Leadership Self-Efficacy and Managers' Motivation For Leading Change
Leadership Self-Efficacy and Managers' Motivation For Leading Change
Leadership Self-Efficacy and Managers' Motivation For Leading Change
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
Wiley is collaborating with JSTOR to digitize, preserve and extend access to Journal of Organizational Behavior.
http://www.jstor.org
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
DOI: 10.1002/job.137
Summary This study develops and tests a leadership model that focuses on managers' motivation for
(1986) social cognitive theory, the primary hypothesis is that high LSE managers will be seen
by direct reports as engaging in more leadership attempts. Relationships are also proposed
between LSE and several factors that are expected to influence this confidence judgment.
The model was tested through surveys distributed to managers (n = 150) and their direct
reports (n = 415) in a real estate management company and an industrial chemicals firm.
Positive relationships (p < 0.05) were found between the first two dimensions of LSE and
was discovered for the LSE/overcoming obstacles dimension (p < 0.05). Several positive
relationships were found between LSE dimensions and proposed antecedents, including
self-esteem (p < 0.05), subordinates' performance abilities (p < 0.05), and managers' job
autonomy (p < 0.05). Copyright ? 2002 John Wiley & Sons, Ltd.
Introduction
Recently, writers have discussed how the role of managers in American industry has changed. Rather
than emphasizing balance, stability, and control, managers are being asked to rapidly respond to chan-
ging environments by actively seeking out new opportunities and leading their followers forward to take
advantage of them. Managers need to continually assess how things could be done better, get employees
to share their change goals, and work together with them to achieve those goals (e.g., Capowski, 1994;
* Correspondence to: Laura L. Paglis, School of Business Administration, University of Evansville, 1800 Lincoln Ave.,
E-mail: lp39@evansville.edu
tThis article is based on the first author's doctoral dissertation and was supported in part by a grant from the Purdue Research
Foundation.
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
Harari, 1995; Kotter, 1990). Consequently, today's organizations need to be able to identify individuals
who are motivated to confront organizational inertia and spearhead continuous improvement efforts.
Our position is that perceptions of leadership self-efficacy (LSE) are an important source of a manager's
motivation for taking on the difficult task of attempting change initiatives at work. Accordingly, a model
of LSE, its antecedents, and its consequences is presented and empirically tested.
the behaviors that are required to produce desired outcomes (Bandura, 1977; Gist & Mitchell, 1992).
behavior. People get involved in activities that they judge themselves capable of handling; once
engaged, their efficacy beliefs influence how much effort they devote to the task and how long they
persist in the face of obstacles (Bandura, 1977, 1986). Extending this research to the leadership area,
managers who favorably judge their ability to lead change are expected to be seen by others at work as
initiating more change efforts and persisting longer at realizing their change goals even in the face of
obstacles. In sum, this research explores one potential answer to the perplexing question, why do some
managers seek out opportunities for leading change within their units, while other managers are con-
The consensus among leadership scholars seems to be that there is no agreed upon definition of leader-
ship (Bass, 1990; House & Podsakoff, 1994; Stogdill, 1974; Yukl, 1998). However, common themes
can be identified among the many different definitions that have been proposed. First, leadership is
(e.g., Hemphill & Coons, 1957; House & Baetz, 1979; Rauch & Behling, 1984; Yukl, 1998). Fol-
lowers' voluntary acceptance of this influence, rather than forced compliance, is implicit in many
authors' views of leadership (e.g., House & Baetz, 1979; Jacobs & Jaques, 1990; Jago, 1982; Kotter,
objective and corresponding strategies (Kotter, 1990; Krech & Crutchfield, 1948; Yukl & Van Fleet,
1990). If common threads are pulled together, what remains is a definition along these lines: 'leader-
ship is a process of identifying a group goal and corresponding strategy, and influencing others to
direct their efforts voluntarily in pursuit of it.' This definition, however, seems to lack the dynamic
quality that is typically associated with great leadership, and so to refine our definition further we
First, both leaders and managers carry out position responsibilities and use their delegated authority,
but only leaders are said to influence followers' commitment (Yukl, 1989; Yukl & Van Fleet, 1990).
This commitment engenders a sense of energy and enthusiasm among employees, and, over time, their
satisfaction becomes tied to the accomplishment of group goals (House & Podsakoff, 1994;
Nanus, 1992). A second distinction between leaders and managers is one of degree of change. While
management has traditionally been concerned with the functions of planning and budgeting,
organizing and staffing, and controlling and problem solving, leaders are agents of change within
Copyright ? 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
organizations (Kotter, 1990; Nanus, 1992). They diagnose their work group's strengths, deficiencies,
and opportunities, and determine the changes needed in order for the unit to survive and excel. Leaders
motivate others to commit to these changes, while helping them overcome obstacles encountered along
the way to realizing them. Thus, for this study we developed a definition of leadership that specifically
Leadership is the process of diagnosing where the work group is now and where it needs to be in the
future, and formulating a strategy for getting there. Leadership also involves implementing change
through developing a base of influence with followers, motivating them to commit to and work hard
in pursuit of change goals, and working with them to overcome obstacles to change.
The definition of LSE follows from the three general leadership tasks implicit above:
LSE is a person's judgment that he or she can successfully exert leadership by setting a direction for
the work group, building relationships with followers in order to gain their commitment to change
The model guiding this research is shown in Figure 1. At its centre is the LSE construct, representing a
manager's self-perceived capabilities for the general leadership tasks of direction-setting, gaining
I-----------------
Perceived
* trait self-esteem
Subordinate antecedents
gaining _ Leadership
commitment attempts
Superior antecedents
* leadership modeling
* coaching behavior
Organizational antecedents
- - - --------------I
Organizational
commitment
Copyright ( 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
followers' commitment, and overcoming obstacles. Four categories of antecedents are cast as
important influences on managers' LSE judgments. The central hypothesis of this study is the pro-
posed positive relationship between LSE and leadership attempts. Managers' organizational
commitment and perceptions of crisis are included as potential moderators of the relationship between
LSE Antecedents
The proposed LSE antecedents in Figure 1 reflect fairly stable individual differences, such as trait
self-esteem, as well as more proximal influences in the work environment, such as superior and sub-
to propose specific links between antecedents and LSE dimensions until its dimensionality can be
tested. Thus, the following hypotheses are set forth in the form of relationships between antecedents
Individual antecedents
The first category of antecedents consists of individual characteristics that may influence a manager's
perceived efficacy for leading change. Personal mastery experiences are believed to be one of the most
influential sources of efficacy information (Bandura, 1986). A pattern of successes strengthens an indi-
vidual's belief in his or her capabilities, especially when success is achieved by overcoming obstacles
through persistent effort (Wood & Bandura, 1989). Thus, one would expect that the manager with more
The second factor expected to influence leadership efficacy perceptions is locus of control (Rotter,
1966). Individuals with an internal locus of control believe that their actions or personal characteristics
are prominent determinants of their experiences, whereas those with an external tendency feel out-
comes are primarily determined by outside forces such as luck, other people, or other circumstances
(Lefcourt, 1991). Possessing an internal locus of control is expected to positively influence managers'
judgments about whether or not they can create meaningful change in their units. Believing that goals
are achievable largely through one's efforts, rather than a product of chance or circumstance, should
enhance managers' perceptions of their ability to lead change. Support for this idea includes research
linking internal locus of control with self-confidence about academic and job performance (Gurin,
The role of trait self-esteem in a self-efficacy model is somewhat controversial. Bandura (1986)
states 'judgments of self-worth [i.e., self-esteem] and of self-capability have no uniform relation'
(p. 410). On the other hand, Brockner (1988) notes that while it is possible for a low self-esteem indi-
vidual to have high self-efficacy for a given task, in general, those higher in self-esteem have more
positive beliefs about their abilities than do people with low self-esteem. Likewise, Gist and Mitchell
(1992) consider self-esteem as one determinant of self-efficacy. They note that people with low
self-esteem are more likely to suffer anxiety in performance situations and tend to excessively focus
on their perceived inadequacies. A low self-esteem manager, then, may feel more anxious and
self-critical about his chances for success when confronting leadership opportunities, compared to
Copyright @ 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
Hypothesis 1(a): Successful experience in leadership roles will be positively related to LSE.
Subordinate antecedents
The second category of LSE antecedents acknowledges the critical role others play in the leadership
process. Whereas routine managerial tasks can sometimes be accomplished via autocratic methods,
A recent line of research has examined the prevalence and effects of organizational cynicism among
employees (Dean, Brandes, & Dharwadkar, 1998; Reichers, Wanous, & Austin, 1997; Wanous,
Reichers, & Austin, 1994). Reichers et al. (1997) found that 23 per cent of managers were cynical about
the likelihood of successful organizational change, while 43 per cent of hourly employees held a similar
view. A consequence of cynical attitudes about change will likely be an unwillingness to exert much
effort on behalf of change goals (Wanous et al., 1994). As managers evaluate their efficacy for leading
change efforts, they may be expected to take into account subordinates' attitudes toward change.
Because managers rely on subordinates to help accomplish change goals, their skills, abilities, and
other performance characteristics should be significant factors in whether or not problems standing
in the way of successful change can be overcome. In a review of the group effectiveness literature,
Hackman and Morris (1975) concluded that the determinants of group performance could be summar-
ized in three themes: the effort applied by group members, their level of knowledge and skill, and the
performance strategies they use. Similarly, Yukl's (1971, 1994) multiple linkage model of leadership
presents a number of work unit performance characteristics that help explain the impact of a manager's
Superior antecedents
Another set of factors relates to the behavior of the manager's superior. Individuals partly evaluate
their own capabilities by observing others (Bandura, 1986); working with a superior who is seen as
an effective leader helps the manager develop a belief that he or she can perform in a similar manner.
Observing a leader can provide valuable information about the abilities, resources, and strategies that
can be applied to the difficult task of affecting organizational change (Gist & Mitchell, 1992; Wood &
Bandura, 1989), and can help the subordinate manager view situational constraints and resistance as
A superior can also help shape the subordinate manager's self-perceived leadership capabilities
through verbal persuasion, or coaching (Bandura, 1986). This may take the form of specific perfor-
mance feedback, or more general expressions of encouragement designed to convince the manager
that he or she is capable of doing more than currently thought possible. These messages communicate
high performance expectations, which can be a persuasive influence on an individual's perceived self-
efficacy (Gist, 1987; Korman, 1970). Conveying high performance expectations helps subtly persuade
Copyright ? 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
the subordinate manager to compare his or her abilities favorably against the requirements of the
leadership role.
Organizational antecedents
The organizational context in which a manager works will be an important influence on what he or she
can do (Bolman & Deal, 1991). A contextual characteristic that is expected to influence managers'
LSE is the degree to which the work environment is open to change (Kanter, 1983, 1999; Scott &
Bruce, 1994; Tichy & Ulrich, 1984). Research studying 'champions of change' has identified top man-
pushing change initiatives (Howell & Higgins, 1990). More specifically, an atmosphere that is suppor-
tive of change is one that encourages creative thinking, encourages risk taking over maintaining the
status quo, tolerates diversity of opinions, and promotes trying different approaches for solving famil-
iar problems (Howell & Higgins, 1990; Kanter, 1983, 1999; Scott & Bruce, 1994; Siegel &
Kaemmerer, 1978). Support for the proposed relationship between environmental receptivity to
change and LSE is provided by a managerial simulation study in which the malleability of an organi-
zational environment was manipulated (Bandura & Wood, 1989). Results showed that participants
who managed the simulated organization under the premise that it was not easily changeable quickly
lost faith in their managerial capabilities. In contrast, perceived organization malleability led to
In addition to a conducive atmosphere, introducing changes to work processes often requires resources
in the form of personnel, equipment, budget dollars, or, most simply, time. Lack of resources can be a
serious roadblock in a manager's path to accomplishing continuous improvement in his unit (Scott &
Bruce, 1994; Stewart, 1982). Thus, to the extent that resources are viewed as inadequate for supporting
change efforts, the manager's self-efficacy for successfully leading change may be diminished.
The last organizational factor expected to influence LSE is managers' job autonomy. In order for
managers to feel confident in their ability to lead change efforts, their jobs need to provide them the
opportunity to set new directions, build relationships and gain followers' commitment, and take the
actions necessary to overcome obstacles. More simply, the manager must have some choice about what
to do and how to do it (Stewart, 1982; Yukl, 1994). The aforementioned study of champions of change
found that these managers had broadly defined jobs and a significant amount of autonomy in their work
(Howell & Higgins, 1990). The presence or absence of such aspects of managerial authority as defining
goals for the unit, making work assignments, selecting new employees, and controlling expenditures
(Hammer & Turk, 1987; Stewart, 1982) is expected to influence managers' LSE.
A key proposition of this study is that managers' perceptions of their capabilities will be an important
influence on whether or not they attempt leadership. Specifically, subordinates of high LSE managers
Copyright J 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
are expected to rate them as more frequently pushing change and continuous improvements in their
work units, compared to low LSE managers. Support for linking LSE with leadership attempts can
be found in research on the causal mechanisms through which self-efficacy affects behavior
(e.g., Bandura, 1977, 1986; Wood & Bandura, 1989). Although testing these mediating processes is
beyond the scope of this study, they are summarized here to provide the theoretical rationale for this
key hypothesis.
First, a person's self-efficacy influences the activities that he or she chooses to engage in. People
approach and explore situations within their perceived capabilities, while avoiding situations they
think exceed their ability (Bandura, 1977). Also, research has shown that the greater confidence indi-
viduals have in their capabilities, the more vigorous their effort and persistence (Bandura, 1986).
Lastly, efficacy perceptions influence how much stress a person experiences when faced with difficult
challenges. Research indicates those who believe they are ill-prepared to handle difficult situations
experience more stress and anxiety and anticipate a greater number of difficulties (Bandura, 1986;
Wood & Bandura, 1989). Compared to low LSE managers, then, the high LSE manager should be less
likely to shy away from the difficult challenge of attempting change at work, and can be expected to
exert more effort and persist longer as obstacles are encountered. Relatively unburdened by the stress
and anxiety associated with self-doubts, the high LSE manager may be expected to respond more
the decision to attempt leadership, in that a manager who strongly believes in the organization's goals
and values should be more willing to contribute extraordinary efforts on its behalf (Mowday, Steers, &
Through meta-analysis, the average correlation between organizational commitment and job perfor-
mance has been found to be relatively small (average r = 0.14; Mathieu & Zajac, 1990); a stronger
relationship has been found between commitment and organizational citizenship behavior (OCB)
(Organ, 1988) (average r= 0.25 to 0.32; Organ & Ryan, 1995). This finding is noteworthy because
it indicates that an employee's decision to engage in discretionary work behavior is related to his or
her level of commitment to the organization. Leadership attempts can be viewed as 'discretionary'
work behavior; managers often have a choice about whether they get involved in introducing contin-
uous improvement efforts in their units. Some managers take the initiative to identify weak areas and
work with subordinates to improve things, while others seem content maintaining the status quo. The
OCB findings suggest that organizational commitment may amplify the relationship between LSE and
leadership attempts.
Hypothesis 6: Organizational commitment will moderate the relationship between LSE and leader-
ship attempts, such that this relationship will be stronger for those high in organizational
commitment.
An implicit assumption in this research has been that 'leaders make things happen'; sometimes, how-
ever, 'things make leaders happen' (Bolman & Deal, 1991). A review of past research on situational
Copyright ? 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
determinants of leadership has concluded that leader behavior is strongly influenced by the situation
(Yukl, 1989). Just as contextual conditions may constrain what a leader is able to do, they may also
provide an opportunity for managers to take on the challenge of leading change and continuous
improvement. An organization in crisis, where critical unit goals are at stake, achievement of these
goals is in jeopardy, and great time pressure exists, may present conditions that actually make it easier
Subordinates report their managers exercising more power and influence, acting more goal-directed,
and making higher quality decisions in crisis situations compared to non-crisis circumstances (Mulder,
de Jong, Koppelaar, & Verhage, 1986; Mulder, Ritsema van Eck, & de Jong, 1971). In terms of leading
change efforts, crisis situations by their nature call for new perspectives and a turning away from the
status quo (Bass, 1990). Managers may be given more leeway by their superiors to propose and imple-
ment new ways of doing things; then too, subordinates may be more amenable to changes that they
may have resisted in more stable times (House & Spangler, 1991).
Hypothesis 7: Perceived crisis will moderate the relationship between LSE and leadership attempts,
such that this relationship will be stronger when crisis perceptions are higher.
Organizational Context
The companies
The first organization, which contributed 113 managers to the final sample (75 per cent), was in the
business of managing commercial real estate properties across the United States. Managers in the
sample were the top-ranking employees at their properties. Each manager had four direct reports:
marketing director, operations director, office administrator, and security supervisor. The man-
agers' superiors were regional vice presidents who worked in a geographic location some distance
away from the managers' workplace. Our contact at company headquarters, the Corporate Direc-
tor of Training and Development, confirmed for us that these managers had substantial latitude in
how they managed their operations and were in a position to initiate continuous improvement
initiatives if they so desired. Indeed, the Director indicated that this behavior would be looked
The second site, the headquarters of an industrial chemicals firm in the Eastern United States,
was a more traditional corporate setting. In most cases, managers, direct reports, and superiors all
worked in the same geographic location. Managers from two product divisions along with man-
agers from a Corporate Staff group were surveyed. Those in the product divisions came from func-
tional areas including manufacturing, engineering and projects, marketing, and divisional
controllership. Corporate Staff managers worked in areas such as finance, human resources, envir-
onmental health and safety, and community relations. Our company contact, a Vice President and
General Manager of one of the product divisions, noted that a leadership education programme
was under development, a part of which would focus on promoting continuous improvement activ-
ities. He indicated that managers were beginning to get the message that this sort of leadership
behavior was encouraged and would increasingly be expected of them in the future.
Copyright G 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
Method
Sample
Surveys were distributed in two organizations to managers (total n = 191) and their direct reports
(total n = 679). The first organization, representing 75 per cent of the managers in the final sample,
was a firm engaged in managing commercial real estate properties throughout the United States.
The other data site was the corporate headquarters of an industrial chemicals firm. We explained to
the company contact in each organization (not study participants themselves) the outline of the project
and our overarching interest in studying the leadership of change and continuous improvement. The
company contact in each firm identified a level of management high enough that the incumbents would
have enough flexibility to push change in their units, and then provided us those managers' names. All
of the managers held positions at least one level above 'first-line' management. The combined sample
consisted of 150 managers (79 per cent response rate) and 415 direct reports (61 per cent). There was a
Measures
Data sources for the variables are listed in Table 1, along with means, standard deviations, and relia-
bility estimates (coefficient alpha). Managers provided data for LSE, organizational commitment, and
perceived crisis. Information on LSE antecedents was gathered from both managers and direct reports.
With the exception of internal locus of control, reliability estimates were 0.70 or higher for all
measures.
Variable M SD Alpha
Note: n = 150 for manager data; 415 for direct report data. Letter after variable name indicates data source: M = Manager,
Copyright ( 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
LSE antecedents. Successful experience in leadership roles was measured with biodata items adapted
from a published inventory (Glennon, Albright, & Owens, 1966). Participants responded to three
questions that asked about leadership experiences in school and community organizations, for
example: 'How often were you elected to a position of authority in high school or college
organizations?' While these kinds of activities may tend to be correlated, the presence of one does not
Locus of control was measured with five items from Rotter's (1966) original scale that represent a
'personal control' factor, reflecting individuals' beliefs about their ability to control what happens in
their lives (Gurin et al., 1978). Previous research using subsets of items from the Rotter scale with a
forced-choice response format have reported low coefficient alphas, below 0.70 (e.g., study one,
Greenberger, Strasser, Cummings, & Dunham, 1989; Gurin et al., 1978; Howell & Avolio, 1993).
Therefore, for this research we converted the items to a Likert format with a 5-point agree/disagree
response scale, following an approach successfully used by Greenberger et al. (study two) to improve
scale reliability. A sample item is 'When I make plans, I am almost certain that I can make them work.'
Modifying the scale did not produce the desired result, however, as a coefficient alpha of 0.63 was
obtained.
Self-esteem was measured by averaging responses to seven items from the Rosenberg Self-Esteem
Scale (Rosenberg, 1965), for example: 'I am able to do things as well as most other people.' A 5-point
The measure used to operationalize the superiors' leadership modelling variable needed to conform
to the definition of leadership adopted in this research; that is, the modelled behaviors needed to per-
tain to leading change. No established scale was found, and so four items were written that dealt with
pushing change, seeking continuous improvement, stimulating new directions for the department, and
persisting in efforts at improvement. Managers were asked to rate how effective the individuals to
whom they reported were at these behaviors, using a 5-point response scale ranging from 'not at all
A 6-item scale measuring superiors' coaching behavior was created from items developed by Yukl
as part of his leader behavior research programme (G. Yukl, personal communication, January 20,
1998). Managers reported how frequently their superiors engaged in these behaviors, such as
... expresses confidence in your ability to carry out a difficult project.' A 5-point response scale, ran-
Direct reports provided the data on subordinates' cynicism about organizational change, responding
to four items derived from an existing scale (Reichers et al., 1997). Direct reports were asked to think
about the overall attitude of the employees in their manager's unit, and rate the extent to which they
agreed with statements such as, 'On average, the employees in this manager's work unit seem to
For subordinates' performance characteristics, managers rated the overall, typical performance of
their work unit on seven dimensions of job performance, including quality of work, cooperation, and
initiative (Wayne & Ferris, 1990). A 5-point response scale with anchors of 'unsatisfactory' to
Support for change and resource supply, evaluated by direct reports, were measured with scales
developed by Scott and Bruce (1994). In the first case, seven items captured subordinates' relative
agreement with descriptions such as, 'This division seems to be more concerned with the status quo
than with change.' Resource supply was measured with five items, for example, 'Lack of funding to
For the job autonomy variable, managers were asked to review a list of seven managerial tasks and
rate their level of authority for each using a frequency response scale (1 rarely; 5 = always). These
Copyright ? 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
tasks included controlling expenditures of budget monies, selecting new employees, and initiating
discussions about work flow between units (Hammer & Turk, 1987).
Leadership self-efficacy. LSE reflects managers' judgments of their capabilities for leading change;
specifically, the construct captures managers' convictions that they can accomplish the following
leadership tasks with their work groups: (1) setting a direction for where the work group should be
headed; (2) gaining followers' commitment to change goals; and (3) overcoming obstacles standing in
the way of meeting change objectives. Accordingly, four items were written to align with each of these
three dimensions, resulting in a 12-item scale. A sample item from each of these three subscales is (1)
'I can develop plans for change that will take my unit in important new directions'; (2) 'I can obtain the
genuine support of my employees for new initiatives in the unit'; and (3) 'I can figure out ways for my
unit to solve any policy or procedural problems hindering our change efforts.'
Before final data collection, the LSE scale was pilot-tested with a sample (n = 55) of university
department heads and MBA students with prior managerial experience. As a result, some item mod-
ifications were made to improve clarity and reliability. In the instrument's final form, a 100-point prob-
ability response scale with 10-point increments was used. Respondents were asked to rate their degree
of confidence in their ability to perform various leadership tasks by circling a number from 0 to 100
per cent. Zero per cent reflected 'not at all confident', 50 per cent reflected 'intermediate level of con-
fidence', and 100 per cent reflected 'completely confident'. This type of 'strength of efficacy' response
scale has been used in previous research (Bandura & Wood, 1989; Locke, Frederick, Lee, & Bobko,
1984). An extensive evaluation of measurement quality for the LSE scale was conducted, including
factor analysis, reliability, and construct validity testing. These results are briefly summarized in
Leadership attempts. Eight items were written to correspond with the definition of leadership used in
this research. Subordinates rated their managers on the frequency with which they engaged in the
following behaviors: pushing change within the unit, seeking continuous improvement in the way
work gets done, quickly changing work processes that are not effective, persisting in efforts at
improving unit effectiveness, stimulating important new directions for the unit, making plans for
improvement that challenge the established way of doing things, using creative approaches for
accomplishing new goals within the unit, and initiating reviews of work processes in order to find more
effective methods. Five response options ranging from 'rarely' to 'frequently' were provided. Factor
analysis of these items indicated a single underlying factor, and a reliability estimate of 0.92 was
obtained.
(Mowday et al., 1979) was used to measure managers' organizational commitment. Sample questions
are 'I find that my values and the organization's values are very similar,' and 'I talk up this organization
to my friends as a great organization to work for.' A 5-point agree/disagree response scale was used.
Perceived crisis. Based on previous research, three elements of an organizational crisis situation were
identified: unit goals at stake that are critical to the larger organization, goal achievement in jeopardy,
and great time pressure (Mulder et al., 1971, 1986). Four statements were written to capture these
elements. Managers were asked to reflect back on their tenure in their current positions, and rate how
frequently these statements characterized the work environment in their units. Factor analysis revealed
a low factor loading for one of the items, and the reliability estimate was below 0.70 with this item
included in the composite variable. This item was subsequently deleted, improving the alpha estimate
to 0.70.
Copyright ? 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
Results
Before hypothesis testing, the factor structure, reliability, and construct validity of the newly devel-
oped LSE measure were evaluated. First, exploratory factor analysis was conducted on the 12-item
scale, using the principal factor method of factor extraction followed by promax (oblique) rotation
(Hatcher, 1994). LSE was proposed as a three-dimensional construct made up of the components of
not these three dimensions were distinct enough in these data to be considered separate variables for
analysis was investigated using several sources of information: the scree plot, proportion of variance
The scree plot did not provide a clear answer as to the number of factors. In terms of evaluating the
proportion of variance accounted for, no hard-and-fast rules exist for determining what amount should
be deemed large enough to retain a factor. A conservative recommendation is that the factor should
account for at least 5 to 10 per cent of the common variance (Hatcher, 1994). The factor analysis results
showed that the first three factors accounted for 80, 16, and 9 per cent, respectively of the common
variance in these data. In addition, the eigenvalues for the first three factors exceeded the average
eigenvalue of the reduced correlation matrix, indicating that each of these factors explained more var-
iance than the average variance contributed by one variable. Finally, the interpretability of retained
factors was evaluated. Loadings in the factor pattern matrix were examined, with the objective of find-
ing a solution in which all items loaded highly (>0.40) on a single factor, each factor had at least three
items with high loadings, and items loading on a particular factor shared some conceptual meaning
(Hatcher, 1994). A three-factor solution produced a factor pattern matrix that was relatively easy to
interpret, reflecting the three dimensions proposed a priori for the LSE construct. With the exception
of one item (subsequently deleted), each LSE item loaded highly on the dimension it was written to
reflect. The inter-factor correlations between the three LSE dimensions ranged from 0.50 to 0.63, with
the highest correlation found between gaining commitment and overcoming obstacles factors.
Based on the factor analysis results as well as the theoretical rationale for the three dimensions
presented earlier, a decision was made to proceed with the hypothesized three-dimensional LSE struc-
ture for the hypothesis tests.' Three LSE variables were formed by averaging the items associated with
the dimensions of direction-setting, gaining commitment, and overcoming obstacles. The first two
variables were 4-item measures, while overcoming obstacles was a 3-item composite. Zero-order
correlations between the three LSE variables ranged from 0.49 to 0.64 (see Table 2). The gaining com-
mitment and overcoming obstacles variables exhibited different relationships with other variables in
the matrix (e.g., with leadership attempts), providing additional support for retaining these as separate
LSE dimensions. Strong reliability estimates were obtained for the LSE variables. Coefficient alphas
for direction-setting, gaining commitment, and overcoming obstacles were 0.86, 0.92, and 0.86,
respectively.
Another preliminary step was the evaluation of construct validity of the new LSE measure. Toward
that end, criterion-related, convergent, and discriminant validity tests were conducted. The relationship
between LSE and leadership attempts (Hypothesis 5) provided a test of concurrent criterion-related
validity for LSE. As well, two established leader behavior rating scales, motivating/inspiring and
'Given the substantial inter-factor correlations, however, we also performed a supplementary analysis, replicating the hypothesis
tests with a unidimensional LSE variable (the average of the 12 items). Differences between the two analyses are summarized in
Copyright J 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
t-
C,..
Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1. LSE direction -
4. LSE total - - - -
0f
10. Sub. cynicism (16)t (19)* (10) (17)* 04 (05) 01 (22)* (25)* -
11. Sub. abilities 29* 39* 23* 36* 18* 20* 27* 02 01 (06)
13. Resource supply 19* 24* 20* 25* (05) 06 06 23* 20* (59)* 05 60*
14. Job autonomy 37* 31* 50* 47* 04 38* 27* 26* 08 (09) 12 09 16t
ta
15. Org. commitment 12 16* 28* 23* 15t 28* 23* 26* 29* 02 12 (01) 04 33*
16. Crisis (18)* (13) (28)* (23)* (09) (34)* (20)* (11) (03) (11) (01) 06 05 (34)* (23)* b
17. L'ship attempts 21* 20* 13 21* (07) 09 01 21* 27* (55)* (07) 55* 55* 19* 09 02
Note: Parentheses indicate negative correlations. Decimals have been omitted. For correlations involving data from direct reports, only cases with two or more raters were included.
Sample size for correlations varies with source of data, as follows: n (M, M) = 150; n (DR, DR) = 130; n (M, DR) = 130.
t0
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
problem-solving (Yukl, Wall, & Lepsinger, 1990), were collected from managers' direct reports in
order to enable further criterion-related validity testing. (These variables were not part of the theore-
tical model.) These leader behaviors were expected to be significantly related to two of the three LSE
dimensions, due to their similarity with the general leadership tasks of gaining followers' commitment
Convergent validity evidence was compiled by including measures similar to leadership confidence in
the surveys. Three existing measures of self-perceived leadership abilities were included (Singer, 1991).
The three ratings were: (1) effectiveness ('If you were in a leadership position, how effective would you
be as a leader?'), (2) ability-match ('How well does your own ability fit requirements for a leadership
position?'), and (3) ease of success ('How easy would it be for you to succeed in a leadership position?').
Lastly, discriminant validity evidence took the form of relationships between LSE, internal locus of
control, and trait self-esteem. The latter two constructs are relatively stable, dispositional variables that
reflect a generalized expectancy regarding personal control over outcomes and a generalized assessment
to vary with situational factors such as resource supply and subordinate attitudes. While internal locus
of control and trait self-esteem are hypothesized to positively relate to LSE when controlling for situa-
tional factors, LSE is expected to be distinguishable from these two dispositional variables.
(r = 0.21, p < 0.05) and LSE/gaining commitment (r = 0.20, p < 0.05). The relationship between lea-
dership attempts and LSE/overcoming obstacles was not significant. In the second criterion-related
validity test, LSE/gaining commitment and LSE/overcoming obstacles were significantly related
(p < 0.05) to the problem-solving leader behavior scale. The motivating/inspiring measure was not
The convergent validity test was positive. All three LSE dimensions were significantly related to the
three existing measures of self-perceived leadership abilities, although these relationships may be
inflated somewhat by same-source method variance. With respect to discriminant validity, although
significant relationships were found between LSE, trait self-esteem, and internal locus of control,
the size of the correlations indicated that these were perceived as sufficiently distinct constructs by
respondents. Specifically, for trait self-esteem, the correlations with the three LSE dimensions ranged
from 0.29 to 0.39, while the relationships between LSE dimensions and internal locus of control ran-
ged from 0.26 to 0.36. In summary, although the construct validity tests were not 100 per cent positive,
they seemed to provide adequate support for moving forward with the tripartite construction of LSE.
As noted in the description of measures, direct reports in each manager's work unit rated the manager
on leadership attempts, as well as providing information for three of the LSE antecedents: cynicism
about change, support for change, and resource supply. Before scores on these variables were averaged
to form work-unit level variables, inter-rater agreement and within-group variability relative to
between-group variability were examined (James, Demaree, & Wolf, 1984; Bartko, 1976; Shrout &
Fleiss, 1979). Inter-rater agreement within work units was evaluated with rwG (James et al., 1984)
and within-group variability relative to between-group variability was assessed with the intra-class
Inter-rater agreement was calculated for the 130 cases in which responses were received from at
least two direct report raters. The rwG value averaged 0.70 or greater for the four measures
(range =0.72 to 0.83). The 0.70 threshold has been suggested as representative of a 'good' amount
of within-group agreement (George & Bettenhausen, 1990). With respect to the number of work units
Copyright ? 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
that had adequate agreement (>0.70), the frequency ranged from a low of 74 per cent for the resource
supply variable to a high of 87 per cent for the leadership attempts variable. In terms of inter-rater
reliability, the intra-class correlation coefficient (ICC) will be high when within work unit variance
is small relative to between work unit variance (Bartko, 1976). For the set of four variables, signifi-
cance tests in each case led to the rejection of the null hypothesis that the population correlation equals
zero. In sum, the rwG and ICC results appear to justify aggregation to the work-unit level for these four
variables. Where data from direct reports were used in the following analyses, only those cases with
two or more direct report raters per manager were used (n = 130), in order to take advantage of the
Hypothesis tests
Hypotheses 1 proposed three individual antecedents of LSE. As shown in Table 2, the first of these,
successful leadership experience, was significantly related to the direction-setting component of LSE
(p < 0.05). Internal locus of control and trait self-esteem were significantly correlated with all three
LSE dimensions (p < 0.05). Thus, Hypothesis 1(a) received partial support, and 1(b) and 1(c) were
fully supported. Subordinates' cynicism about organizational change was negatively related to LSE/
gaining commitment (p < 0.05), a cross-source finding. Subordinates' cynicism was also modestly
negatively related with LSE/direction-setting (p < 0.10). These results partially supported Hypotheses
2(a). Subordinates' performance abilities was positively correlated with all three LSE components
(p < 0.05), as proposed in Hypothesis 2(b). Weak results were found for the two variables relating
to superiors (Hypothesis 3). The only marginally significant correlation (p < 0.10) occurred between
superior's leadership modelling and LSE/overcoming obstacles. The three organizational factors, sup-
port for change, resource supply, and job autonomy, were significantly related to all three LSE dimen-
Some of the significant correlations between proposed antecedents and LSE may be inflated by
same-source method variance, although the positive results for the subordinates' cynicism, support
for change, and resource supply variables would not be influenced by this. Also, intercorrelation
among some of the antecedents was evident. In order to control for this in examining the influence
of individual antecedents, separate regressions were run with the three LSE dimensions as dependent
variables (one-tailed tests, 0.05 significance level).2 Results are shown in Table 3. For LSE/direction-
setting, self-esteem, subordinates' performance abilities, and job autonomy emerged as significant
antecedents. For the gaining commitment component, subordinates' performance abilities and job
autonomy were significant. Self-esteem and job autonomy were significant when LSE/overcoming
Lastly, a canonical correlation analysis was performed to investigate overall relationships between the
LSE antecedents and LSE dimensions. One significant canonical correlation emerged (r = 0.64;
F = 3.26(33,387), p < 0.0001), indicating 40 per cent overlapping variance between the first variate pair.
Although a fair amount of judgment comes into play in drawing conclusions, an examination of the
canonical cross-loadings (Hair, Anderson, Tatham, & Black, 1998) indicated that a combination of high
self-esteem, an internal locus of control, high job autonomy, and a work unit with high-performing sub-
ordinates corresponded to greater LSE for direction-setting and overcoming obstacles to change.
2A dichotomous firm variable was included in the regressions to control for any differences between the two organizations that
comprised the sample. None of the following other potential control variables had significant bivariate correlations with LSE
dimensions: gender, manager's undergraduate GPA (self-reported), tenure in current position (all respondents) and in
Copyright ? 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
LSE antecedents b SE b SE b SE b SE
L'ship experience 0.11 0.09 0.06 0.09 -0.02 0.11 0.06 0.07
Locus of control -0.01 0.18 0.00 0.18 0.10 0.22 0.05 0.15
Sub. cynicism -0.11 0.15 -0.14 0.16 -0.01 0.19 -0.08 0.13
Sub. abilities 0.38* 0.19 0.63* 0.19 0.28 0.23 0.46* 0.16
Sup. modelling -0.02 0.12 0.06 0.12 0.11 0.15 0.06 0.10
Sup. coaching -0.16 0.12 -0.04 0.12 0.00 0.14 -0.07 0.10
Support change 0.18 0.18 0.12 0.18 0.12 0.22 0.13 0.15
Resource supply 0.05 0.20 -0.07 0.20 -0.14 0.24 -0.04 0.16
Job autonomy 0.57* 0.18 0.37* 0.19 0.95* 0.22 0.58* 0.15
Note. n = 130.
aControl variable.
*p < 0.05; one-tailed tests.
Hypothesis 5 stated that managers' LSE should be related to subordinates' ratings of their attempts
at leading change. As can be seen in Table 2, significant correlations were observed between subordi-
nates' leadership ratings and two of the three components of LSE: direction-setting (r = 0.21, p < 0.05)
and gaining commitment (r = 0.20, p <0.05). Thus, Hypothesis 5 received substantial but not
complete support.
and leadership attempts. This hypothesis was tested for each of the three LSE dimensions using hier-
archical regression. First, leadership attempts was regressed on the independent variable (the LSE
dimension) and the moderator (organizational commitment). Secondly, leadership attempts was
regressed on these two variables and the interaction term (LSE dimension x organizational commit-
ment). The results supported a moderator effect when LSE/overcoming obstacles was the independent
variable; in this model, the interaction term added unique explanatory power beyond the main effects
(t = 1.77, p < 0.05, one-tailed). Plotting the interaction (Cohen & Cohen, 1983) supported the form of
the hypothesized relationship; that is, the relationship between LSE/overcoming obstacles and leader-
Hypothesis 7 also proposed a moderator effect, setting forth managers' crisis perceptions as an influ-
ence on the relationship between LSE and leadership attempts. An analysis identical to the one
described above was performed. The interaction term was not significant in any of these regressions.
It was noted, however, that perceived crisis was significantly and negatively correlated with two of the
Supplementary analysis
As mentioned earlier, while we believe the factor analysis results provided enough support to proceed
with the three-dimensional LSE structure for the analysis, there were sizeable correlations found
Copyright 0 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
between the three LSE subscales. Therefore, we replicated the hypothesis tests using a total LSE vari-
able (12-item average; coefficient alpha = 0.92). The results are shown in Tables 2 and 3; findings that
are different from those presented above are summarized here. Leadership experience, which was
significantly correlated with only the direction-setting component of LSE (p <0.05), was not
significantly related to total LSE. Likewise, superior's leadership modelling, modestly related to
LSE/overcoming obstacles (p < 0.10), was not significantly correlated with total LSE. Subordinates'
cynicism about organizational change, which had mixed results across the three subscales, was a sig-
nificant antecedent of the total LSE variable (p < 0.05). The central argument of this research was that
managers' LSE would positively relate to subordinates' ratings of their leadership attempts
(Hypothesis 5). This hypothesis was supported for two of the three LSE dimensions in the initial
analysis, and was also supported for total LSE (p < 0.05). The moderator effect for organizational
commitment (Hypothesis 6) was not found using the total LSE scale.
Discussion
'Today, more than ever, it is more-not less-risk taking that American business vitally needs'
(McNair, 1954). These words were written almost 50 years ago, yet they portend a theme echoed with
increasing frequency by contemporary writers (e.g., Capowski, 1994; Harari, 1995; Kotter, 1990). As
environments rapidly change, managers are being asked to respond by actively seeking out new oppor-
tunities and leading their followers to exploit them to full advantage (Kotter, 1990).
The purpose of this study was to develop and test a leadership model that focused on a manager's
motivation for stepping forward to lead change efforts at work. Accordingly, a three-dimensional con-
struct, LSE, was developed to reflect managers' self-perceived capability for successfully executing
the behaviors required to effect change in the workplace. We examined the dimensionality, reliability,
and construct validity of the new measure, and while it appeared to perform acceptably for an initial
study, the validity tests indicate the need for additional work.
Consistent with self-efficacy theory (Bandura, 1977, 1986), the central hypothesis was that high
LSE managers would engage in more leadership attempts, compared to self-doubters. Mostly suppor-
tive findings were discovered, as LSE/direction-setting and LSE/gaining commitment were signifi-
cantly correlated with subordinate ratings of managers' leadership attempts. As for the third LSE
dimension, overcoming obstacles, an interaction effect was found. This suggests that a high level of
organizational commitment is necessary for managers' efficacy for overcoming obstacles to be trans-
lated into leadership action. In sum, our findings concerning the relationship between LSE and leader-
ship attempts were encouraging. An interesting avenue for future research will be investigating the
Several factors, both characteristics of the individual as well as features of the work context, were
proposed as influences on managers' LSE. Across several analyses, some of the most consistently sig-
nificant results were obtained for self-esteem, subordinates' performance abilities, and job autonomy.
For each of these variables, significant, positive relationships were found for at least two of the three
LSE dimensions in the regression analysis, with job autonomy exhibiting particularly strong results.
This makes sense, as leading change efforts should be easier for managers with greater control over
Substantial intercorrelation was evident among some of the LSE antecedents, particularly those for
which the data were obtained from direct reports (i.e., subordinates' cynicism, support for change, and
resource supply; average r = 0.60). In each case, significant zero-order correlations were found
Copyright ? 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
between these antecedents and one or more of the LSE dimensions. None of the regression results was
significant, however, as the intercorrelation likely constrained the emergence of any one of these as
significant when all three were in the model. Thus, some of the more rigorous hypothesis tests,
Limitations of this study include reliance on self-report data for several of the hypothesis tests, and
the accompanying risk of inflated correlations due to method variance. It should be noted, however,
that the primary hypothesis linking LSE with leadership attempts involved data from separate sources,
and this hypothesis was substantially supported. As well, several significant cross-source findings at
the correlation level were found between proposed antecedents and LSE dimensions. Another limita-
tion concerns the relatively low number of direct report raters per manager in this study (mean = 2.7).
Although managers in our sample generally had small spans of control (75 per cent had five or fewer
direct reports), this sample characteristic limits the generalizability of our findings to other contexts.
Three aspects of the results were particularly surprising to us. First, mastery experiences are tradi-
tionally viewed as one of the most important influences on self-efficacy, 'because they provide the
most authentic evidence of whether one can muster whatever it takes to succeed' (Bandura, 1997,
p. 80). In this study, the successful leadership experience variable was related to only one of three
LSE dimensions, and only at the correlation level. It may be that the biodata items were deficient
in capturing the efficacy-building aspect of leadership experience, as they tended to focus on frequency
of involvement (e.g., 'How often were you elected to a position of authority... ') rather than actual
success or failure in previous leadership experiences. Further work to develop a stronger measure for
Second, the lack of influence of superiors' behavior on managers' LSE was interesting. We noted,
however, that superiors' modelling and coaching were both significantly correlated with managers' lea-
dership attempts. For instance, managers who evaluated their superiors as effective at pushing change
(i.e., leadership modelling) were rated by direct reports as engaging in more leadership attempts them-
selves. It appeared that superiors were having an effect on managers' leadership behavior, yet this influ-
ence did not flow through LSE. Perhaps superiors who are 'change agents' themselves offer more
rewards to their subordinate managers who display similar types of behavior. This reinforcement
may be a potent motivator of managers' actions, eclipsing the proposed leadership modelling effect
on self-efficacy. On the other hand, reverse causality may be at work: managers who frequently engage
in leadership attempts may stimulate more leadership modelling and coaching from their superiors.
Lastly, although managers' crisis perceptions did not moderate the relationship between LSE and
leadership attempts as proposed, this variable was significantly and negatively correlated with two of
three LSE dimensions. This result calls to mind Stewart's (1976, 1982) distinction between managerial
activities that are self-generating versus those that are primarily reactive. Stewart's theory proposes that
more initiative and planning can be expected from a manager in a primarily self-generating job than
from a manager in a job that mainly involves responding to others' problems. In this study, a manager's
characterization of the work environment as being in a reactive 'crisis mode' was related to diminished
self-efficacy for initiating change efforts. Rather than directly influencing leadership attempts, this con-
Author biographies
Indiana. She earned her Ph.D. degree in Organizational Behavior and Human Resource Management from
Copyright ? 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
the Krannert Graduate School of Management, Purdue University, in 1999. Her research interests include
Stephen G. Green is the Basil S. Turner Professor of Management at the Krannert Graduate School of
Management, Purdue University, West Lafayette, Indiana. He earned his Ph.D. degree in Psychology
at the University of Washington. His current research interests include leadership, socialization, and
the management of innovation. Steve's research has been supported by the National Science Founda-
tion, the Center for Innovation Management Studies, the Department of Defense, Krannert's Centerfor
International Business and Economic Research, and the Center for the Management of Manufacturing
Enterprises.
References
Bandura A. 1977. Self-efficacy: toward a unifying theory of behavioral change. Psychological Review 84: 191-215.
Bandura A. 1986. Social Foundations of Thought and Action: A Social Cognitive Theory. Prentice-Hall:
Bandura A. 1997. Self-efficacy: The Exercise of Control. W. H. Freeman and Company: New York.
Bandura A, Wood R. 1989. Effect of perceived controllability and performance standards on self-regulation of
complex decision making. Journal of Personality and Social Psychology 56: 805-814.
Bartko JJ. 1976. On various intraclass correlation reliability coefficients. Psychological Bulletin 83: 762-765.
Bass BM. 1990. Bass & Stogdill's Handbook of Leadership: Theory, Research, and Managerial Applications
Bolman LG, Deal TE. 1991. Reframing Organizations: Artistry, Choice, and Leadership. Jossey-Bass: San
Francisco.
Brockner J. 1988. Self-esteem at Work: Research, Theory, and Practice. Lexington Books: Lexington, MA.
Capowski G. 1994. Anatomy of a leader: where are the leaders of tomorrow?. Management Review 83: 10-17.
Cohen J, Cohen P. 1983. Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences (2nd ed.).
Dean JW Jr, Brandes P, Dharwadkar R. 1998. Organizational cynicism. Academy of Management Review 23:
341-352.
George JM, Bettenhausen K. 1990. Understanding prosocial behavior, sales performance, and turnover: a group-
Gist ME. 1987. Self-efficacy: implications for organizational behavior and human resource management.
Gist ME, Mitchell TR. 1992. Self-efficacy: a theoretical analysis of its determinants and malleability. Academy of
Glennon JR, Albright LE, Owens WA. 1966. A Catalog of Life History Items. The Creativity Research Institute of
Greenberger DB, Strasser S, Cummings LL, Dunham RB. 1989. The impact of personal control on performance
and satisfaction. Organizational Behavior and Human Decision Processes 43: 29-51.
Gurin P, Gurin G, Lao RC, Beattie M. 1969. Internal-external control in the motivational dynamics of Negro
Gurin P, Gurin G, Morrison BM. 1978. Personal and ideological aspects of internal and external control. Social
Hackman JR, Morris CG. 1975. Group tasks, group interaction process, and group performance effectiveness: a
review and proposed integration. In Advances in Experimental Social Psychology (Vol. 8), Berkowitz L (ed.).
Hair JF Jr, Anderson RE, Tatham RL, Black WC. 1998. Multivariate Data Analysis (5th ed.). Prentice-Hall: Upper
Hammer TH, Turk JM. 1987. Organizational determinants of leader behavior and authority. Journal of Applied
Harari O. 1995. The successful manager in the new business world. Management Review 84: 10-12.
Copyright ? 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
Hatcher L. 1994. A Step-by-step Approach to Using the SAS System for Factor Analysis and Structural Equation
Hemphill JK, Coons AE. 1957. Development of the leader behavior description questionnaire. In Leader
Behavior: Its Description and Measurement, Stogdill RM, Coons AE (eds). Bureau of Business Research, Ohio
House RJ, Baetz ML. 1979. Leadership: some empirical generalizations and new research directions. In Research
in Organizational Behavior (Vol. 1, pp. 341-423), Staw BM (ed.). JAI Press: Greenwich, CT.
House RJ, Podsakoff PM. 1994. Leadership effectiveness: past perspectives and future directions for research. In
Organizational Behavior: The State of the Science (pp. 45-82), Greenberg J (ed.). Erlbaum: Hillsdale, NJ.
House RJ, Spangler WD. 1991. Personality and charisma in the U.S. presidency: a psychological theory of leader
Howell JM, Avolio BJ. 1993. Transformational leadership, transactional leadership, locus of control, and support
for innovation: key predictors of consolidated business unit performance. Journal of Applied Psychology 78:
891-902.
Howell JM, Higgins CA. 1990. Champions of change: identifying, understanding, and supporting champions of
Jacobs TO, Jaques E. 1990. Military executive leadership. In Measures of Leadership (pp. 281-295), Clark KE,
Jago AG. 1982. Leadership: perspectives in theory and research. Management Science 28: 315-336.
James LR, Demaree RG, Wolf G. 1984. Estimating within-group inter-rater reliability with and without response
Kanter RM. 1983. The Change Masters. Simon & Schuster: New York.
Kanter RM. 1999. Change is everyone's job: managing the extended enterprise in a globally connected world.
Korman AK. 1970. Toward an hypothesis of work behavior. Journal of Applied Psychology 54: 31-41.
Kotter JP. 1988. The Leadership Factor. Free Press: New York.
Kotter JP. 1990. A Force for Change: How Leadership Differs from Management. Free Press: New York.
Krech D, Crutchfield RS. 1948. Theory and Problems of Social Psychology. McGraw-Hill: New York.
Lefcourt HM. 1991. Locus of control. In Measures of Personality and Social Psychological Attitudes (pp. 413-
499), Robinson JP, Shaver PR, Wrightsman LS (eds). Academic Press: San Diego, CA.
Locke EA, Frederick E, Lee C, Bobko P. 1984. Effect of self-efficacy, goals, and task strategies on task
Mathieu JE, Zajac DM. 1990. A review and meta-analysis of the antecedents, correlates, and consequences of
McNair MP. 1954. Tough-mindedness and the case method. In The Case Method at the Harvard Business School,
Mowday RT, Steers RM, Porter LW. 1979. The measurement of organizational commitment. Journal of
Mulder M, de Jong RD, Koppelaar L, Verhage J. 1986. Power, situation, and leaders' effectiveness: an
Mulder M, Ritsema van Eck JR, de Jong RD. 1971. An organization in crisis and non-crisis situations. Human
Organ DW. 1988. Organizational Citizenship Behavior: The Good Soldier Syndrome. Lexington Books:
Lexington, MA.
Organ DW, Ryan K. 1995. A meta-analytic review of attitudinal and dispositional predictors of organizational
Porter LW, Steers RM, Mowday RT, Boulian PV. 1974. Organizational commitment, job satisfaction, and turnover
Rauch CF, Behling 0. 1984. Functionalism: basis for an alternate approach to the study of leadership. In Leaders
and Managers: International Perspectives on Managerial Behavior and Leadership, Hunt JG, Hosking DM,
Reichers AE, Wanous JP, Austin JT. 1997. Understanding and managing cynicism about organizational change.
Rosenberg M. 1965. Society and the Adolescent Self-image. Princeton University Press: Princeton, NJ.
Rotter JB. 1966. Generalized expectancies for internal versus external control of reinforcement. Psychological
Copyright ) 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions
Scott SG, Bruce RA. 1994. Determinants of innovative behavior: a path model of individual innovation in the
Shrout PE, Fleiss JL. 1979. Intraclass correlations: uses in assessing rater reliability. Psychological Bulletin 86:
420-428.
Siegel S, Kaemmerer W. 1978. Measuring the perceived support for innovation in organizations. Journal of
Singer M. 1991. The relationship between employee sex, length of service and leadership aspirations: a study from
valence, self-efficacy and attribution perspectives. Applied Psychology: An International Review 40: 417-436.
Stewart R. 1982. Choices for the Manager: A Guide to Understanding Managerial Work. Prentice-Hall:
Stogdill RM. 1974. Handbook of Leadership: A Survey of Literature. Free Press: New York.
Tichy N, Ulrich D. 1984. Revitalizing organizations: the leadership role. In Managing Organizational Transitions
(pp. 240-264), Kimberly JR, Quinn RE (eds). Richard D. Irwin: Homewood, IL.
Wanous JP, Reichers AE, Austin JT. 1994. Organizational cynicism: an initial study. In Academy of Management
Best Paper Proceedings 1994 (pp. 269-273), Moore DP (ed.). Academy of Management.
Wayne SJ, Ferris GR. 1990. Influence tactics, affect, and exchange quality in supervisor-subordinate interactions:
a laboratory experiment and field study. Journal of Applied Psychology 75: 487-499.
Wood R, Bandura A. 1989. Social cognitive theory of organizational management. Academy of Management
Yukl G. 1971. Toward a behavioral theory of leadership. Organizational Behavior and Human Performance 6:
414-440.
Yukl G. 1989. Managerial leadership: a review of theory and research. Journal of Management 15: 251-289.
Yukl G. 1994. Leadership in Organizations (3rd ed.). Prentice Hall: Englewood Cliffs, NJ.
Yukl G. 1998. Leadership in Organizations (4th ed.). Prentice Hall: Upper Saddle River, NJ.
Yukl G, Van Fleet DD. 1990. Theory and research on leadership in organizations. In Handbook of Industrial and
Organizational Psychology (2nd ed., Vol. 3, pp. 148-197), Dunnette MD, Hough L (eds). Consulting
Yukl G, Wall S, Lepsinger R. 1990. Preliminary report on validation of the Managerial Practices Survey. In
Measures of Leadership (pp. 223-238), Clark KE, Clark MB (eds). Leadership Library of America, Inc.: West
Orange, NJ.
Copyright ( 2002 John Wiley & Sons, Ltd. J. Organiz. Behav. 23, 215-235 (2002)
This content downloaded from 176.31.0.28 on Tue, 01 Mar 2016 04:00:40 UTC
All use subject to JSTOR Terms and Conditions