!
CITY MANAGERS’ LEADERSHIP CREDIBILITY:
EXPLAINING THE VARIATIONS OF SELF-OTHER
ASSESSMENTS
NATHAN J. GRASSE
Central Michigan University
BRIANNE HEIDBREDER
Kansas State University
DOUGLAS M. IHRKE
University of Wisconsin-Milwaukee
ABSTRACT
Leadership skills are clearly important for city managers. Credible
leaders have the potential to have a positive impact on employee
attitudes, beliefs, and behaviors, as well as organizational outcomes.
Despite the importance of leadership and the credibility of leaders, we
have little understanding of how city managers assess their own
leadership credibility. In order to improve our understanding, this
research explores factors that influence managers’ awareness of their
credibility as leaders. Using data from Michigan municipalities, we
find that both individual and organizational characteristics are
correlated with a city manager’s ability to accurately assess their
leadership skills when compared to their subordinates’ assessments.
PAQ WINTER 2014
545
INTRODUCTION
The concept of “leadership” is arguably one of the most
studied topics in organizational behavior (Fernandez, 2005;
Kaiser, Hogan, & Craig, 2008; Van Wart, 2003). For decades,
scholars have attempted to understand effective leadership at the
individual level and the relationship between quality leadership
and success at the organizational level (for some classic
examples, see Boyatzis, 1982; Burns, 1978; Kouzes & Posner,
1987; Stogdill, 1948, 1974; Terry, 2003). Even though the
public administration literature has generated a substantial
volume of leadership scholarship, our understanding of
leadership both conceptually and as it relates to organizational
outputs remains elusive (Van Wart, 2003).
As James
MacGregor Burns states, “leadership is one of the most observed
and least understood phenomena on earth” (1978, p. 3).
Therefore, our efforts to understand leadership must continue
(Van Wart, 2003).
The public-sector literature has notably failed to create
one unifying or dominant theory of leadership (Terry, 2003; Van
Wart, 2003). However, it has produced a vast literature on a
variety of different aspects of leadership (Van Wart, 2013). One
particular path of scholarship examines the role of the
transformational leader in organizations (Bass, 1985, 1990,
1996; Burns 1978). We build on this research by examining one
of the defining traits of transformational leaders: leadership
credibility. Namely, we argue leadership credibility exists when
leaders are highly self-aware (Sosik & Megerian, 1999).
In this article, we develop a measure of a leader’s selfawareness by asking them to rate their own performance and
comparing these ratings to evaluations by their subordinates.
We then attempt to identify characteristics that influence the
accuracy of manager’s self-evaluations in order to improve our
understanding of leadership credibility.
Our findings are
important, as research from the field of organizational
psychology has shown individuals who are aware of their
performance behave quite differently than those who are not
self-aware (Atwater, Ostroff, Yammarino, & Fleenor, 1998;
Atwater & Yammarino, 1992; Sosik & Megerian, 1999; Tekleab,
!
546
PAQ WINTER 2014
Sims, Yun, Tesluk, & Cox, 2008). Before turning to our model,
it is important to review the relevant literature on organizational
leadership.
LITERATURE REVIEW
The topic of “leadership” has been an interest to public
administration scholars for decades, yet there are still many gaps
in our understanding of what makes a leader great and the impact
that leadership has on organizations (Van Wart, 2003). The gap
in our knowledge may be related to the fact that as a concept
“leadership” can be quite amorphous. As Stodgill (1974, p. 259)
concludes, “there are almost as many definitions of leadership as
there are persons who have attempted to define the concept.”
This definitional problem has yet to be resolved in any coherent
fashion, but there are reasons to be encouraged.
Scholars have long noted the general distrust toward
government and hostility toward bureaucratic growth (Meier &
Bohte, 2006; Terry, 2003; West, 1995). However, the fact
remains that over time we have come to expect more and more
from our government, making the behavior of individuals within
the bureaucracy at all levels critical to the daily lives of so many
(Terry, 2003). This has led some scholars to question the role
that leaders play in organizational success (Bass, 1985, 1990,
1996; Burns, 1978; Fernandez, 2004; Hennessey, 1998). While
many scholars (Fernandez, 2005; Terry, 2003) assert the
importance of leaders in the public-sector, others question
whether leadership truly matters (Boyne, 2002). Wright and
Pandey (2010, p. 75-76) argue that “transformational leaders are
expected to be both less common and less effective in public
sector organizations than private sector organizations because
the former rely more on bureaucratic control mechanisms.”
These arguments are typically founded on the conception of
public-sector bureaucracies as being bogged down by political
control, red tape, and low levels of managerial autonomy
(Boyne, 2002). If this assertion is true, we should expect
leadership to have very little impact on organizational success in
the public sector. On the contrary, there is a growing tradition in
the literature that focuses on how leadership affects the ability of
!
PAQ WINTER 2014
547
government to provide high quality goods and services
effectively and efficiently (Fernandez, 2004; Hennessey, 1998;
Van Wart, 2003) and seeks to combine leadership practices and
behaviors with organizational outcomes (Golembiewski, 1989;
Schein, 1992; Church, 1995; Kaiser, Hogan, & Craig, 2008;
Tekleab et al., 2008; Moynihan, Wright, & Pandey, 2012;
Wright, Moynihan, & Pandey, 2012).
Research suggests that leadership does exist and, at a
minimum, involves exerting influence on others (Bass, 1985,
1990, 1996, Burns, 1978; Kouzes & Posner, 1987; Terry, 2003;
Van Wart, 2013). In an organizational context, leaders influence
employees toward goal achievement through their actions and
behaviors over time (Fernandez, 2005; Hennessey, 1998;
Trottier, Van Wart, & Wang, 2008). Effective leaders can
stimulate organizational goal achievement both directly and
indirectly. For instance, leaders can influence or change
organizational cultures for the better, as measured by the impact
of leadership on factors such as employee motivation and
mission valence (Wright, Moynihan, & Pandey, 2012), as well as
employee experience with red tape (Moynihan, Wright, &
Pandey, 2012). Employee attitudes and behaviors are, in turn,
influenced by the culture of their organizations. The actions and
behaviors of employees, along with the culture of the
organization, add up to “more or less” organizational
performance. These successful leaders are therefore referred to
as transformational in that they transform the organization into
something unlike it was in the past (Bass, 1985, 1990, 1996;
Burns, 1978; Moynihan, Wright, & Pandey, 2012; Wright,
Moynihan, & Pandey, 2012).
Kouzes and Posner (1987) argue that leadership, or the
ability to influence others and transform organizations, depends
on the level of credibility leaders have with their employees.
They assert that the process of developing credibility involves 5
broad practices and 10 commitments (Table 1). They argue that
when leaders practice these behaviors and carry out these
commitments, their followers tend to believe they are credible.
Credibility, in turn, enables leaders to transform their
organizations (Kouzes & Posner, 1987).
!
548
PAQ WINTER 2014
Credibility, like a bank account, is built up over time
through deposits in the form of the practices and commitments.
Withdrawals from the account, in the form of negative
reinforcing actions and behaviors, can be very costly to leaders
and their organizations. The goal for leaders is to build the
account in the form of employee commitment, which is
absolutely essential for transformational change to come about
(Kouzes & Posner, 1987).
Table 1.
Kouzes and Posner’s Practices and Commitments of Credible
Leadership.
CHALLENGING THE PROCESS
Search for opportunities
Experiment and take risks
INSPIRING A SHARED VISION
Envision the future
Enlist others
ENABLING OTHERS TO ACT
Foster collaboration and trust
Strengthen others
MODELING THE WAY
Set the example
Plan small wins
ENCOURAGE THE HEART
Recognize individual contribution
Celebrate accomplishments
The primary interest of this research concerns disparities
in perceptions of leader credibility between leaders themselves
and their subordinates. We explore the literature on “self-other”
perceptions as a means for providing the reader with insight into
this measurement technique, and the implications it has for
leader effectiveness and development. It is important that public
administration literature include research on the psychological
factors and processes that might mediate the effectiveness of
leadership efforts. Much of current work relies on assumptions
about these factors and processes borrowed from organizational
psychology, without verifying whether or not they function in
!
PAQ WINTER 2014
549
same way across sectors (Atwater, Ostroff, Yammarino, &
Fleenor, 1998; Atwater & Yammarino, 1992; Sosik & Megerian,
1999).
SELF-OTHER PERCEPTIONS OF LEADERSHIP
Arguably, one way administrators can work toward
increasing their credibility as leaders is to demonstrate emotional
intelligence, which includes the trait of self-awareness (Sosik &
Megerian, 1999). Self-awareness, as measured using self-other
ratings, is “integral to transformational leadership effectiveness”
(Sosik & Megerian, 1999; p. 368). Studies have demonstrated
that a leader’s self-awareness has implications for a wide array
of individual and organizational outcomes including employee
self-esteem, satisfaction with supervision (Baird 1977), and
leadership effectiveness (Fleenor & McCauley, 1996; Atwater,
Ostroff, Yammarino, & Fleenor, 1998; Ostroff, Atwater, &
Feinberg, 2004, Fleenor, Smither, Atwater, Braddy & Sturm,
2010). Unfortunately, while these perceptions have received a
good deal of attention from a variety of fields ranging from
psychology to human resource management, it is hard to be
certain that the underlying theory will be applicable in public
organizations (e.g., Fleenor & McCauley, 1996, Farh & Dobbins,
1989, Gibbons 1983, Froming & Carver, 1981).
To the keen observer, it should not be surprising that
self-ratings tend to be “inflated, invalid, biased, inaccurate, and
generally suspect when compared to the ratings of others or more
‘objective’ criteria” (Yammarino & Atwater, 1993, p. 231).
While most people tend to have an inflated image of self, selfratings are not without merit. Comparing how we perceive
ourselves to how others perceive us can yield invaluable
information for leadership development as it can indicate a
leader’s self-awareness (Sosik & Megerian, 1999; Tekleab et al.,
2008).
A self-other ratings comparison can result in three
possible outcomes. Estimators can be categorized as accurate
estimators, over-estimators or under-estimators.
Accurate
estimators are those focal individuals (the person receiving the
feedback) whose self-ratings are in agreement with the ratings of
!
550
PAQ WINTER 2014
the relevant others. Over-estimators are those focal individuals
whose self-ratings are significantly inflated above the ratings of
relevant others. Under-estimators are those focal individuals
whose self-ratings are significantly deflated below the ratings of
the relevant others (Yammarino & Atwater, 1993).
Research indicates that individuals who are capable of
integrating outside feedback into their self-evaluations are more
likely to assess their performance accurately (Atwater &
Yammarino, 1992). Individuals’ abilities to self-evaluate have
implications for their behavior; over-estimators are highly
unlikely to accept feedback regarding their performance and are
therefore unlikely to make modifications to their behavior, while
individuals who accurately assess or under-estimate their
performance are more likely to accept feedback and modify their
behavior (Ashford 1989).
Differences between self-other
assessments can also affect employees. Studies have identified
that subordinates’ attitudes and performance are influenced by
the accuracy of leaders’ self-assessments when compared to their
subordinates’ assessments (Atwater, Ostroff, Yammarino, &
Fleenor, 1998). Specifically, subordinates of over-estimators are
less satisfied and productive than the employees of accurate or
under-estimators (Moshavi, Brown, & Dodd, 2003).
We seek to understand local government managers’
awareness of their leadership credibility. Psychology has
identified perceptions related to leadership as particularly useful
for comparing self-other evaluations, as understanding ones’
leadership performance depends upon interactions with
subordinates (Atwater & Yammarino, 1993), and subordinates’
perceptions have implications for superiors’ abilities to lead
(Bass, 1990; Ashford, 1989), but we have little knowledge about
how these perceptions influence the practice of management in
the public sector. Leadership credibility has been identified as
an immensely important resource for city managers, conditioning
their ability to manage their organizations (Kouzes & Posner,
1993). Unfortunately, in the context of local government, we
have yet to identify what factors influence leaders’ awareness of
their credibility. This work is a first step toward understanding
self and other perceptions in this important area of management,
comparing managers’ and subordinates’ evaluations of
!
PAQ WINTER 2014
551
leadership credibility, seeking to understand incongruencies
between these sets of perceptions.
We utilize personal,
organizational, and environmental factors to explain differences
in self-other perceptions.
HYPOTHESES
Research on private sector organizations has found that
individual characteristics influence self-other perceptions of
performance (Atwater & Yammarino, 1992; Fleenor, McCauley,
& Brutus, 1996; Ostroff, Atwater, & Feinber, 2004). We believe
that the personal or background characteristics of individuals
will also influence whether city managers are accurate, overestimators, or under-estimators of their leadership capabilities.
In this study, we examine whether the background characteristics
of tenure, age, education, and gender correlate with our three
categories of self-other perceptions.
We believe that age will have a negative association with
accurate self-perception. Age has been demonstrated to lead to
overestimation (Ostroff, Atwater, & Feinber, 2004). In in the
private-sector literature, this has been attributed to decreased
receptivity (Ryan, Brutus, Gregurus, & Hakel, 2004). In our
opinion, this may hold true for city managers, but it also could be
that life experiences increase their ability to incorporate new
information into their self-perceptions.
City management is a challenging profession, and
reports by the International City/County Management
Association (ICMA) indicate that an average tenure is less than
eight years for these leaders (ICMA 2009). We expect that longserving managers have earned their tenure, in part, by being
aware of how they are perceived by others, particularly their city
councils, but also by their department heads. It may be that
experienced city managers have learned the importance of
others’ perceptions, informing self-assessments with this
information. This would comport with some of the literature
from organizational psychology, in which experience has been
identified as a positive predictor of accurate assessment (Atwater
!
552
PAQ WINTER 2014
& Yammarino, 1992; Ostroff, Atwater, & Feinberg, 2004,
Fleenor, Smither, Atwater, Braddy & Sturm, 2010).4
Education also may lead to more accurate assessment, as
managers with more training are better able, or prone, to utilize
the information others can provide to improve their management
skills.
There are two potential explanations for this
phenomenon.
The first comes from the literature on
organizational psychology and relies on the relationship between
education and cognitive ability.
Specifically, education
increases one’s capability to utilize information effectively to
make choices, including decisions related to their relationships to
subordinates (Ostroff, Atwater, & Feinberg, 2004). In this sense,
education serves as a crude measure of intelligence, which has
been found to correlate with leader behavior, although not
always in the expected direction (Atwater & Yammarino, 1992;
Bass, 1990; Lord el al., 1986; Mabe & West, 1982). A second
explanation relies on the specific skills education can provide.
We argue that city managers, often educated with specific skills
in the area of performance evaluation, are predisposed to possess
the management skills specifically needed to process information
related to their own performance. This should lead to a more
constructive use of the information provided by interactions with
employees.
Gender is the final demographic trait that we posit will
affect accurate self-assessment. We expect that female leaders
may more
accurately assess their leadership capabilities
compared to their male colleagues (Appelbaum, Audet, &
Miller, 2003; Roberts 1991). As leaders, women have been
found to be particularly effective communicators (Fox &
Schuhmann, 1999) with strong “people skills” compared to their
male counterparts (Appelbaum, Audet, & Miller, 2003). This
personality feature, among others, has led some scholars to view
female leaders as having the ability to direct their leadership
toward transformational goals (Appelbaum, Audet, & Miller,
2003). Research by Roberts (1991) suggests that women change
their self-evaluations to be more congruent with the evaluations
4
There is some disagreement in this literature however, as some studies find
that experience leads to over assessment, functioning much like age (Ashford
1993: Van Velsor, Ruderman, & Phillips, 1991).
!
PAQ WINTER 2014
553
of others. Women tend to view evaluations by others to be more
informational than do men, and are more likely to use this
information when evaluating their own performance (Fleenor &
McCauley, 1996). Therefore we hypothesize that gender will
correlate with self-other perceptions, particularly with the
category of managers who self-rate accurately.
The elements of organizational life may also influence
self-other perceptions.
Specifically, we hypothesize that
organizations with healthy cultures will allow for better
communication and will therefore better inform managers of
their performance. We examine culture from two perspectives,
employee change receptivity and propensity to identify with the
organization, expecting that organizations with adaptive and
committed employees will be more likely to provide useful
information to city managers about their performance. This is
due to more meaningful interactions with these employees,
allowing for more constructive feedback between managers and
their subordinates, which should result in increased selfawareness (Atwater & Yammarino, 1992).5 Therefore, they will
be more “in tune” with the manager, leading us to hypothesize
that employee change motivation will be significantly associated
with accurate estimators.
Relatedly, employees who are
committed to the organization may be more likely to coordinate
with others and practice open communication, part of the cycle
of regenerative interactions identified by Golembiewski (1989).
Finally, we explore the two variables that indirectly
measure the environment in which municipalities are nested.
Here we are attempting to measure the characteristics of cities
that research on urban politics has consistently shown to
influence city governance and management (Svara, 1991).
Specifically, we hypothesize smaller communities will have
more accurate estimators, as will those with more homogeneous
populations. These hypotheses are consistent with Svara’s work,
where he suggests communities with council-manager
governments, which tend to be relatively small and
5
!Functioning similar to the influence of healthy leader member exchange
(Gerstner & Day, 1997).!
!
554
PAQ WINTER 2014
homogeneous, have more cooperative rather than conflictual
relations within their governments.
Data
In order to examine self-other perceptions of leadership,
this study draws data from a 2006 survey of Michigan local
government officials serving communities with populations of
10,000 or more. This survey collected data from department
heads, council members, city managers, and mayors. A total of
1272 individuals from 92 communities were surveyed with 466
returning survey materials, resulting in a 36.64% response rate.
This study incorporates additional data from the US Bureau of
the Census (2000) and Legislative Service Bureau of Michigan
(2004) to supplement the survey data.
Methods
The analysis utilizes a few variations of one primary
dependent variable, which measures the accuracy of city
managers perceptions of their leadership credibility. In order to
create this variable, we compared city managers’ selfassessments with the assessments of key subordinates, in this
case department heads in their local governments.
The
differences between these self and other assessments reflect city
managers’ self-awareness.
Leadership credibility is measured using an eight-item
scale developed from the work of Kouzes and Posner (1993) by
Gabris and Mitchell (1991).6 This variable assesses whether or
not employees feel the city manager is trustworthy,
communicates well, and behaves with integrity.7 The average
scores of department heads were differenced from managers’
self-assessments. These scores are utilized to create leadership
credibility difference scores, which indicate the degree to which
these supervisors and their subordinates differ when assessing
managers’ credibility.
The literature on leadership suggests that individuals
who over-estimate their own leadership capabilities are less
6
This scale has been utilized in other studies of leadership credibility (Gabris,
Golembiewski and Ihrke 2001; Gabris and Ihrke, 1996).
7
All eight items in this scale are described in Appendix A.!
!
PAQ WINTER 2014
555
effective than individuals who are aware of their leadership skills
(Bass & Yammarino, 1991; Atwater & Yammarino, 1992;
Tekleab et al., 2008). This seems to be due to their inability (or
unwillingness) to use feedback to improve performance. This
indicates that the accuracy of city managers self-ratings will
have important implications for local governments, affecting the
performance of these key local government administrators.
Due to the unreliability of difference scores (Johns,
1981, Atwater & Yammarino, 1992), we did not utilize raw
scores for statistical analysis. Instead, these differences were
utilized to categorize individuals’ perceptions of their leadership
capabilities as accurately estimated, over-estimated, and underestimated. Individuals were assigned to these categories based
on the difference of their self-other evaluations. The standard
deviation of the absolute value of these scores was 4.34.
Therefore, individuals with a self-other difference score greater
than 4.34 are categorized as over-estimators, individuals with a
self-other difference score below -4.34 are categorized as underestimators, and individuals between -4.34 and 4.34 are
categorized as accurate self-assessors. Membership in these
categories will have implications for manager’s abilities to
manage their organizations, influencing their reaction to
feedback and employees’ responsiveness (Atwater &
Yammarino, 1992, Bass, 1990; Ashford, 1989).
The three categories generated from our difference
scores allow us to assess three types of managers, utilizing
binomial logistic regressions to predict membership in each
category. The first category examines managers who are aware
of their leadership credibility in order to determine which
variables are associated these managers. The second accounts
for managers who over-estimate their credibility, while the third
measure accounts for managers who under estimate their
credibility. We include these three types of managers as we
speculate that different factors will lead managers to accurately
assess, over-estimate or under-estimate their credibility.
Certainly many factors could influence differences
between subordinates’ and supervisors’ assessments of
leadership credibility. We rely on individual, environmental,
and organizational characteristics to explain these differences.
!
556
PAQ WINTER 2014
Previous studies of leadership traits demonstrate that individual
characteristics can explain variations in leadership qualities
(Atwater & Yammarino, 1992). Following in this tradition, we
utilize city managers’ age, tenure, education and gender to
predict these variations.8 Tenure is measured using the number
of years the manager has served in their current position. Age is
a numerical response to an open ended question regarding the
respondent’s age in years. Education is measured using an
ordinal scale, with respondents indicating their highest level of
educational attainment.
Finally, gender was coded as a
dichotomous indicator where a value of 0 was given to females
and 1 was given to males.
We utilize two organizational characteristics in order to
measure the organization’s culture. These variables examine the
change motivation of employees and the degree to which
employees identify with the entire organization (rather than their
specific work group). Once again, survey responses provided
the data for these variables. These variables are measured on
five point scales and averaged across responding department
heads in order to create an organizational average.
In order to assess the environmental context in which
these managers operate, we also included municipal level
measures of population density and population homogeneity.
These variables rely on US Census data: population density is
operationalized as residents per square mile, whereas
homogeneity by the percent of the population comprised of
Caucasians.
RESULTS
A brief examination of the three categories of city
managers reveals interesting differences between the three
categories of self-assessors. Table 2 includes the means and
standard deviations of each of the variables representing these
categories.
As this table indicates, over-estimators rated themselves
higher than any other group, but their employees’ assessments
8
!
See Appendix A for survey items used to create the independent variables.
PAQ WINTER 2014
557
were the lowest of any of our groups of managers. Accurate
estimators were rated higher than any other group by their
employees, while under-estimators’ self-ratings varied greatly
from their subordinates’ assessments. Our findings indicate that
over-estimators’ self-ratings were not solely responsible for their
classification, as they were also rated lower than other groups by
their subordinates.9 Table 2 also displays department heads’
ratings of city managers’ effectiveness when dealing with their
city councils. As with department heads’ assessments of
leadership credibility, city managers who accurately estimated
their credibility received higher ratings on this measure when
compared to either under-estimators or over-estimators, while
over-estimators received lower ratings than any other category.
This underscores the potential import of accurate selfperceptions.
To further examine leadership credibility, we utilized
binomial logistic regressions to examine the associations
between individual and organizational characteristics and the
accuracy of manager self-ratings.
The three categories
generated from our difference scores allow us to assess three
types of managers in order to determine which variables are
associated with the accuracy of managers’ self-perceptions. The
first category represents accurate self-assessment. The second
accounts for over-estimators, while the third measure accounts
for under-estimating managers. We examine each of these three
types of managers as we speculate that different factors will lead
managers to accurately assess, over-estimate or under-estimate
their credibility. To this end, we utilize three models, one for
each of our three self-evaluation types. The models examining
accurate and over-estimators assessors of their credibility
identify many factors associated with these traits, while our
model for under-estimators does not perform well.
The
regression models predict the likelihood of a manager’s
membership in each category and can be found in Table 3.
9
This echoes the findings of Bass and Yammarino (1991), as well as Atwater
and Yammarino’s study of transformational leadership (1992).
!
558
PAQ WINTER 2014
Table 2
Evaluations by Manager Type.
Manager Type
Credibility
Accurate Estimators
(N=20)
Self-Assessment
31.00
1.94
Other-Assessment
29.35
2.51
4.03
.56
Self-Assessment
34.10
4.45
Other-Assessment
26.10
4.745
3.60
.93
Self-Assessment
18.25
7.27
Other-Assessment
28.28
6.58
3.95
.90
Effectiveness Dealing with
City Council
Over-Estimators
(N=18)
Effectiveness Dealing with
City Council
Under-Estimators
(N=4)
Effectiveness Dealing with
City Council
Mean
Standard
Deviation
The findings of the first model indicate that many of the
independent variables are associated with accurate selfassessment. The positive and significant coefficient for manager
tenure indicates that city manager who have served longer in
their positions are more likely to be aware of their credibility, as
expected (Atwater & Yammarino, 1992; Ostroff, Atwater, &
Feinberg, 2004).
Predicted probabilities demonstrate the
substantive significance of this coefficient. For example, a city
manager with five years of experience in the position has only a
5.7% chance of accurately estimating their credibility, a manager
with 10 year of experience a 22.9% chance, and a manager with
15 years of experience has a 58.8% chance to be an accurate
estimator.10
10
Holding all other independent variables at their mean values or median
categories.
!
PAQ WINTER 2014
559
Table 3
Predicting Accurate Perceptions of Leadership Credibility.
Accurate
Assessment Model
Variable
Coefficient
(odds ratio)
Pvalue
Over-Estimators
Model
Under-Estimators
Model
Coefficient
(odds ratio)
Pvalue
Coefficient
(odds ratio)
P-value
Manager
Tenure
.314**
(1.369)
.018
-.246**
(.782)
.027
.314
(ns)
.991
Manager Age
-.353**
(.702)
.022
.224**
(1.252)
.022
-.353
(ns)
.992
Manager
Education
-2.62**
(.072)
.045
2.402**
(11.047)
.038
-2.62
(ns)
.991
Manager
Gender
-5.55*
(.004)
.054
2.873
(17.703)
.192
dropped
(ns)
dropped
Population
Density
-.001
(.999)
.129
.001**
(1.001)
.048
-.001
(ns)
.993
Percent
White
-.382**
(.682)
.018
.440***
(1.553)
.008
-.382
(ns)
.991
Employee
Change
Motivation
Employees
Identification
with
Organization
Model Fit
.386*
(1.472)
.070
-.382*
(.682)
.056
.386
(ns)
998
-1.96
(.141)
.121
.873
(2.393)
.421
-1.96
(ns)
.991
Prob.
>Chi2
Adj. R2
.000
Prob. >Chi2
.000
Prob. >Chi2
.000
2
.521
2
.561
1.00
N
40
N
40
N
40
Adj. R
Adj. R
*pChi2
.000
2
.521
2
.561
1.00
N
40
N
40
N
40
Adj. R
Adj. R
*p
Purchase answer to see full
attachment