on creative
performance.
No main effect relationship was found between charismatic leadership and
creative performance. This is surprising given that previous research has found a
positive relationships between transformational leadership and creative
performance (Bono & Judge, 2003). One possible reason for this is that
transformational leaders tend to stimulate their followers to think outside the box
and to take risks in their decisions-making. This behavior is not included in most
formal definitions of charismatic leadership; charismatic leaders through their
vision, inspire their followers to go beyond expected performance. In many ways,
72


charismatic leaders are directive and might increase conformity from their
followers because they have a vision that they wish their followers to pursue.
Charismatic leaders may have the expected effect on creative performance
for less creative employees. There was an interaction between openness to
experience and charismatic leadership on creative performance such that
employees with high openness to experience had high creative performance
regardless of charismatic leadership while employees low on openness to
experience had higher levels of creative performance with a charismatic leaders
than a non-charismatic leader. Because the only measure of creative performance
in the present study was supervisor (leader) ratings, one possible explanation is
that charismatic leaders have higher standards and are harsher judges of creativity
with highly creative employees. If charismatic leaders have more confidence in
their own creative ability, that could also contribute to tougher standards.
Narcissism is considered a possible dark side of charisma (Deluga, 1998).
Possibly, some charismatic leaders set tougher standards with creative
subordinates because they have a need to perceive themselves as superior.
Additional analyses should be done to determine if charismatic leaders have higher
creative self-efficacy themselves which could provide some indirect evidence of
higher standards based on their own self-view. If charismatic leaders do indeed
have tougher standards for evaluating employee creativity, charismatic leadership
73


could still be related to higher creative performance in employees with high
openness to experience even though no relationship was found here. Future
research should be conducted utilizing additional measures of creative job
performance such as objective measures or peer evaluation to test that possibility.
Another possible explanation for the finding that charismatic leadership
only influences creative performance when employees are low on openness to
experience is that employees high on openness to experience are already so
inclined to be creative that some environmental factors do not have much impact.
Employees who do not have personality factors related to creativity may benefit
more from the effects of charismatic leaders.
While no main effect was found between charismatic leadership and
creative performance, creative self-efficacy was predicted by charismatic
leadership and creative self-efficacy was positively related to creative job
performance. This provides some indirect evidence that charismatic leadership
may be related to creative performance. These findings would be consistent with
previous studies that charismatic leaders boost their followers self-efficacy and
this leads to higher levels of creative performance (Howell & Frost, 1987;
Kirkpatrick & Locke, 1996). Although these studies did not focus on creative
performance, they did find that charismatic leaders tended to increase their
followers self-efficacy.
74


Limitations
There are several limitations in this study. First, the reasons for some of the
findings might be due to limitations in the measure that were used. Because both
creative self-efficacy and personality were based on self-report, there is the
potential for common method variance. A confirmatory factory analysis should be
conducted to determine if the factors are distinguishable from each other. In
particular, some of the openness to experience items, especially creative and
uncreative may not be distinguishable from the creative self-efficacy items.
From a theoretical standpoint, self-efficacy is a proximal construct and acts to
motivate, whereas personality is considered a distal construct. If the items do not
load onto separate factors, it would call into question whether the constructs in this
present study were actually measuring what they purported to measure. In order
to make a strong case that creative self-efficacy mediates the relationship between
openness to experience and creative performance, openness to experience and
creative self-efficacy need to be distinguishable from each other.
The creative self-efficacy measure developed by Tierney and Farmer,
(2002) was only three items long since additional items did not add to the
reliability. A fourth item was later added (Tierney, personal communication,
75


November 2003). In spite of high reliability, it is possible that four items were not
enough to capture all of the variability that exists in creative self-efficacy.
While the Saucier (1994) personality measure has acceptable reliability, it
is also relatively new and may not be as good as longer personality measures such
as the Adjective Checklist. With fewer items, it may not capture the variance in
personality was well as a longer measure or have the same reliability. A
confirmatory factor analysis should be done in order to make sure that the Saucier
measure loads onto the Big Five factors.
Another limitation is the reliance on supervisors ratings to measure
creative performance. The creative performance measure did not include a
definition of creativity on the measure itself. It is possible that some supervisors
did not have a well-developed concept of creativity. For example, some
supervisors may view only tasks such as advertising, marketing, or art as involving
creativity. The cover letter did contain a longer explanation of creativity, and
stated that creativity could be mean to see a new way to solve a difficult problem
or to have an idea for improving an existing process. However, many
respondents may not have read the cover letter carefully, and it is likely that some
supervisors could have benefited from having a specific definition of creativity on
the employee creativity rating form.
76


While the present study did use data from two sources (employees and
supervisors) and past researchers have found the supervisor rating corresponded
with other performance measures, an objective measure of creativity would have
strengthened the present study. Because this study included employees in a wide
variety of job categories, there was no single objective standard for creativity, such
as patents or published articles.
While it was seen as a strength to include such a wide variety of different
job since creative performance could potentially come from anyone in an
organization, it is a possible explanation for the lack of a main effect relationship
between leadership and creative performance. If the sample had been limited to
job categories where creativity was seen as central to success, such as in
advertising or research and development, there might have been different findings.
In the organization where the study was conducted, the sample was too small to
group these jobs separately in a meaningful way. For example, there were less
than a dozen people in the arts department. No difference in creative performance
was found between the broad categories used in the study. Further research is
recommended to test the proposed model in organizations where large numbers of
people are doing creative work, such as an advertising agency or scientific
research firm. This would also enable an objective measure of creativity to be
used.
77


Directions for Future Research
Future research should be done using additional measures of creative
performance to explore the possibility that charismatic leadership does have a
positive impact on creative performance for employees high in openness to
experience in spite of the null findings in the present study. Because organizations
often rely very heavily on supervisor ratings for performance evaluation and
decisions about promotions, pay increases, and layoffs, it may be worth
investigating whether charismatic leaders are harsher judges of more talented
employees in order to better understand bias in personnel appraisal. If charismatic
leaders are harsher evaluators or non-charismatic leaders are more lenient
evaluators, it could have implications for fairness in evaluating employee
performance. Perhaps less popular supervisors use favorable evaluations as a way
to keep employees satisfied whereas charismatic leaders have other means to
motivate and retain employees.
Implications for Organizations
The findings suggest that organizations may want to test for openness to
experience when selecting employees for positions where creativity is important.
Validation studies should be done within the organization to confirm that pre-hire
personality tests are related to subsequent creative performance on the job.
78


Alternatively, validation studies could be conducted with existing employees.
Organizations may want to select a personality test that is less transparent than an
adjective checklist if using for selection. In the present study, employees were told
that the results would have no impact on any organizational decisions and so
would have less reason to distort their responses in order to appear more desirable
as an applicant.
Secondly, the findings suggest that charismatic leaders may help an
organization get creative contributions from a broader number of its employees.
Employees less inclined towards creative because of their personality may be more
likely to be creative with a charismatic leader. As organizations strive for
innovation, it would seem to be an advantage to be maximizing creativity from all
employees. Employees may have unique contributions to make, especially if few
employees are doing the same tasks. Organizations may want to select supervisors
with a charismatic leadership style or train supervisors to be more charismatic
leaders (Towler, 2003).
The findings in the present study also suggest that when organizations
value creativity, this leads to higher levels of creative performance. This suggests
that organizations should develop and encourage a climate where creativity is
valued regardless of the types of jobs in the organization. For example, this
research suggests even when employees have mundane jobs such as clerical work,
79


a climate of creativity can enable them to think outside the box and to be creative
in every-day aspects of their work.
80


APPENDIX A
Personality Measure
Please use the following list of common human traits to describe yourself as accurately as
possible. Describe yourself as you see yourself at the present time, not as you wish to be
in the future. Describe yourself as you are generally or typically, as compared with other
persons you know of the same sex and of roughly the same age.
Very Inaccurate Moderately Inaccurate Neither Accurate nor Inaccurate Moderately Accurate Very Accurate
1 2 3 4 5
1. Jealous
2. Extraverted
3. Energetic
4. Bold
5. Temperamental
6. Unenvious
7. Unintellectual
8. Practical
9. Complex
10. Organized
11. Unsympathetic
12. Cold
13. Disorganized
14. Envious
15. Imaginative
16. Kind
17. Warm
18. Cooperative
19. Talkative
20. Intellectual
81


21. Systematic
22. Relaxed
23. Efficient
24. Rude
25. Creative
26. Deep
27. Sympathetic
28. Withdrawn
29. Bashful
30. Harsh
31. Shy
32. Moody
33. Fretful
34. Sloppy
35. Uncreative
36. Inefficient
37. Touchy
38. Careless
39. Quiet
40. Philosophical


APPENDIX B
Charismatic Leadership Measure
Not at all Once in a while Sometimes Fairly Often Almost Always
0 1 2 3 4
Please answer the following questions with regard to your current supervisor:
______Talks about most important values and beliefs
______Talks optimistically about the future
______Instills pride in being associated with him/her
______Talks enthusiastically about what needs to be accomplished
______Specifies the importance of having a strong sense of purpose
______Displays a sense of power and purpose
______Articulates a compelling vision of the future
______Emphasizes the importance of having a collective sense of mission
______Goes beyond his/her self-interest for the good of the group
______Acts in ways to build your trust
______Considers the moral and ethical consequences of his/her decisions
83


APPENDIX C
Creative Self-Efficacy Measure
Strongly
disagree
1
Disagree
2
Somewhat
disagree
3
Neither
agree nor
disagree
4
Somewhat
agree
5
Agree
6
Strongly
agree
7
<
->
Please use the above scale to respond to the following items:
______ I feel that I am good at generating novel ideas.
______ I have confidence in my ability to solve problems creatively.
______ I have a knack for further developing the ideas of others.
______ I am good at finding creative ways to solve problems.
84


APPENDIX D
Demographic Questions
Please select the job category that best fits:
____Sales
____Customer Service/Order Administration
____Manufacturing
____Finance/Information Technologies/Marketing/Art
____Other
Please indicate your current level of education:
____Some high school
____High school diploma
____Some college or technical school
____4-year degree
____Graduate degree
How long have you been with this company?
____Less than 3 months
____3 to 6 months
____6 months to 1 year
____1 year to 3 years
____3 to 5 years
____5 to 10 years
____More than 10 years
How long have you been with your current supervisor?
____Less than 3 months
____3 to 6 months
____6 months to 1 year
____1 year to 3 years
____3 to 5 years
____5 to 10 years
____More than 10 years
Please indicate your age:
____18-30 ____31-40 ___41-50 ____Over50
____Male_____Female
____African American
____Asian
____Latino/Hispanic
____Caucasian
other
85


APPENDIX E
Supervisor Creativity Rating Form
Direct report name:
(Please detach here and then return the bottom portion only.)
Code:
For the direct report named above, please answer the following questions using the scale:
Neither
Very Moderately Accurate Moderately Very
Inaccurate Inaccurate nor Inaccurate Accurate Accurate
| (1) | (2) j (3) ___ | (4) | (5)
Comes up with creative ideas related to work.
Develops sufficient plans for implementing new ideas.
Solves problems that have been difficult for others.
Tries out new ideas on the job.
Sees opportunities for new products or processes.
Identifies ways to improve the current products or processes.
Creativity is important in this job position.
Creativity is a valuable asset in this job position.
86


APPENDIX F
Cover Letter to Employees
University of Colorado at Denver
Psychology Department
Mail Box 173
P.O. Box 173364
Denver, CO 80217-3364
Dear Employee:
March 4, 2004
National Pen Corporation has chosen to participate in a research study investigating the link
between leadership style and on-the-job creativity. Creativity can be manifest in a variety of ways.
For example, creativity might help you to draw a design for a pen, see a new way to solve a
difficult problem, or have an idea for inproving an existing process. Effective leaders can help
enhance the creative process, enabling National Pen to obtain a competitive advantage in the
marketplace.
Your participation is entirely voluntary, and we hope you will be part of this exciting research. In
order that the results will truly represent National Pen Corporation, it is important that all
questionnaires be completed and returned. The University of Colorado at Denver requires that
participants in research be respected and protected. The following information is provided to help
you decide whether you wish to participate in the present study. You will be asked to assess your
supervisor on his/her leadership behaviors, to assess your own personality, and to assess your level
of confidence regarding the creative process. Your supervisor will be asked to assess your
creativity in accomplishing work tasks. All questionnaires will take only about 15 minutes to
complete. Some respondents may experience minimal psychological discomfort.
All information you provide on this questionnaire will be held in strictest confidence. We
fully realize that if this confidentiality were to be breached, there could be social or economic
consequences. Therefore, we have taken the following steps to preserve your confidentiality and to
encourage your honest and forthright responses: Your responses will not be reported with any
personally identifying information. No individual data will be reported; only aggregated
(summarized) data will be reported. None of the information from this study will be placed in your
records or otherwise used for or against your future advancement. Your participation is for
research purposes only. Codes were placed on the questionnaires in order to be able to pair your
responses with the responses of other employees with the same supervisor and with your
supervisors responses. In order to conceal your identity, the coding was done by someone with no
other involvement in either this research or National Pen Corporation. We (Dr. Annette Towler
and Susan Strickland) will receive and process the completed questionnaires, and we will not know
87


which name is associated with each code. Your name will never be attached to the completed
questionnaire.
By participating in this research, you may help leaders at your company become more effective at
developing and fostering employee creativity. As a token of our appreciation, five respondents will
be chosen at random to receive mystery gifts. One of these mystery gifts will be a $50 gift
certificate.
We would be most happy to answer any questions that you might have concerning this research.
Please contact Dr. Annette Towler at the University (e-mail: atowler@carbon.cudenver.edu) or call
Susan at 303-860-7756. If you have any questions concerning your rights as a research subject you
can contact the CU-Denver Office of Academic Affairs, 14* Floor Lawrence Street Bldg., 303-
556-4060. Please return the completed questionnaires by March 18.2004 in the envelope
provided.
Sincerely,
Annette Towler, Ph.D.
Assistant Professor in Psychology
University of Colorado at Denver
Susan Strickland
Graduate student
University of Colorado at Denver
88


APPENDIX G
Cover Letter to Supervisors
University of Colorado at Denver
Psychology Department
Mail Box 173
P.O. Box 173364
Denver, CO 80217-3364
Dear Supervisor:
March 4,2004
National Pen Corporation has chosen to participate in a research study investigating the link
between leadership style and on-the-job creativity. Creativity can be manifest in a variety of ways.
For example, creativity might help you to draw a design for a pen, see a new way to solve a
difficult problem, or have an idea for improving an existing process. Effective leaders can help
enhance the creative process, enabling National Pen to obtain a competitive advantage in the
marketplace.
Your participation is entirely voluntary, and we hope you will be part of this exciting research. In
order that the results will truly represent National Pen Corporation, it is important that all
questionnaires be completed and returned. The University of Colorado at Denver requires that
participants in research be respected and protected. The following information is provided to help
you decide whether you wish to participate in the present study. You will be asked to assess the
work-related creativity of each employee you supervise. Each of the employees you supervise will
be asked to assess your leadership behaviors. (In a separate letter, you may also be asked to assess
the leadership behaviors of your supervisor.) All questionnaires will take only about 15 minutes to
complete. Some respondents may experience minimal psychological discomfort.
All information you provide on this questionnaire will be held in strictest confidence. We
fully realize that if this confidentiality were to be breached, there could be social or economic
consequences. Therefore, we have taken the following steps to preserve your confidentiality and to
encourage your honest and forthright responses: Your responses will not be reported with any
personally identifying information. No individual data will be reported; only aggregated
(summarized) data will be reported. None of the information from this study will be placed in your
records or otherwise used for or against your future advancement. Your participation is for
research purposes only. Codes were placed on the questionnaires in order to be able to pair your
responses with the responses of employees you supervise. Names were placed on the employee
creativity questionnaires so that you would know which employee to rate on each one. Please
remove the names before returning the completed forms to us. In order to conceal your identity
and the identity of those you supervise, the coding was done by someone with no other
involvement in either this research or National Pen Corporation. We (Dr. Annette Towler and
Susan Strickland) will receive and process the completed questionnaires, and we will not know
which name is associated with each code. No names will ever be attached to the completed
questionnaires.
89


By participating in this research, you may help leaders at your company become more effective at
developing and fostering employee creativity. As a token of our appreciation, five respondents will
be chosen at random to receive mystery gifts. One of these mystery gifts will be a $50 gift
certificate.
We would be most happy to answer any questions that you might have concerning this research.
Please contact Dr. Annette Towler at the University (e-mail: atowler@carbon.cudenver.edut or call
Susan at 303-860-7756. If you have any questions concerning your rights as a research subject you
can contact the CU-Denver Office of Academic Affairs, 14th Floor Lawrence Street Bldg., 303-
556-4060. Please return the completed questionnaires by March 18.2004 in the envelope
provided.
Sincerely,
Annette Towler, Ph.D.
Assistant Professor in Psychology
University of Colorado at Denver
Susan Strickland
Graduate student
University of Colorado at Denver
90


REFERENCES
Abdel-Halim, A. A. (1981). Personality and task moderators of subordinate
responses to perceived leader behavior. Human Relations, 34, 73-88.
Ackerman, P. L., & Heggestad, E. D. (1997). Intelligence, personality, and
interests: Evidence for overlapping traits. Psychological Bulletin, 121,
219-245.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting
interactions. Newbury Park, CA: Sage.
Amabile, T. M. (1979). Effects of external evaluation on artistic creativity.
Journal of Personality and Social Psychology, 37, 221-233.
Amabile, T. M. (1983). The social psychology of creativity. New York: Springer-
Verlag.
Amabile, T. M. (1988). A model of creativity and innovation in organizations. In
B. M. Staw & L. L. Cummings, (Eds.), Research in organizational
behavior, (10), 123-167. Greenwich, CT: JAI Press.
Amabile, T. M., Conti, R., Coon, H., Lazenby, J., & Herron, M. (1996). Assessing
the work environment for creativity. Academy of Management Journal, 39,
1154-1184.
Amabile, T. M., DeJong, W., & Lepper, M. (1976). Effects of externally imposed
deadlines on subsequent intrinsic motivation. Journal of Personality and
Social Psychology, 34, 92-98.
Amabile, T. M., & Gitomer, J. (1982). Childrens artistic creativity: Effects of
choice and task materials. Unpublished manuscript, Brandeis University.
Amabile, T. M., Goldberg, N., & Capotosto, D. (1982). Effects of reward and
choice on adults artistic creativity. Unpublished manuscript, Brandeis
University.
91