 

 

 

 

 

 

 

 

 

 

 

 

3
.5" ..
JV... 9

t.

w 4:

 

 

 

 

 

)9
n , a 1-

6%

«Car...

$33.32 .
.. i

4.. My; “mm.“
a“..».._..§, “33%
$35.11...
. . .wnarmai

n!
h

{RE .

 

11111315

 

(2&0:
52/72 9/40
LIBRARY
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

GENDER STEREOTYPES AND BSD-DEGREE LEADERSHIP
FEEDBACK RATINGS: AN APPLICATION OF
ROLE CONGRUITY THEORY

presented by

CORI A. DAVIS

has been accepted towards fulﬁllment
of the requirements for the

Ph.D. degree in Psychology

 

 

(/ Major Professor’s Signature
44:}4/‘9/‘22 j"OCH;

Date

MSU is an Afﬁrmative Action/Equal Opportunity Institution

 

 

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65-p.15

 

GENDER STEREOTYPES AND 360-DEGREE LEADERSHIP FEEDBACK
RATINGS: AN APPLICATION OF ROLE CONGRUITY THEORY

By

Cori A. Davis

A DISSERTATION

Submitted to
Michigan State University
in partial ﬁJlﬁIIment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Psychology

2003

ABSTRACT

GENDER STEREOTYPES AND 360-DEGREE LEADERSHIP FEEDBACK
RATINGS: AN APPLICATION OF ROLE CONGRUITY THEORY

By

Cori A. Davis

This study examined gender differences in a 360-degree leadership development
assessment. First, consistent with Role Congruity Theory (Eagly & Karau, 2002), the
. context of the work environment was expected to impact the relationship between leader
gender and leader ratings such that female leaders would be rated as being less effective
. and needing more development than their male counterparts in male-dominated work
contexts. The types of context variables explored were the level of male domination of
the industry and ﬁmctional areas within which the leaders worked, and the proportion of
male raters for each leader. Second, gender differences in ratings were expected for
different types of skills being assessed. In this study, both communal and agentic
leadership skills were included in the 360 instrument. It was expected that women would
generally receive ratings indicating higher effectiveness and less need for development
with regard to communal, as opposed to agentic, leadership skills. Finally, the impacts of
rater gender, rater level relative to the leader (i.e. boss, peer, subordinate), and leader
level on leadership ratings (including self-ratings) were explored.

No gender differences were found for any type of rating: overall leadership
effectiveness, agentic development needs, or communal development needs. Further,
gender differences in ratings were not impacted by the organizational context variables
included in this study, thereby failing to support Role Congruity Theory. Finally, no

gender differences were found as a function of rater gender, rater level, or leader level.

H'

.\Ii

 

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................... v—i
LIST OF FIGURES ................................. ' ...................................................................... vii
INTRODUCTION ........................................................................................................... 1
LITERATURE REVIEW ................................................................................................ 7
Gender Stereotype Theories ................................................................................. 7

Social Role Theory ................................................................................... 7

Lack ofF it Model ..................................................................................... 9

Role Congruity Theory ........................................................................... 11

Changing Nature of Leadership .......................................................................... 17

360 Degree Leadership Development Assessments ............................................ 23
Summary ........................................................................................................... 25
HYPOTI—[ESES .................................................................. . .......................................... 28
Work Context .................................................................................................... 29
Industry .................................................................................................. 31

Functional Area ...................................................................................... 32

Rater Gender Composition ..................................................................... 34

Types of Ratings ................................................................................................ 36
Development vs. Effectiveness Items ..................................................... 36

Communal vs. Agentic Development Items ............................................ 38
Organizational Level .......................................................................................... 4O
Leader/Rater Relationship ...................................................................... 41

Leader Level .......................................................................................... 50

METHODS .......................................................................................... . ........................ 53
Design ............................................................................................................... 53
Participants ........................................................................................................ 54
Measures ........................................................................................................... 57
Denison Leadership Development Survey .............................................. 57

Leadership Effectiveness Items .............................................................. 6]
Organizational Level .............................................................................. 62

Functional Area ...................................................................................... 63

Industry Type ......................................................................................... 64

Proportion of Male Raters ...................................................................... 66

Analyses ............................................................................................................ 67

HLM .................................................................................... . ................. 67

iii

DI

RE}:

 

TABLE OF CONTENTS (CONTINUED)

RESULTS ......................................................................... . ........................................... 72
Overview ........................................................................................................... 72
General Descriptives .......................................................................................... 72
Main Effects ...................................................................................................... 75

Leader Gender Main Effects ................................................................... 76
Leader Level Main Effects ........ . ............................................................ 76
Rater Gender Main Effects ..................................................................... 76
HLM Null Models ............................................................................................. 77
HLM Model 1: 3-Level, 2-Way Interactions ...................................................... 78
Hypotheses 1, 5a, 6a, 7a, and 8a ............................................................. 78
Hypotheses 1, 5b, 6b, 7b, and 8b ............................................................ 79
HLM Model 2: 2-Level, 2-Way Interactions ...................................................... 79
Hypotheses 3, 5c, 60, 7c, and 8c ............................................................. 80
Hypothesis 12 ..... . ................................................................................... 80
Hypothesis 14 ......................................................................................... 80
HLM Model 3: 2-Level, 2-Way Interactions ...................................................... 81
Hypothesis 4 .......................................................................................... 81
Hypotheses lOa-b ................................................................................... 81
HLM Model 4: 2-Level, 3-Way Interaction ........................................................ 82
Hypothesis 9 .......................................................................................... 82
Analysis of Variance Tests ................................................................................. 82
Hypothesis 10c ....................................................................................... 82
Hypothesis ll ......................................................................................... 83
Hypothesis l3 ......................................................................................... 83

DISCUSSION ............................................................................................................... 85

Findings ............................................................................................................. 87
Main Effects ........................................................................................... 89
Descriptive Data ..................................................................................... 90
Context Effects ....................................................................................... 91
Rater Effects .......................................................................................... 93

Limitations ........................................................................................................ 94
Nature of Rating Dimensidns ................................................................. 94
Leader Characteristics ............................................................................ 96
Standards of Performance ....................................................................... 96

Future Research ................................................................................................. 97
Promotion Criteria .................................................................................. 98
Development Opportunities .................................................................... 99
Changing Nature of Leadership ............................................................ 102

Conclusion ....................................................................................................... 103

REFERENCES .................................................................................. . ......................... 105

TABLE OF CONTENTS (CONTINUED)

APPENDIX A

Denison Leadership Development Survey Demographics Questions ............................. 166
APPENDIX B

Procedure for Administering and Reporting the DLDS .................................................. 167
APPENDIX C

Description of Denison Leadership Development Survey Items .................................... 169
APPENDIX D

Communal and Agentic Development Items ................................................................... 175
APPENDIX E

Denison Leadership Development Survey Effectiveness Items ...................................... 177
APPENDIX E

Denison Leadership Development Survey Organizational Level Options ...................... 178
APPENDIX G

Denison Leadership Development Survey Functional Area Options and '
Corresponding DOC Codes ............................................................................................. 179

LIST OF TABLES

Table 1. Number of Raters for Male and Female Leaders ............................................ 121

Table 2. Average Number of Raters Per Leader ........................................................... 122
Table 3. Leader and Rater Demographics .................................................................... 123

Table 4. Denison Leadership Development Survey Skills, Dimensions, and

Subdimensions ............................................................................................................ 125
Table 5. Denison Leadership Development Survey Items ............................................ 126
Table 6. Agentic and Communal Development Item Ratings ....................................... 128
Table 7. Number of Leaders in Each Organizational Level .......................................... 132
Table 8. Number of Leaders in Each Functional Area ................................................. 133
Table 9. Number of Leaders in Functional Area Percent-Female Ranges .................... 134
Table 10. Number of Leaders in Each Major Industry ............................. . ................... 135
Table 1 1. Number of Leaders in Industry Percent-Female Ranges .............................. 136
Table 12. Number of Leaders in Proportion of Male Raters Ranges ............................ 137
Table 13. Summary of Study Hypotheses, Analyses, and Results ............................... 138
Table 14. Means and Standard Deviations for Male and Female Leaders ..................... 146

Table 15. Summary of HLM Descriptive Main Effects of Leader Gender
and Leader Level ......................................................................................................... 147

Table 16. Summary of Analysis of Variance Descriptive Main Effects of
Leader Gender, Leader Level, and Rater Gender ......................................................... 148

Table 17. Results of the Null Model: Testing for Between Group Variance ................ 149

Table 18. Summary of Results for 3-Level, 2-Way Interaction HLM Analyses:
Hypotheses 1, 2, 5a-b, 6a—b, 7a-b, and 8a-b ................................................................. 152

vi

LIST or TABLES (CONTINUED)

Table 19. Summary of Results for 2-Level, 2-Way Interaction HLM Analyses:

Hypotheses 3, 5c, 6c, 7c, 8c, 12, and 14 ....................................................................... 154
Table 20. Summary of Results for 2-Level, 2-Way Interaction HLM Analyses:
Hypotheses 4 and lOa-lOb ..................... . ..................................................................... 157
Table 21. Summary of Results for 2-Level, 3-Way Interaction HLM Analysis:
Hypothesis 9 ................................................................................................................ 158
Table 22. Analysis of Variance Results for Hypotheses 10c, l l, and 13 ...................... 159

vii

LIST OF FIGURES

Figure 1. Denison Leadership Development Model .................................................... 160
Figure 2. Example 3-Level, 2-Way Interaction HLM Model for
Hypotheses 1, 2, 5a-b, 6a-b, 7a-b, and 8a—b (Taken from Hypothesis 1) ....................... 161
Figure 3. Example 2-Level, 2-Way Interaction HLM Model for
Hypotheses 3, 5c, 6c, 7c, 8c, 12, and 14 (Taken from Hypothesis 3) ............................ 162
Figure 4. Example 2-Level, 2-Way Interaction HLM Model for
Hypotheses 4 and lOa-b (Taken from Hypothesis 4) .......................... . ......................... 163
Figure 5. 2-Level, 3-Way Interaction HLM Model for Hypotheses 9 .......................... 164
Figure 6. Example 2 and 3-Level HLM Null Models ................................................... 165

viii

INTRODUCTION

Over the past three decades, the percentage of female managers in the workforce
has increased from 18% to 45% (U. S. Bureau of Labor Statistics, 2001a). In addition,
women are earning 56% of Bachelors degrees and 45% of advanced degrees including
42% of PhDs and 43% of professional degrees (e. g. law and medicine, U. S. Bureau of
the Census, 2000). Yet, despite the growing number of educated women in the
workforce, women constitute only 4% of the top officers in Fortune 500 companies, 3%
of the most highly paid officers, and 0.4% of CEOs (Catalyst, 2000).

The “pipeline” problem is a common response to the huge disparity between the
number of educated, working women and the number of women in leadership positions.
The pipeline explanation states that there are simply not enough women with the
appropriate education and background to fill top leadership positions. However, this
explanation is obviously inconsistent with the statistics on the current number of
educated, female managers in the United States (Carli & Eagly, 2001).

Heilman (1995) also suggests that the pipeline explanation is insufficient. She
notes that research indicates that women who moved into management in the 1970’s have
not progressed through the ranks as quickly as males who entered management at around
the same time. In fact, a report by the Feminist Majority Foundation states that at the
present rate of advancement, it will take until the year 2465 for women to be as equally
represented as men in executive suites (1991', cited in Morrison, White, & Van Velsor,
1992) despite research evidence that there are few, if any, actual gender differences in
qualiﬁcations to lead (see Dobbins & Platz, 1986; Morrison & Von Glinow, 1990,

Powell, 1993', Stroh, Brett, & Reilly, 1992; Cleveland, Stockdale, & Murphy, 2000).

This indicates that despite progression in education and movement to management
positions, women are still experiencing problems scaling the hierarchy (Bartol, 1999).
This holds true even for women in female-dominated ﬁelds. In fact, some research has
shown that men are promoted even more quickly in female-dominated as opposed to in
male-dominated ﬁelds (eg. Maume, 1999; Ott, 1989; Williams, 1992, 1995; Yoder &
Sinnett, 1985).

If the pipeline explanation cannot explain why there are so few women in
leadership positions, what can? There have been many explanations proposed in
response to this question. Kaufmann, Isaksen, and Lauer (1996), for example, tested
three different explanations for this phenomenon which they called Experience Theory,
Personality Theory, and Stereotype Theory. The ﬁrst two theories state that women
simply have not yet built up the qualifying experiences or do not possess the traits
necessary to becoming successful leaders. The last theory, Stereotype Theory, states that
the glass ceiling is product of attitudes toward women. They found the most support for
Stereotype theory.

In fact, most explanations for why more women are not in the executive suite
stem from Stereotype Theory. The theory, when applied to gender, generally states that
individuals have a preconceived notion about what actions are appropriate for men and
women. Although some may argue that gender stereotypes are diminishing, particularly
in the workplace, recent Gallop Poll responses conﬁrm that stereotypical attitudes
regarding leadership and gender still exist. For instance, when it comes to choosing a

boss, Americans still prefer men (Simmons, 2001: January 1 1). In addition, Americans

see women as emotional and affectionate, and men as more aggressive (Newport, 2001:
February 21).

Powell, Butterﬁeld, and Parent (2002) performed a recent study which assessed
attitudes about women managers. They gathered data from 348 undergraduate and part-
time graduate business students. They found that although stereotypes about managers
place less emphasis on masculine characteristics than in earlier studies, a good manager is
still perceived as having a predominately masculine skill set (i.e. autocratic, decisive,
assertive, etc.) (Powell et al., 2002).

Heilman (2001) has noted that the glass ceiling is a natural consequence of gender
stereotypes and the subsequent expectations stereotypes produce about what women are
like and how they should act. She also proposed that stereotypes in the workplace often
manifest in evaluation bias. She claims that bias in performance evaluation is the
primary cause for the glass ceiling because when women receive inadequate leadership
performance ratings, they are less likely to be selected for leadership positions. Bartol
(1999) also noted that an important area of study relating to the advancement of women
in organizations is whether there are systematic gender differences in performance
evaluations. Unfortunately, there have been few consistent ﬁndings with regard to the
direct effect of leader gender on performance ratings.

Several laboratory studies have suggested that ratee gender has little to no
inﬂuence on performance evaluations (e.g., Frank & Drucker, 1977; Giannantonio, Olian,
& Carroll, 1995; Hall & Hall, 1976), whereas others have found signiﬁcant gender effects
sometimes in favor of women (e.g., Bigoness, 1976; Hamner, Kim, Baird, & Bigoness,

1974), and sometimes in favor of men (e.g., Pazy, 1986; Woehr & Roch, 1996). As with

Ct

 

(it)

as:

de

E135

these lab studies, field studies have not led to a clear conclusion about the impact of
leader gender on performance evaluations.

Bartol (1999) has noted that there are relatively few ﬁeld studies considering ratee
gender effects in official performance appraisals, particularly at the managerial levels.
Some ﬁeld studies found that men were evaluated more positively than women (e.g.
Deaux, 1979, Eagley, Makhijani, & Klonsky, 1992, Pulakos, White, Oppler, & Borman,
1989), others found that women were rated more highly than were men (e. g.Bennett,
1982', Mobley, 1982', Pulakos & Wexley, 1983), and still others found equal ratings
between men and women (e. g. Kirnan, Farley, & Geisinger, 1989', Yammarino &
Dubinsky, 1988). The inconsistent results of these studies may suggest that more
complex issues are involved (Bartol, 1999).

Some researchers have noted that gender differences in performance evaluation
do not always explain the disparity between men and women in the top ranks. There is
also evidence that female employees receive less developmental opportunities than their
male counterparts. For instance, females are less likely to have mentors, and more likely
to experience barriers to gaining a mentor (Ragins & Cotton, 1991). In addition, women
are less likely to be offered challenging experiences that stretch their skills (Lyness,
2002). Lyness (2002) notes that high-quality organizational career management and
assessment systems, including coaching and feedback, can help to offset some of the
developmental disadvantages women face. However, if bias exists in these sorts of
assessment systems, then they are not serving their purpose.

Given the already disparate developmental opportunities afforded women, it is

important to understand how gender stereotypes impact current leadership development

fie

systems. An increasing number of organizations are moving toward a leadership
development model that includes some form of multisource performance assessment,
often referred to as 360-degree feedback (Atwater & Waldman, 1998, Becker, Ayman, &
Korabik, 2002; London & Smither, 1995). Such systems usually incorporate
performance and deveIOpment evaluation feedback from a variety of sources, including
subordinates, peers, and bosses.

Antonioni (1996) reported that approximately 25% of companies use some type
of multirater or 360 degree feedback process. One of the reasons for its increased
popularity is the belief that developmental feedback will improve leadership
performance. Studies have, in fact, found improvements in overall performance as a
result of 360 feedback (e.g. Atwater, Roush, & Fischthal, 1995', Reilly, Smither, &
Vasilopoulos, 1996).

Although there is a growing body of literature on 360 feedback, little is known
about whether gender differences that are said to occur in performance appraisals carry
over into 360 developmental feedback assessments. Do these same differences exist
when individuals receive negative 360 feedback that is conﬁdential, developmental, and
not used for reward or promotion decisions? In addition, relatively little is known
regarding gender effects related to evaluations provided by sources other than supervisors
within a 360 feedback system (Bartol, 1999).

Given that a 360 assessment is essentially an informal performance evaluation
from every type of person, using this type of developmental tool raises potential
questions about possible gender bias. There is an opportunity here to explore how past

ﬁndings regarding gender differences in performance evaluation can be generalized to

360 assessment systems. In addition, we need to explore what unique questions need to
be answered regarding the role of stereotypes in the context of a 360 feedback system.
Therefore, the purpose of this research is to explore the theories which help to explain
gender differences in performance evaluation and generalize and test those theories in the
context of a 360 leadership development tool. The next few sections review Gender
Stereotype theories more speciﬁcally as well as the major ﬁndings in the ﬁeld of gender
and leadership. Hypotheses are then offered relevant to the theories and ﬁndings

presented.

LITERATURE REVIEW

Gender Stereotype Theories

 

There are basically two forms of thought surrounding the origin of stereotypes.
There is the biological view and the social view. Those who support the biological view
claim that the differences between men and women are a byproduct of evolution and are
rooted in physiological characteristics that have evolved over time. Alternatively, those
supporting the social perspective support the claim that differences between men and
women stem from societal values and subsequent manufactured gender roles. With
regard to examining stereotypes in the workplace, the social perspective has received the
most attention and is most relevant for this study.

Social Role Theory. Some claim that the workplace is an extension of our social
environment, and therefore spillover effects of male/female social stereotypes affect how
men and women are treated at work (Gutek & Morasch, 1982). Eagly and colleagues
have made the strongest arguments for this viewpoint using Social Role Theory. Social
Role Theory suggests that gender stereotypes are similar to other social stereotypes in
that they reﬂect “perceivers’ observations of what people do in daily life” (Eagly &
Steffen, 1984; p. 735). In other words, because we observe particular groups of peOple
performing speciﬁc behaviors, we are prone to believing that the abilities and personal
qualities required for performing those behaviors are typical of those groups. A key
proposition of social role theory is that most of the behaviors we ascribe to men and
women fall into two categories: agentic and communal (Bakan, 1966; Eagly, 1987). Men
are stereotypically seen as displaying agentic behaviors, and women are seen as

displaying communal behaviors.

 

9?.

Pie

Ma.

 

Communal characteristics generally describe a concern with the welfare of other
people or as social or service-oriented. Examples include being helpful, affectionate,
kind, sympathetic, sensitive, gentle, and nurturing. Agentic characteristics generally
describe an achievement-oriented, assertive, controlling, and confident orientation.
Examples include being ambitious, aggressive, dominant, forceful, independent or self-
sufﬁcient, self-conﬁdent, decisive, and prone to act as a leader. (Eagly & Karau, 2002;
Heilman, 2001). Not only are these agentic and communal conceptions of women and
men different, but they are also often oppositional. Members of one sex are seen as
lacking what is considered to be prevalent in members of the other sex (Heilman, 2001).

Heilman (2001) has noted that these agentic and communal social stereotypes
about the attributes of men and women are pervasive, widely shared, and resistant to
change (see Dodge, Gilroy, & Fenzel, 1995; Leuptow, Garovich, & Leuptow, 1995).
Some researchers have found that gender provides the strongest basis for categorizing
people, even when compared with other personal characteristics such as race, age, and
occupation (Fiske, Haslam, & F iske, 1991; Eagly & Karau, 2002', Stangor, Lynch, Duan,
& Glass, 1992; van Kippenberg, van Twuyver, & Pepels, 1994). In addition, some have
noted that stereotypes about women and men are easily and automatically activated
(Banaji & Hardin, 1996; Banaji, Hardin, & Rothman, 1993', Blair & Banaji, 1996; Eagly
& Karau, 2002).

Research also supports the notion that traditional stereotypes of women and men
predominate in work settings as well as in non-work settings. Heilman, Block, and
Martell (1995), for example, showed that even when women are depicted as managers,

they are characterized as less agentic than men. In their study, working managers from a

Sit

€11

al

 

 

Pct

fag

“’0

t‘,
01‘

mo

the

to]E

range of industries described women managers as more competent, active, and potent
than women in general. However, they described women managers as decidedly more
deﬁcient in these same attributes than male managers. Heilman et al. (1995) concluded
that the increased presence of women in the workplace and their assumption of new roles
do not appear to preclude gender-stereotypic perceptions. This led Heilman to propose

the Lack ofFit model (Heilman, 1983, 1995, 2001).

 

Lack of Fit Model. Heilman (1983, 1995, 2001) developed the Lack ofFit Model
to help specify the circumstances in which gender discrimination occurs. Her model
suggests that the way individuals perceive job roles may be a crucial component in
determining the extent to which gender influences work-related outcomes (Cleveland et
al,2000)

In Heilman’s (2001) view, some jobs are more closely associated with gender
stereotypes than others. This may be due to the typical number of males or females
employed in these roles, or to the types of activities performed on the job (Cleveland et
al., 2000). For instance, common stereotypes exist for nurses and police. When most
people think of a nurse, they think of a woman. This may be because most nurses are, in
fact, female. This may also be because nurses care for people, and many think that
women are more likely than men to have care-giving roles. Similarly, when people think
of a police ofﬁcer, they usually picture a man. Again, this stereotype may arise because
most police officers are,iin fact, men. It may also arise because police ofﬁcers enforce
the law, and many think that men are more likely than women to hold an authority-type

role.

Heilman (1983, 1995, 2001) also suggests that certain jobs may carry gender
stereotypes because of personal qualities considered to be necessary for success. With
regard to management and leadership, she noted that success is often associated with
ambitiousness, competitiveness, analytical skills, and interest in power (Cleveland, et al.,
2000). These are all agentically-oriented qualities. Alternatively, women are typically
considered to have an entirely different set of skills including interpersonal skills and
interest in consensus (Cleveland et al., 2000). Although these are also work-related, they
are not skills usually associated with the stereotypic leader.

The lack of ﬁt model implies that the role of a leader may be problematic for
women because the schemas, or preconceived ideas, that people have regarding
leaders/managers are different from those they have regarding women. There is much
evidence to suggest that the prevailing image of a “leader” is more similar to that of a
man than a woman. This has been found to be true for respondents who are young and ‘
old, male and female, and is generalizable across culture and time (Becker etal., 2002;
Schein, 1973, Schein, Mueller, & Jacobson, 1989; Schein, Mueller, Lituchy, & Liu,
1996). A woman in a leadership role likely activates two conflicting schemas: a
feminine, or communal schema associated with her gender and a masculine, or agentic
schema associated with leader role (Becker et al., 2002).

Beginning in the 70’s, researchers have repeatedly shown that many traits
associated with leadership (e. g. competitiveness, self-conﬁdence, and desire for
responsibility) are seen as male-oriented or agentic traits (Brenner, Tomkiewicz, &
Schein, I989; Heilman, Block, Martell, & Simon, 1989', Powell & Butterﬁeld, 1979,

1989, 1994; Schein, 1973, 1975). For example, in early studies by Schein (1973, 1975),

10

male and female managers gave their impression of either women, men, or successful
middle managers. The participants perceived successful middle managers as being
signiﬁcantly more similar to men than women. This was particularly true for a number of
agentically-oriented characteristics such as competitiveness, self-conﬁdence, objectivity,
aggressiveness, and ambitiousness. Other researchers have replicated Schein’s (1973,
1975) findings in the United States (e.g. Heilman et al., 1989; Massengill & di Marco,
1979) as well as in the United Kingdom, Germany, Japan, China, and Singapore (Lee &
Hoon, 1993', Schein, 2001). These studies show that the typical perceptions of the
leadership role overlaps with the male stereotype (Cleveland et al., 2000).

Based on the incongruence between the stereotype of a leader and the stereotype
ofa woman, Heilman (I983, 1995) suggested that gender discrimination is most likely to
occur when the qualities of the person do not ﬁt with the stereotype of the role (Cleveland
et al., 2000). Heilman suggested that this effect should hold true for both women and
men. In other words, both women in masculine roles and men in feminine roles are more
likely to experience gender discrimination than women in feminine roles and men in
masculine roles, respectively. However, wage surveys suggest that women are more
likely than men to be penalized in sex-role incongruent situations (Cleveland et al., 2000',
Renner, Rives, & Bowlin, 2002).

Role Congruity Theory. Recently, Eagly and Karau (2002) presented their Role
Congruity Theory which builds on the ideas in Social Role Theory and the Lack of Fit
Model. The theory itself integrates the social psychology perspectives regarding
stereotyping with the industrial/organizational psychology perspectives on perceptions of

leadership and management roles.

There is an emphasis in Role Congruity Theory on the “gender” role as opposed
to the “social” role discussed in Social Role Theory. Where social roles are socially
shared expectations applied to persons belonging to a certain social position or social
category (Biddle, 1979', Sarbin & Allen, 1968), gender roles are consensual beliefs about
the attributes of men and women (Eagly & Karau, 2002). Role Congruity Theory
contrasts with classic stereotype theories which portray prejudice as arising from
unfavorable stereotypes toward a social group (e.g., Allport, 1954; Esses, Haddock, &
Zanna, 1993). From that perspective, the lack of women in top leadership positions
would be attributed to a negative stereotype toward women in general. However, such a
context-free theory could not explain why there is bias against women in some roles and
not others (Eagly & Karau, 2002).

The basic philosophy behind Role Congruity Theory, then, is that women in
organizational leadership roles are perceived as being role incongruent because the role
of a leader is inherently agentic. Eagly and Karau (2002) claim that “the potential for
prejudice exists when social perceivers hold a stereotype about a social group that is
incongruent with the attributes that are thought to be required for success in certain
classes of social roles. ‘When a stereotyped group member and an incongruent social role
become joined in the mind of the perceiver, this inconsistency lowers the evaluation of
the group member as an actual or potential occupant of the role” (Eagly & Karau, 2002;
p. 574). Therefore, Role Congruity Theory suggests that bias toward female leaders
generally follows from the incongruity that individuals perceive between the

characteristics of women and the requirements of a leadership role.

 

 

 

 

\‘x

am-

Car

 

According to Role Congruity Theory, the perceived incongruity between
leadership roles and the female gender role leads to two types of bias against female
leaders. First, individuals may perceive women less favorably than men as potential
occupants of leadership roles. Second, individuals may evaluate behavior that fulfills the
prescriptions of a leader role less favorably when these behaviors are performed by a
woman. Consequences of these forms of bias include less favorable attitudes toward
female as opposed to male leaders, and increased difficulty for women trying to advance
into leadership positions. In presenting their Role Congruity Theory, Eagly and Karau
(2002) reviewed evidence from various studies that substantiates these proposed
consequences.

One of the main ﬁndings Eagly and Karau (2002) reported is that context
variables of the work environment which make gender more salient are likely to enhance
biases against female leaders. Speciﬁcally, women working in male-dominated settings
will experience more bias. There are some contexts, at multiple levels of analysis, that
have been shown to be important in affecting female leader biases. Three of these
contexts include industry, job functional area, and gender composition of evaluators.
These context variables are being described in the current study for a few reasons. First,
“they each represent different levels of interest. Industry can be investigated across
organizations, functional area can be investigated both across and within organizations,
and the gender composition of evaluators is speciﬁc to each individual. The argument
can be made that even if women are working in agentic industries, they still may be

working in a communal job role mainly surrounded by other women. Second, no single

 

it:
til

Ci.

Ex;

IEL

Jen

 

study has ever included multiple context variables at varying levels of analysis to
determine which context variable has the most impact on female rating bias.

With regard to industry, Scandura (1991) performed a study including 176
female executives from service industries, manufacturing, government, retail industries,
and small businesses. Eighty percent ofthe women in this sample came from small
companies and most were in service and retail industries. Thus, Scandura (1991)
concluded that the glass-ceiling breakers were unlikely to come from the largest, most
powerful organizations and were more likely to come from female-dominated industries.
This study was limited in the number of industries explored and provided no information
about the possible mechanisms preventing women in large companies and/or male
dominated industries from reaching the top leadership positions.

With regard to functional area, research has found that males are more likely to
become leaders when the tasks being performed are male-oriented (e. g. assembling
mechanical components) (Butterﬁeld & Grinnell, I999, Carbonell, 1984, Megargee,
1969; Nyquist & Spence, 1986). Studies from the Goldberg research paradigm also
tested whether women suffer disadvantages relative to men. The Goldberg studies
typically consisted of experiments in which participants evaluate men and women whose
characteristics have been set equal. This method of examining potential bias against
women has been labeled the Goldberg Paradigm in honor of Goldberg’s (1968) initial
experiment of this nature (Eagly & Karau, 2002). A typical Goldberg study would
involve presenting a resume or application for evaluation, with half of the participants
receiving it with a female name and half receiving it with a male name (e. g. Rosen &

Jerdee, 1974). In a thorough meta-analysis of a subset of Goldberg studies, Davison and

14

 

Iii.

SL1.

P’s:

of,

0fL

(B.

Burke (2000) showed that men were preferred over women for male gender-typed jobs
and women were preferred over men for female gender-typed jobs. Given that leadership
roles are usually gender-typed as masculine, this line of research supports Role Congruity
Theory’s prediction of bias against female candidates for leadership positions (Eagly &
Karau, 2002).

The ﬁnal context variable explored in this study is the gender composition of
evaluators. Many of the researchers who have reported rater gender composition effects
describe the composition of raters as a context variable. For instance, Bowen, Swim, and
Jacobs (2000) found little evidence of overall gender bias in performance appraisals in
ﬁeld studies, however found signiﬁcant pro-male biases when only men served as raters.
Eagly and colleagues also found that men were perceived as being more effective than
women to the extent that leader and subordinate roles were male-dominated numerically
(Eagly, Karau, & Makhijani, 1995). In addition, other research has shown that female
leaders are seen as being less effective relative to male leaders as the proportion of male
subordinates increased (Eagly & Karau, 2002).

In presenting the importance of context in their Role Congruity Theory, Eagly and
Karau (2002) reviewed ﬁndings from both ﬁeld and lab studies. However, they chose to
focus on lab studies because lab studies provide more control over contextual factors
such as job tenure, interpersonal relationships between leaders and raters, and job
performance (Eagly & Karau, 2002). However, although not discounting the importance
of lab studies in gender-work research, other researchers have emphasized the importance
of examining ﬁeld data, especially when exploring issues of organizational context

(Bartol, 1999).

15

1h

 

pe

.c
U:

‘3

1

be

x‘,‘ i

the

Researchers have found that particularly in the study of gender, ﬁndings differ
between lab studies and ﬁeld studies such that gender effects are more likely to be found
in laboratory settings (Dobbins & Platz, I986; Osborn & Vicars, 1976). Lab studies
almost always present participants with paper people or a confederate team leader with
whom they have had little experience. Some researchers have criticized the lack of
context, and therefore external validity, of such studies and have pointed to the need for
more ﬁeld research in this area (Bartol, 1999; Eagly & Karau, 2002).

Although there have been some field studies in the gender-work arena, many are
limited by the samples they use. Many include middle managers only from a single
functional area or industry. Martell, Parker, Emrich, and Crawford (1998) also noted that
almost every study of sex stereotyping in the workplace has looked at characterizations of
men and women for entry and middle levels of management, with little attention to
whether those in middle management positions are perceived differently on the attributes
associated with successful executives.

Bartol has also commented on the need for ﬁeld research in this area. She noted
that “the limited data that exist on the link between gender of rater or gender of ratee and
performance evaluations in real and simulated workplaces suggest that simply using
gender as a predictor does not capture the complexities of the area of inquiry” (Bartol,
1999; p. 173). She also noted that the number of studies conducted in ﬁeld settings has
been small and has largely concentrated on positions near the bottom of the hierarchy
where women have experienced more success in being promoted. In addition, she notes
that it has usually been necessary to include a broad mix of positions to attain adequate

sample sizes (Bartol, 1999).

I6

Finally, Ragins (1991) supported the view that the research setting is one artifact
that contributes to inconsistent ﬁndings regarding gender effects, particularly with regard
to subordinate evaluations of leaders. She notes that ﬁeld studies may lead to more
realistic leadership appraisals given that raters have multiple opportunities to observe
leaders’ behaviors and compare leaders with others in the organization.

In sum, Role Congruity Theory is derived from Social Role Theory and the Lack
of Fit Model. It speaks to gender stereotypes as they pertain to leadership roles in
organizational settings. Role Congruity Theory emphasizes the effect of context
variables on the relationship between leader gender and performance evaluation.
Although there is some support for Role Congruity Theory, there are few ﬁeld studies
which have successfully shown the effects of context variables such as industry,
ﬁJnctional area, and gender composition of raters. Also, few studies have examined
leader ratings in a 360 paradigm which includes evaluation of development needs. The
next few sections discuss the Changing Nature of Leadership and the Use of 360
Leadership Assessments. These sections will help us to understand how current
leadership theory and past ﬁndings regarding 360 assessments can help us to further
determine the circumstances in which male and female leadership ratings may differ.

ChangingNature of Leadership

 

Although Role Congruity Theory addresses the research pertaining to context
effects on the relationship between leader gender and evaluation ratings, it does not
speciﬁcally address the nature of the skills being evaluated (i.e. the content of the
evaluation tool). In order to understand what particular skills are valued by raters, it is

important to explore current leadership theory as it pertains to gender.

17

Unfortunately, examining leadership in the organizational domain as it relates to
gender presents a unique problem. Not only does the gender-laden nature of the
organizational context need to be considered, but so does the gender-laden nature of
leadership. In organizational, political, and religious domains, individuals are inundated
with examples of male leaders. The typical male leader is often characterized as being
agentic and therefore ideas about what constitutes good leadership tend to emphasize
stereotypically masculine traits and skills.

Leaders have traditionally been viewed as strong, assertive, take—charge
individuals who more or less dictate followers’ actions. However, more recent trends in
the leadership ﬁeld highlight a mix of desired leadership skills ranging from communal to
agentic. The transformational theory of leadership, particularly, incorporates a blend of
both the agentic and communal leadership skills necessary for effective leadership.
Investigating gender through a transformational leadership lens might allow us to better
understand the types of skills and behaviors emphasized in the leadership ﬁeld. In
addition, the transformational leadership model has become increasingly popular in
recent years and is one of the few models of leadership to be studied in relation to gender
(Butterﬁeld & Grinnell, 1999; Howell & Avolio, 1993; Howell & Higgins, 1990).

Transformational leadership is “the process of influencing major changes in the
attitudes and assumptions of organizational members. . .and building commitment for
major changes in the organization’s objectives and strategies” (Yukl & Van Fleet, 1992,
p. 174). Transformational leadership transcends the daily reward/punishment-based
styles of leadership, and whereas more traditional styles such as transactional leadership

produce incremental changes in the way followers behave, transformational leadership

18

produces fundamental changes in others’ beliefs and attitudes. Many scholars have
argued that transformational leadership is crucial for organizational effectiveness and
adaptability in an age of increasing turbulence and uncertainty because it requires leaders
to provide vision, share leadership (empower others), earn credibility, communicate, and
promote integration and strategizing (Cleveland et al., 2000).

A few researchers have argued that some characteristics of the transformational
leadership model (compared with other models such as transactional leadership) overlap
with the stereotypical communal qualities of women (Maher, 1997). The
transformational model emphasizes democratic relationships, participatory decision
making, delegation, and team-based leadership skills (Eagly & Johnson, 1990). These
characteristics of the transformational leadership model are expected to lead to positive
organizational outcomes. For example, proponents of learning organizations (e. g,
Garvin, 1993', Senge, 1990 ) emphasize supportiveness, participation, and team-based
learning as key elements of organizational effectiveness. Similarly, continuous quality
improvement scientists and practitioners (e.g., Deming, 1986; Juran, 1988) emphasize
building cooperation, developing teamwork skills, building relationships, and
empowering employees in order to improve organizational effectiveness (Eagly & Karau,
2002', Hackman & Wageman, 1995).

There is some evidence that women are more skilled in these aspects of
transformational leadership than men (Denmark, 1993), although the number of studies
examining gender differences in transformational styles of leadership is relatively low
(Cleveland et al., 2000). However, some have studied speciﬁc leadership styles

characteristic of the transformational model such as participative leadership and

I9

I I I I I I I ll l
I. ~. .

 

consideration behaviors. For instance, Eagly and Johnson's (1990) meta-analysis of I62
studies compared the leadership styles of women and men found a tendency for women
to lead in a more participative and interpersonally-oriented style (e.g. concern with
maintaining interpersonal relationships by tending to others’ morale and welfare. . . i.e.
helping, explaining, friendly, available) than men. They concluded that female leaders
may be more interpersonally-oriented and democratic whereas male leaders may be more
task-oriented and autocratic (concern with accomplishing assigned tasks by organizing
task-relevant activities. ..i.e. rules, procedures, high performance, deﬁnite roles).

Despite this evidence that women actually perform some leadership skills better
than men, proficiency at communal-type skills does not guarantee others’ perceptions of
overall effectiveness as a leader. Recent studies in the area of gender and leadership style
have found that “women in performance settings face a catch-22” (Rudman & Glick,
2001', p. 743). Female leaders who display communal qualities are oftentimes liked, but
can be perceived as incompetent. Alternatively, women who display agentic qualities in
order to be perceived as qualiﬁed for leadership roles may be perceived as competent, but
may also suffer social repercussions (Glick, Zion, & Nelson, 1988; Rudman, 1998,
Rudman & Glick, 2001). Research has shown that agentic women are not liked as much
as their agentic male counterparts and are viewed as socially deficient and insufﬁciently
feminine (Rudman & Glick, 2001). Therefore, Rudman and Glick (2001) noted that
women vying for leadership positions face a double bind. They can either display
communal qualities and be seen as nice but incompetent, or they can display agentic
qualities and be seen as competent but socially inept. In both cases, women risk being

disqualiﬁed for leadership roles.

20

 

(it

be

Car

 

An actual example where such a situation occurred was in the Supreme Court
case, Price Waterhouse vs. Hopkins. Hopkins (a woman) was denied partnership in the
ﬁrm despite better qualiﬁcations (e. g. more billable hours, higher customer satisfaction
evaluations, more business sold) than her male counterparts who received the promotion.
Members of her ﬁrm informed her that she was denied a promotion due to interpersonal

problems. When pressed for details, her evaluators let her know that she was too

,3 6‘ 3’ L‘

“macho, needed a course at charm school,”

overcompensated for being a woman,
and might beneﬁt from walking, talking, and dressing more femininely (F iske, Bersoff,
Borgida, Deaux, & Heilman, 1991; p. 339). Hopkins sued her ﬁrm, and the ﬁnal decision
by the Supreme Court supported the notion that Hopkins was denied a leadership role in
the male-dominated ﬁrm expressly based on unfair sex-stereotype expectations. The
judge ruled that she was expected to be more congenial than her male counterparts solely
on the basis that she was a woman (Fiske et a., 1991).

This is an important example because although by all performance accounts,
Hopkins was ﬁt for a leadership position, subjective Opinions about her personal style and
subsequent development needs played a role in Price Waterhouse’s decision to deny her
promotion. Given the discrepancies between men and women with regard to their
ascension into leadership roles, it is important to understand the differences in the
developmental nature of the feedback they are each given. This area has not been
studied.

Many practitioners and researchers alike have pointed out the discrepancies

between the types of developmental opportunities offered to male and female leadership

candidates. Lyness (2002) pointed out that women are less likely to have mentors within

21

their organizations as well as informal social networks. Although not specifically
addressed in this literature, feedback is a ﬁJndamental component of most types of
development opportunities. Other studies have also pointed out the differences in female
and male mentoring relationships, feedback, job assignments, etc. (see Morrison & Von
Glinow, 1990).

The importance of feedback is not unique to the leadership literature, of course.
However, it is not well studied in this area. Given that male and female leadership
candidates are known to be evaluated differently in some cases, and in other cases they
are evaluated similarly but not necessary socially accepted in similar ways, it is not
unreasonable to suspect that men and women may be receiving different forms of
feedback for their development needs as a function of their gender.

In sum, the nature of leadership is changing as organizations move toward models
of continuous learning and improvement. The Transformational Leadership model is
gaining increasing popularity, and has been hypothesized to lead to positive
organizational outcomes. A few studies have shown that some characteristics of
Transformational Leadership are communal in nature, and therefore it is suspected that
women will be perceived as being more proﬁcient in some transformational skills than in
other, more traditional leadership skills. Research and case study ﬁndings suggest that, in
fact, women can sometimes be seen as effective as men when their leadership behaviors
are agentic, however they may also tend to be seen as lacking the communal skills
congruent with their gender role. This raises the question: Do women need to act more
agentically in order to be perceived as effective leaders? If women are given the

appropriate effectiveness evaluations for their performance, is the dislike for agentic

22

women going to influence the types of developmental feedback they receive, primarily
from male followers?

360 Degree Leadership Development Assessments

 

Most 360 degree feedback tools are subjective in nature, meaning that the rater is
not provided with a frame of reference such as behavioral anchors by which to compare
the leader’s performance (Biernat & Fuegen, 2001; Biernat, Manis, & Nelson, 1991;
Biernat & Vescio, 2002). Although procedures vary, typically the leader is rated by
people who interact frequently with him or her, who are knowledgeable about the
leader’s performance, and whose opinions are valued by the leader (Mount, Judge,
Scullen, Sytsma, Hezlett, 1998).

Aside from the use of multiple raters, 360 feedback systems differ from
traditional evaluation methods in a few ways. First, 360 systems are used most
frequently to enhance personal development and growth. Second, ratings provided in 360
systems are made anonymously or conﬁdentially (often with the exception of the
immediate supervisor) and are not accompanied by face to face discussion. The
conﬁdential nature of 360 systems is believed to promote honesty on the part of the raters
and, therefore, be more beneﬁcial to the ratee (Mount et al., 1998).

Multirater systems are believed to have some advantages over traditional
evaluation systems (Hazucha, Hezlett, & Schneider, 1993', London & Beatty, 1993;
London & Smither, 1995'; Tornow, 1993). First, because job performance is
multidimensional, raters other than the supervisor may be better suited to evaluate some
aspects of performance. Second, even if raters have the same opportunities to observe a

leader’s performance, they may perceive it differently and therefore evaluate it

differently. Generally speaking, 360 feedback systems are assumed to provide ratees
with job relevant information that would otherwise not be available (Mount et al., 1998).

Research in 360 degfee assessments has primarily investigated the level of
agreement between self and other ratings (e. g. Atwater, Ostroff, Yammarino, & F leenor,
1998; Johnson & Ferstl, 1999) and, to a lesser degree, method variance (e. g, Mount et
al., 1998), measurement equivalence (e.g., Maurer, Raju, & Collins, 1998), and interrater
reliability (e.g., Greguras & Robie, 1998). However, few studies have explored
differences between various types of raters on different types of skills. In addition, few
studies have examined the differences in self-ratings on various types of skills.

There are four possible sources of ratings in most 360-degree assessments:
subordinate, peer, boss, and self. Many studies in the 360-assessment ﬁeld focus on the
level of rater-ratee agreement (primarily in studies assessing rater accuracy). However,
few studies in the 360-assessment literature have examined the circumstances in which
differences between types of raters arise. In addition, there are less than a handful of
studies which have examined whether or not raters differ in the speciﬁc leadership
behaviors they associate most with effectiveness. For instance, a few researchers have
found male preferences in ratings provided by bosses (Hooijberg & Choi, 2000; Tsui &
O’Reilly, 1989) while one study found a positive female bias from female peer raters
(Martell & Desmet, 2001).

Researchers have also asked managers at differing levels to rate the importance of
various skills for their own jobs (e. g. Alexander, 1979; Gomez-Mejia, McCann, & Page,

1985; Paolillo, 1981; Pavett & Lau, 1983). At the ﬁrst line levels, managers favored

abilities involved in supervision such as monitoring potential problems and managing

conﬂict. At the middle management levels, respondents reported that theirjobs involved
greater human relations skills that involved encouraging cooperation, motivating others,
and developing subordinates. At the executive levels, managers reported that their jobs
required a broad range of skills including monitoring information, serving as a liaison,
and engaging in long-term planning. Finally, in a study of male managers’ perceptions of
the successful executive skills, respondents reported that successful executives need the
ability to act as change agents, have managerial courage, display a results orientation.
These are all highly agentic in nature (Martell et al., 1998).

In sum, 360 feedback assessments are usually intended to be developmental in
nature and are also believed to have many beneﬁts over more traditional performance
assessment systems. Given the nature of multirater assessment tools, the relationship
between the leader and rater, as well as the organizational level of the leader are expected
to inﬂuence the leader ratings. However, few studies have explored differences between
various types of raters on different types of skills. In addition, few studies have
examined the differences in self ratings on various types of skills.

Summary

Recent research and US. census information tells us that women are still having a
difficult time advancing to leadership positions in organizations, a phenomenon
commonly labeled the “glass ceiling.” Many researchers have explored possible causes
for the glass ceiling, and some have concluded that gender stereotypes play a large role.
More speciﬁcally, researchers have proposed that stereotypes affect the quality of the

performance evaluations and development opportunities provided to women.

25

Although researchers have examined gender differences in both the performance
evaluation and development ﬁelds, most empirical studies have dealt with gender
differences in evaluation. However, there are many mixed ﬁndings regarding the main
effects of leader gender (as well as rater gender) on performance evaluations. Some
research has found that women are generally rated lower than men, and other research
has found that men are rated lower or that there is no difference at all between genders
with regard to performance ratings. This batch of mixed ﬁndings has led some
researchers to begin exploring moderators of gender differences and performance
evaluations. Recently, Eagly and Karau (2002) proposed the Role Congruity Theory to
explain the more complex relationship between gender and performance evaluation.

Role Congruity Theory generally states that women will be rated lower when
there is a mismatch between their gender role and their job role. In the context of
leadership positions, this implies that when a woman’s stereotypically communal (e. g.
helpful, affectionate, kind, sympathetic, sensitive, gentle, etc.) gender role contrasts with
the stereotypically agentic (e. g. ambitious, aggressive, dominant, forceful, independent,
self-conﬁdent, etc.) leadership role, individuals will see her as a less effective leader.
Role Congruity Theory suggests that work context plays a large role in moderating the
relationship between gender and performance evaluation to the extent that cues in the
environment make either a person’s gender or leadership role more salient.

There are many lab studies which generally support the notions pr0posed in Role
Congruity Theory. These studies are typically limited to looking at bottom up ratings of
paper leaders or confederate peers. These studies are also limited in the types of

measures used. Oﬁentimes the performance evaluation instruments are developed for the

26

 

study and are not based on leadership theories about the types of skills good leaders
should possess. Also, many of these studies do not assess leadership performance
specifically. Rather, they address general performance on a lab task.

There are few ﬁeld studies to date exploring the tenets of Role Congruity Theory,
and the ﬁeld studies that do exist are limited in scope. Many are limited by the industry
in which the participants work, or only include women in middle management positions.
Given the importance of the work context in this field, many researchers have called for
the need for more ﬁeld research in this area.

In addition, there are no studies to date that have taken an integrative approach to
studying the glass ceiling with regard to performance evaluation and leadership
development. In other words, few have examined those areas in which leadership
development is intermixed with assessment methodologies. Many organizations are
beginning to merge evaluation assessments with development programs, and more and
more companies are using multi-rater development assessments such as 360-degree
feedback surveys (Peiperl, 2001). Unfortunately, little is known about the effects of
gender stereotypes on such tools. This recent trend toward using more assessment-based
leadership development programs raises some interesting questions about how
stereotypes may manifest themselves in these types of development paradigms. The

following hypotheses section speciﬁes some of these questions.

HYPOTHESES

As outlined in the introduction, there are three main areas of interest in this study.
First, the context of the work environment is expected to impact the relationship between
leader gender and leader ratings. Second, differences are expected in ratings for
communal and agentic leadership skills. Third, the impacts of rater and leader level on
leadership ratings are going to be explored.

The current study attempts to address these areas of inquiry through the use of a
360-degree feedback, ﬁeld data set. The proposed data set contains self and others’
(bosses, peers, and subordinates) ratings of male and female leaders working in various
contexts. In addition, the measure used to collect the data contains both communal and
agentic skill development items, as well as a set of overall leader effectiveness items.

The context variables in this data set include industry, functional area, and
proportion of male raters. The degree of male or female domination in each of these
contexts will be measured via the percentage of men and women operating in each
context. In other words, the percentage of males and females working in each industry
and job function will be used in the analyses. For proportion of male raters, the
percentage of male raters (combined bosses, peers, and subordinates) for each leader will
be calculated.

In addition to various context variables being used in this analysis, there are also
various types of ratings. There are communal development ratings, which assess the
leaders’ development needs with regard to communal leadership skills. There are also
agentic development ratings, which assess the leaders’ development needs with regard to

agentic leadership skills. These two types of development ratings are not expected to

differ in every circumstance. In cases where they are not expected to differ, they will be
combined to form a rating for overall development. Finally, there are effectiveness
ratings. The content of the effectiveness items focuses on leaders’ general leadership
ability, leadership potential, and expected success. The effectiveness items are not
communally or agentically loaded, and represent the raters’ overall perception of each
leader’s effectiveness. In some of the following hypotheses, there is reason to believe
that the independent variables will affect development and effectiveness ratings
differently, and in others there is reason to believe that they will be affected in the same
manner. The types of ratings being used as the dependent variables in each hypothesis
will be speciﬁed. Speciﬁc research ﬁndings and hypotheses are provided below.

Work Context

 

Eagly and Karau’s (2002) Role Congruity Theory suggests that the role of leader
and the stereotypical female gender role are incongruent. The theory supports the notion
that women are perceived as being relatively ineffective when compared to men to the
extent that leadership roles are deﬁned in predominately masculine terms. Although
Eagly and Karau (2002) discuss many ﬁndings that support their theory, there are still
gaps in the research in this area with regard to the speciﬁc circumstances or environments
in which the incongruence between gender roles and leadership roles is manifest.

Some researchers have noted that a major factor in enhancing the saliency of the
gender of female leaders may be organizational context (Becker et al., 2002; Eagly &
Karau, 2002; Ridgeway, 1992). Speciﬁcally, a woman’s gender is more likely to be
noticed when she is working in a male-dominated, or non-traditional context. Research

in both laboratory and organizational settings support this notion and have indicated that

29

 

when phys
prominent
al., 1995, I

Tim
make more
less effectix
in male-do;

compafEd tr

 

 

 

COntexts
Titer
females‘ 110'
Some studit
individual_]
leadership I?
€11,200”F
the militar)
FOU’ & de \
ﬁmctiOna] a
Heilman, 2

Few

speciﬁCaHV

when physical characteristics such as gender and color are novel, they become more
prominent and can inﬂuence observers’ perceptions in a stereotypical manner (Eagly et
al., 1995; Kanter, 1977).

The current research, therefore, hypothesizes that in contexts which emphasize or
make more salient the difference between gender and leader roles, women will be seen as
less effective and needing more development than men. In other words, women working
in male-dominated contexts may have a harder time being accepted as a leader when
compared to men in the same context or when compared to women in female-dominated
contexts.

There is some support for the proposition that context affects the evaluation of
females; however the way in which researchers deﬁne context varies from study to study.
Some studies have examined organization-level variables while others have examined
individual-level variables. Contexts that have been explored include male domination of
leadership positions (Eagly & Karau, 2002); proportion of male subordinates (Bowen et
al., 2000); proportion of men among the raters (Bowen et al., 2000); industries including
the military, education, government, and social services (Eagly & Karau, 2002; Lord,
Foti, & de Vader, 1984); organizational level (Hunt, Boal, & Sorenson, 1990); and
functional areas including line jobs, staff jobs, carpentry, and secretarial work (Lyness &
Heilman, 2002; Robbins & DeNisi, 1993).

Few of the studies examining context variables have focused on female leaders
speciﬁcally (most have not had a large enough sample of female leaders to examine
gender differences), and many were limited to exploring only one context area within a

study. In addition, many of the industries and functional areas included in existing

 

 

studies
etc )

explorc'

chosen

rater ge
funCIIOI

section:

Provide:
manufac
due to tl
\K‘ere per
bUI modi
organiza.
Becker e
leadershi.
in bankln
females 11
that “‘Omi
thESe ind”
industries)

leads 10 ”It

studies do not translate well into the business domain (i.e. nursing, military, education,
etc.) With regard to the current study, three types of organizational context will be
explored: industry, ﬁinctional area, and rater gender composition. These contexts were
chosen because they represent a variety in levels of analysis. Industry can be assessed
across organizations, ﬁinctional area can be assessed across and within organizations, and
rater gender composition is speciﬁc to each leader. In addition, the industries and
functional areas explored in this study will be business-oriented. The following three
sections describe each type of context in detail.

Industg. Industry refers to the type of product or service an organization
provides. Examples of industries include healthcare, education, utilities, automotive, and
manufacturing. The types of industries explored in past studies have been limited mainly
due to the lack of sufﬁcient samples. However, some research has shown that women
were perceived as being substantially less effective than men in military organizations,
but modestly more effective than men in education, government, and social service
organizations (Eagly & Karau, 2002). Looking at more business-oriented industries,
Becker et al., 2002) analyzed discrepancies in self-subordinates’ perceptions of
leadership behavior. They found that there were greater discrepancies for female leaders
in banking, accounting, and manufacturing than for males in those settings and for
females in education. These studies support the predictions of Role Congruity Theory
that women in male-dominated industries (i.e. a much higher percentage of men work in
these industries) are less likely to be rated to be as effective as men in male-dominated
industries, but when in female-dominated industries, are seen as effective as men. This

leads to the ﬁrst hypothesis in this study:

 

f.
others '(L
female It
and turn.
effct'm'r’l
expected
and t’ﬂct

E
leader’s }
engineeri
gender t}-
work“ 0r
With func
aSsociatec
argued 1h;

gender an
& Phillips
evaluating
She ‘5 Perl?
Thi
In their me
to
und SIIOr

mm and \V,

prO"ided lo

Hypothesis 1: Industry will moderate the relationship between leader gender and
others ’(combined boss, peer, and subordinate) ratings. In male-dominated industries,
female leaders will receive lower overall development ratings (i. e. needing more agentic
and communal development) and lower effectiveness ratings (i. e. performing less
eﬂectively) than male leaders. Irivgrrale-dominated industries, no diﬂerences are
expected between male and female leaders with regard to others’ overall development
and eﬂectiveness ratings.

 

Functional area. For purposes of this study, functional area is deﬁned as the
leader’s primary work area or general job role. Examples of functional areas include
engineering, production, human resources, and support staff. Functional areas can be
gender typed in the sense that they tend to be viewed as consisting primarily of “man’s
work” or “woman’s work” (Dipboye, 1985). For instance, men are typically associated
with functional areas such as engineering and manufacturing whereas women are often
associated with functional areas such as human resources and support staff. Many have
argued that such gender types associated with certain jobs may interact with the ratee’s
gender and can lead to gender bias in performance evaluations (Bartol, 1999; Freedman
& Phillips, 1988; Knight & Saal, 1984; Nieva & Gutek, 1980). Thus, a critical factor in
evaluating context effects on leader evaluations may be whether the gender of the leader
is congruent with the gender role that is typically associated with the particular job he or
she is performing.

There is some empirical support for the gender/functional area congruence notion.
In their meta-analysis of some of the Goldberg-paradigm experiments, Eagly et al. (1992)
found strong bias against female leaders in the subgroup of experiments that portrayed
men and women as basketball coaches. Robbins and DeNisi (1993) also found that raters

provided lower performance ratings in gender-incongruent situations (such as a female in

a)
Is.)

 

a carpent
secretariz

,\

(2002) fc.
to as “lin.
evaluatioz
external a
jobs are c.
male stere
consistent
stereotype
Ly;
others hat]
Slocum (y
orféanizatio
manage“ it
lEga] eXPEI‘]
line DOSIIIQI
StaffPOSitig
L’nﬁ
Organizathn
ﬁlnher Subs:

a carpenter job) as compared to gender-congruent situations (such as a female in a
secretarial job).

More recently, in a study of 485 upper level managers, Lyness and Heilman
(2002) found that women received lower performance ratings than men in what they refer
to as “line” jobs (business management, operations management, sales) and higher
evaluations than men in what they refer to as “staff” jobs (human resources, legal, and
external affairs). Lyness and Heilman (2002) concluded that the characteristics of line
jobs are consistent with the forceful and achievement-oriented attributes associated with
male stereotypes. Alternatively, they concluded that the characteristics of staff jobs are
consistent with the caring and relationship-oriented attributes associated with female
stereotypes.

Lyness and Heilman’s (2002) ﬁndings are consistent with the ways in which
others have deﬁned the duties for line and staff functions. Hellriegal, Jackson, and
Slocum (2002) described managers in line jobs as directing and controlling essential
organizational activities such as selling or producing products, whereas they described
managers in staff jobs as providing support, such as human resources, public relations, or
legal expertise. Ragins and Sundstrom (1989) also supported the idea that managers in
line positions tend to have greater organizational power and inﬂuence than managers in
staff positions.

Unfortunately, the Lyness and Heilman (2002) study was limited to one
organization in which ratings were provided by supervisors only. The current study can
further substantiate their ﬁndings by looking across organizations and raters. This leads

to the second hypothesis in this study:

33

 

 

1
gender i
function
needing
perform.
dtﬁl’renr
develop,

leadershi
somexx he
to no dire
I974, Let
Stumpfg;
finding {3
issue (Bar
in this are;
Thompson
(Yammarii
few ﬁEId 51
but ramEr e
employe€S
In 0
ratingS OfA

two ofthrec

H )pothesis 2: Functional area will moderate the relationship between leader
gender and others ’(combined boss, peer, and subordinate) ratings. In male-dominated
functional areas, female leaders will receive lower overall development ratings (i. e.
needing more agentic and communal development) and lower eﬂectiveness ratings (i. e.
performing less eﬂectively) than male leaders. Infemale-dominatedfunctional areas, no
diﬂerences are expected between male and female leaders with regard to others’ overall
development and eﬂectiveness ratings.

 

Rater Gender Composition. The direct effect of rater gender on male and female
leadership ratings has been explored in numerous studies. The ﬁndings from this area are
somewhat mixed. Results from a variety of lab studies suggest that rater gender has little
to no direct impact on performance evaluations (Hall & Hall, 1976; Jacobson & Effertz,
1974; Lee & Alvares, 1977; London & Poplawski, 1976; London & Stumpf, I983;
Stumpf & London, 1981). Although such lab studies have been generally consistent in
ﬁnding few rater gender direct effects, relatively few ﬁeld studies have investigated this
issue (Bartol, 1999). Some have noted that a primary reason for the lack of ﬁeld studies
in this area is that raters are often predominately male (Griffeth & Bedeian, 1989;
Thompson & Thompson, 1985; Tsui & Gutek, 1984; Tsui & O’Reilly, 1989) or female
(Yammarino & Dubinsky, 1988), precluding such tests in a single study. In addition, the
few ﬁeld studies that have been conducted do not speciﬁcally address leadership ratings,
but rather examine job speciﬁc performance ratings provided to non-management
employees.

In one such ﬁeld study, Pulakos et al. (1989) analyzed the ofﬁcial performance
ratings of Army personnel. They found that females gave higher ratings than males on
two of three performance dimensions. They also found signiﬁcant, but minimal,
interactions with ratee gender and rating source (peer or supervisor). Two other ﬁeld

studies, however, did not produce ﬁndings to support rater gender direct effects. Mobley

 

(1982) ti
manager:
Shore an
for men z

A
coHeagu<
ratings (I
tendency
in cases \

D
gender or
‘c’eHder Cc
B01361] CI
in fl€ld 51
raters. E2
aSSOQaIe<
eﬁeCll\‘er
female 16;
0f male 51

(1982) found no direct effect in a study of performance appraisal ratings for non-
managerial and non-professional employees in a large supply distribution center, and
Shore and Thornton (1986) found no direct effect in an analysis of performance ratings
for men and women doing assembly work.

Alternatively, there have been some studies, primarily from the work of Eagly and
colleagues, which have shown that rater gender does have a direct effect on leadership
ratings (Eagly & Karau, 2002; Eagly et al., 1992). Their research showed a small
tendency for participants to evaluate female leaders less favorably than males, especially
in cases where the evaluators were men.

Despite the variation in ﬁndings regarding the direct effect ofindividual rater
gender on leadership ratings, one relatively consistent ﬁnding has been reported when the
gender composition of the group of raters is used as the dependent variable. For example
Bowen et al. (2000) found little evidence of overall gender bias in performance appraisals
in ﬁeld studies, however they found signiﬁcant pro-male biases when only men served as
raters. Eagly and Karau (2002) also found that greater proportions of male raters were
associated with less effective ratings for women (relative to men) on measures of leader
effectiveness (see also Bowen et al., 2000). In addition, other research has shown that
female leaders are seen as being less effective relative to male leaders as the proportion
of male subordinates increased (Eagly & Karau, 2002). In this study, we will use the
percentage of all (bosses, peers, and subordinates) male raters to represent rater

composition for each leader. This leads us to the third hypothesis in this study:

35

 

int
2155
de:
0\'t

dex

dis:
inﬁ~
PTO]
Indé

Hypothesis 3: Theproportion of male raters (i. e. percentage of male boss, peer,
and subordinate raters) will moderate the relationship between leader gender and
others ’(combined boss, peer, and subordinate) ratings. When the proportion of male
raters is higﬁ, female leaders will receive lower overall development ratings (i. e. needing
more agentic and communal development) and lower eﬂectiveness ratings (i. e.
performing less effectively) than male leaders. When the Loportion of male raters is
l____,ow no differences are expected between male and female leaders with regard to others
overall development and effectiveness ratings.

 

Types of Ratings

In addition to the context effects on leadership ratings, it is important to consider
the content of the evaluation instrument for possible gender differences. With regard to
360 assessments speciﬁcally, researchers have been primarily concerned with
investigating the psychometric viability and practical value of the tool. Researchers have
focused little effort on studying the content of the evaluation items. However, given that
360 evaluations are unique from other types of assessments in that they are usually
intended for development purposes, it is important to consider the content of these
assessments particularly with regard to how they represent what leadership skills are
desired. With regard to this study, we will be exploring three evaluation content areas:
overall leadership effectiveness, development of agentically-oriented skills, and
development of communally-oriented skills.

Development vs. Effectiveness Items. The Price Waterhouse vs. Hopkins case

 

discussed in the introduction highlights an important distinction in the types of
information used in promotion decisions. In that case, a female leader was denied
promotion despite having performance ratings comparable to her male counterparts. The
judge ruled that she was denied promotion because she was expected to be more
congenial than her male counterparts solely on the basis that she was a woman (Fiske et

al., 1991). Essentially, she did not receive a promotion because her superiors thought she

 

was lack
hypothe'
like for l

to or hig

role ma}
her fulti
evidencr
found th
regarder
l€ss ratii

Successf

rn

at a larg
On One c
[We on“.
0"erall [
highligh
differen:
may 1101
the POSS

ratinL’S p

was lacking the communal-oriented skills expected of her as a woman. If we were to
hypothesize what performance effectiveness and development ratings might have looked
like for Hopkins, we might suggest that her performance effectiveness ratings were equal
to or higher than her male peers while her development ratings were left wanting.

The Price Waterhouse vs. Hopkins case shows that a woman who fulﬁlls a leader
role may elicit negative reactions, even while she may also receive positive evaluation for
her fulﬁllment ofthis role (Eagly & Karau, 2002). Heilman et al. (1995) found some
evidence ofthis combination of positive and negative evaluations in a lab study. They
found that even when researchers described female managers as successﬁil, participants
regarded these women as more hostile (i.e. devious, argumentative, selﬁsh, bitter) and
less rational (i.e. illogical, subjective, unable to separate feelings from ideas) than
successful male managers.

Shore (1992) found similar results in a ﬁeld study of female and male employees
at a large petroleum company. In this study, women scored signiﬁcantly higher than men
on one of the performance dimensions (performance style) and equivalently to men on
two other dimensions (cognitive measures and interpersonal style). However, women’s
overall ratings were not signiﬁcantly higher than the men’s ratings. Shore’s (1992) study
highlights the importance of looking beyond an overall rating and rather investigating
different types of ratings. As Bartol (1999) put it, “Equal ratings for males and females
may not always be equitable” (p. 169). This leads us to Hypothesis 4 which deals with
the possible discrepancy between overall effectiveness and individual development

ratings provided to men and women.

 

 

 

Tl

IOI

hy

[0

rep

par

COT
aSSl

tot

rEltir
CODE

 

Hypothesis 4: LeadeLgender will moderate the relationship between others'
(combined boss, peer, and subordinate) overall development (agentic and communal)
and eﬂectiveness ratings such that the two sets of ratings will be more strongly, positively
related for male leaders than for female leaders. In other words, ratings indicating a
lesser need for development will be more strongly related to ratings indicating greater
eﬂectiveness for men than for women.

Communal vs. Agentic Development Items. In addition to looking at
development items as a whole, they can be separated into those which describe more
agentically-oriented skills and those that describe more communally-oriented skills.
There are many different types of skills that a good leader needs to have, and various 360
tools differentially emphasize some types of skills more than others.

Although not addressed in any current research, it seems reasonable to
hypothesize that there is a difference in the types of development feedback raters provide
to male and female leaders. Much of the logic behind such a hypothesis stems from the
research in gender differences in leadership style. For instance, Eagly and Karau (2002)
reported that women would receive less positive reactions when fulﬁlling their roles in a
particularly dominant, assertive, directive, or self-promoting style.

Across various studies, leadership style has been categorized as either attention to
task accomplishment (or initiating structure) versus attention to relationships (or
consideration behavior) (Bass, 1990). Initiating structure has been shown to be
associated with stereotypic masculine attributes, whereas consideration has been shown
to be associated with stereotypic feminine attributes (Korabik, 1990).

Some researchers have examined the degree to which male and female leader
ratings differ as a ﬁrnction of the general leadership styles of initiating structure and

consideration. In a lab study, Bartol and Butterﬁeld (1976) found that female leaders,

relative to male leaders, received more positive ratings if they adopted a consideration

38

 

style In a
structure st
unable to r t
lzraeli & 12
F iel
category of
(involving (
(V'room & ‘
gender diffc
initiating str
democratica
given to wor
\"room (193
both men an.
Their study 5
leadership th
(Cleveland et
ShOWn that 31
than male lea,
Bari], & Wars

Skills rather ll]

style. In another lab study, Bass (198]) found that raters considered an initiating
structure style to be more suitable for men than for women. Field research has been
unable to replicate these lab ﬁndings (Butterﬁeld & Powell, 1981; Cleveland et al., 2000;
Izraeli & Izraeli, 1985; Rice, Instone, & Adams, 1984).

Field studies have been more successﬁil when investigating another general
category of leadership style: the degree to which leaders use a democratic or participative
(involving others) versus autocratic (directing others) approach to decision making
(Vroom & Yetton, 1973). Powell (1993) reported that meta-analyses revealed larger
gender differences on participative and autocratic dimensions than on consideration and
initiating structure dimensions, with women tending to be seen as leading more
democratically than men. This difference may be partly due to the negative evaluations
given to women who lead autocratically (Cleveland et al., 2000). In addition, Jago and
Vroom (1982) showed that a participative leadership style was rated as appropriate for
both men and women, but an autocratic style was deemed appropriate only for men.
Their study suggests that women may have a more restricted choice in the style of
leadership they can use in order to be perceived in a positive light by their followers
(Cleveland et al., 2000). These ﬁndings are consistent with other studies which have
shown that autocratic or dominating leadership behavior is less well received from female
than male leaders (Copeland, Driskell, & Salas, 1995; Eagly & Karau, 2002; Korabik,
Baril, & Watson, 1993; but see Luthar, 1996). This leads us to hypothesize that female
leaders will be provided with feedback that suggests they should work on their communal

skills rather than build their agentic skills.

 

 

h.
l"(1

35
ﬁx

[)6
ht” .

(It

Pt’t
bet

E

lier
lead
exar
lead

the r

ofrh
IVE r
OIEx
ab0ur
the ,e

leadEr

Hypothesis 5a: Female leaders will receive lower communal development ratings
(i.e. needing more development in communal skill areas) from others (combined bosses,
peers, and subordinates) as compared to their male counterparts. The discrepancy
between female and male ratings is expected to be more pronounced in male-dominated
industries.

Hypothesis 5 b." Female leaders will receive lower communal development ratings
(i. e. needing more development in communal skill areas) from others (combined bosses,
peers, and subordinates) as compared to their male counterparts. The discrepancy

between female and male ratings is expected to be more pronounced in male—dominated
functional areas.

Hypothesis 5c: Female leaders will receive lower communal development ratings
(i. e. needing more development in communal skill areas) from others (combined bosses,
peers, and subordinates) as compared to their male counterparts. The discrepancy
between female and male ratings is expected to be more pronounced as the proportiorLof
male raters increases.

 

Opgtnizational Level

Organizational level is the third major category of interest in the current study.
Here, organizational level refers to two things: the level of the rater and the level of the
leader. The organizational level of the rater is important in a relational sense. For
example, we are not interested in whether the boss providing ratings is an executive
leader or a middle manager. Rather, we are interested in the fact that this individual is
the ratee’s superior.

The organizational level of the leader, however, is a different matter. In the case
of the ratee, we are interested in where he or she is positioned within the organization.
We need to know if the leader being evaluated is a ﬁrst line supervisor, middle manager,
or executive. In a sense, once the organizational level of the leader is known, inferences
about the positions of the raters can be made. The following review, therefore, describes
the research and hypotheses relevant to two sets of variables: leader/rater relationship and

leaderlevel

40

 

 

 

as:
as:
ex
ad

I81

wc
atri
Sh.
der

ind

Fate

cha

Leader/Rater Relationship. There are four possible sources of ratings in most
360-degree assessments: subordinate, peer, boss, and self. Many studies in the 360-
assessment ﬁeld focus on the level of rater-ratee agreement (primarily in studies
assessing rater accuracy). However, few studies in the 360-assessment literature have
examined the circumstances in which differences between types of raters arise. In
addition, there are less than a handful of studies which have examined whether or not
raters differ in the leadership roles they associate most with effectiveness. Due to the
lack of research in these areas in the 360 literature, the hypotheses proposed in this
section of the current study are largely based on the theory of Relational Demography.

The theory of Relational Demography is generally based on the similarity-
attraction paradigm and suggests that individuals are more attracted to other individuals
who share their same demographic characteristics (Tsui & O’Reilly, 1989). These
characteristics may include age, tenure, gender, and race among other things. According
to the research in this area, the attraction between individuals sharing the same
demographic characteristics can manifest itself in many different ways. With regard to
work settings, it has been found that demographic similarity can affect individuals’
attitudes toward their work groups, perceptions of advancement opportunities (Riordan &
Shore, 1997), and extra-role behaviors (Tsui, Porter, & Egan, 2002). In addition,
demographic similarity has also been shown to be related to supervisors’ perceptions of
individual helping behaviors (Tsui et al., 2002).

In the current study, the theory of Relational Demography would suggest that
raters are more likely to rate leaders positively if they share the same demographic

characteristics, in this case gender. A few researchers have found differences in ratings

41

 

 

“‘6'.

int:

tend
L'nfi
their
lllCOl
fEma
that I
Will 1
relati
are IT
more
df’l‘t’li
Perla;

d1 367‘;
”IQ/ex

provided to males and females by their bosses. For instance, Tsui and O’Reilly (1989)
showed that demographic differences between a subordinate ratee and a superior
signiﬁcantly affected performance evaluations of the ratee. Gender differences were a
signiﬁcant predictor, with speciﬁc results indicating that ratees in opposite-gender dyads
were rated lower and were liked less well than ratees in same-gender dyads. Speciﬁc
interactions in this study showed that female superiors gave female ratees higher
evaluations than male ratees and indicated greater liking; however they found no
differences in ratings from male superiors.

Hooijberg and Choi (2000), alternatively, found that superiors (primarily male)
tended to see male managers as being more effective than female managers.
Unfortunately, the results of their study are limited due to the small number of women in
their study. Ninety percent (90%) of their sample were white males. Although relatively
inconclusive, the results of these two studies suggest that bosses might rate male and
female leaders differently depending on their own gender. Speciﬁcally, it is proposed
that male bosses will rate male leaders higher than female leaders, and that female bosses
will rate female leaders higher than male leaders. It is also expected that both of these
relationships will be stronger in male-dominated contexts. In other words, male bosses
are more likely to favor other men when women are the minority, and female bosses are
more likely to favor other women when they are the minority.

Hypothesis 6a: Male bosses will provide higher communal and agentic
development ratings (1'. e. needing less development) and higher effectiveness ratings (i. e.
performing more eﬂectively) to male leaders as compared to female leaders. The

discrepancy between male and female ratings is expected to be more pronounced in
male-dominated industries.

42

 

d

[R
d l

4.
pt

”ll

dc

do
m

Hypothesis 6b: Male bosses will provide higher communal and agentic
development ratings (i. e. needing less development) and higher eﬂectiveness ratings (i. e.
performing more eﬂectively) to male leaders as compared to female leaders. The
discrepancy between male and female ratings is expected to be more pronounced in
male-dominatedfunctional areas.

Hypothesis 6c: Male bosses will provide higher communal and agentic
development ratings (i. e. needing less development) and higher effectiveness ratings (i. e.
performing more eﬂectively) to male leaders as compared to female leaders. The
discrepancy between male and female ratings is expected to be more pronounced as the
proportion of male raters increases.

Hypothesis 7a: Female bosses will provide higher communal and agentic
development ratings (i.e. needing less development) and higher eﬂectiveness ratings (i.e.
performing more effectively) to female leaders as compared to male leaders. The
discrepancy between female and male ratings is expected to be more pronounced in
male-dominated industries.

Hypothesis 7b: Female bosses will provide higher communal and agentic
development ratings (i. e. needng less development) and higher eﬂectiveness ratings (i.e.
performing more eﬂectively) to female leaders as compared to male leaders. The
discrepancy between female and male ratings is expected to be more pronounced in
male-dominated functional areas.

 

Hypothesis 7c: Female bosses will provide higher communal and agentic
development ratings (i. e. needing less development) and higher effectiveness ratings (i.e.
performing more eﬂectively) to female leaders as compared to male leaders. The
discrepancy between female and male ratings is expected to be more pronounced as the
proportion of male raters increases.

Although there are no speciﬁc ﬁndings with regard to peer ratings in the
Relational Demography literature, we can generalize the theory to predict that female
peers will rate female leaders higher than male leaders, and that this effect will be
stronger in female contexts. Martell and Desmet (2001), for instance, reported ﬁnding a
female bias in their study such that female managers rated other female managers more
favorably than male managers on certain skills (e. g. inspiring, mentoring, problem

solving, rewarding, and supporting), whereas male respondents displayed no bias. They

claimed this pattern of results indicates a same gender bias on the part of women raters.

 

 

 

 

 

th”

{an
rhs

fen

dbl

per.
(1,151

ﬂvr

dt’l
Pcf
chm
£211

and

Klal
then
(20c
a pU‘.
lead:
Plutﬁ

exkx

Peer;

Hypothesis 8a: Female peers will provide higher communal and agentic
development ratings (i.e. needing less development) and higher eﬂectiveness ratings (i.e.
performing more eﬂectively) to female leaders as compared to male leaders. The
discrepancy between female and male ratings is expected to be more pronounced in
female—dominated industries.

Hypothesis 8b: Female peers will provide higher communal and agentic
development ratings (i. e. needing less development) and higher eﬂectiveness ratings (i. e.
performing more effectively) to female leaders as compared to male leaders. The
discrepancy between female and male ratings is expected to be more pronounced in
female.dominatedfunctional areas.

 

Hypothesis Re: Female peers will provide higher communal and agentic
development ratings (i. e. needing less development) and higher eﬂectiveness ratings (i.e.
performing more effectively) to female leaders as compared to male leaders. The
discrepancy between female and male ratings is expected to be more pronounced as the
proportion of male raters decreases.

There are only a few researchers (e.g. Pfeffer & Salancik, 1975; Tsui, 1984; Tsui
& Ohlott, 1988) who have empirically investigated systematic differences in the speciﬁc
types of leadership behaviors subordinates, peers, and bosses associate with
effectiveness. Furthermore, these studies are limited to a very narrow range of industries
and have not speciﬁcally examined the issue in relation to leader gender. In a ﬁeld study,
Maher (1997) found that there were no differences in the way subordinates evaluated
their actual male and female managers. However, more recently, Hooijberg and Choi
(2000) performed a ﬁeld study focusing on 360-feedback for 252 middle managers from
a public utility industry. They predicted that subordinates would care more about
leadership skills such as involvement, that peers would care more about skills such as
providing the organization with a mission, and that superiors would care more about the
external side of business such as customer relations. They found mixed results including

partial support for their subordinate and supervisor predictions and little support for their

peer predictions. Hooijberg and Denison (2002) have recently revised Hooijberg and

44

 

 

 

\l

II

but

altei
mal.
C011:
Inca
likei
Hyp
Imp;

rElar

Choi’s (2000) predictions by suggesting that bosses would give higher ratings to leaders
who show agentically-oriented leadership skills (such as providing mission), while direct
reports would give higher ratings to leaders with communally-oriented skills (such as
involvement).

Although these latter propositions have yet to be tested, they raise an interesting
issue regarding the relationship between rater type and the types of skills most related to
effectiveness. In addition, one can hypothesize how leader gender impacts the
relationship between rater type and the types of skills associated with effectiveness.
According to Role Congruity theory, women have a more narrow range of acceptable
behaviors than men. For instance, female leaders must be seen as agentic by their bosses,
but not too agentic as to appear masculine. Likewise, female leaders must appear
communal to their subordinates, but not too communal as to appear soft. Men,
alternatively, may have a more broad set of behaviors they can acceptably engage in. A
male leader who obviously cares about his subordinates is more likely than a female to be
considered as being an effective leader as opposed to being considered soft and
incapable. In addition, men are expected by their bosses to appear strong and are less
likely than women to be considered too assertive or controlling. Based on this logic,
Hypothesis 9 proposes that both rater type (boss or subordinate) and leader gender will
impact the relationships between agentic and effectiveness ratings, as well as the

relationship between communal and effectiveness ratings.

45

 

 

 

Ii

rl

OI

BI

ll‘ii

dil

Hypothesis 9: Rater type (boss or subordinate) and leader gender will interact in
their influence on the relationship between agentic and eﬂectiveness ratings, and on the
relationship between communal and eﬂectiveness ratings. It is expected that the
relationship between agentic and eﬂectiveness ratings will be higher for bosses than for
subordinates, more so for men than for women. In addition, the relationship between
communal and eﬂectiveness ratings will be higher for subordinates than for bosses, more
so for men than for women.

In addition to subordinate, peer, and boss ratings in 360 assessments, leaders are
often asked to provide self-ratings of their performance. Self-ratings are being used more
and more frequently within performance appraisal systems in organizations (Atwater,
1998), and Bartol (1999) suggests that gender differences in self-ratings may have
implications for directly or indirectly affecting the ratings that individuals receive from
their supervisors. There is some evidence to support the idea that there are gender-related
differences in self-ratings of leadership effectiveness, however ﬁndings are conﬂicting.

Deaux (1979), for example, found that male heads of a retail department store
chain viewed themselves as performing signiﬁcantly better and having greater ability
than did their female peers. In addition, the males rated themselves has having higher
intelligence, more difﬁcult jobs, more approval for their work, and better relationships
with the supervisor than did their female counterparts. However, a more recent study by
Tsui and Gutek (1984) indicated a trend toward equivalent or even higher self-ratings by
women rather than by men. Results from their study of middle managers in a multi-
company corporation showed that female manager self-ratings of performance
effectiveness were signiﬁcantly higher than those of males.

Examples of studies showing no differences between male and female self-ratings

include Shore and Thornton (1986), who found no gender differences in ratings on

familiar tasks in a real work setting in which performance feedback was available. Also,

46

in a study focusing on managers, Wohlers, Hall, and London (1993) concluded that male
and female managers from a variety of organizations did not differ in self-ratings.

As these studies show, there is little consistent evidence for a direct relationship
between leader gender and self-rating bias. Therefore, at this point little is known about
the contextual circumstances which affect the likelihood of women producing higher self-
ratings than do men (Bartol, 1999). However, just as Role Congruity Theory helped us
hypothesize about the factors affecting the ratings provided to female leaders from
subordinates, peers, and bosses, it can also help guide predictions about the possible
variables affecting female self-ratings.

Role Congruity Theory claims that context plays a large role in predicting when
gender impacts leadership evaluation. . The three types of contexts proposed for study in
the current research are industry type, functional area, and rater gender composition. If
we were to generalize our predictions concerning others’ ratings within these contexts to
self-ratings, then we can propose that self-ratings are going to be impacted by the context
in which the leader is performing.

Role Congruity Theory also suggests that women will respond to their own
performance similarly to how others respond such that female leaders will rate
themselves lower within contexts inconsistent with their gender role. Eagly and
colleagues have performed research in this area, and attribute lower self-ratings to
stereotype expectancies. They found that expectancies associated with gender roles
foster behaviors consistent with these roles (Eagly, 1987; Eagly & Karau, 2002; Eagly,
Wood, & Diekman, 2000). Others have also reported that gender stereotype expectancies

are especially powerful in yielding demonstrations of behavioral conﬁrmation (Geis,

47

1993; Skrypnek & Snyder, 1982; Snyder, 1981), including studies of stereotype threat
(Steele, 1997). For instance, Spencer, Steele, and Quinn (1999) showed that it is possible
to undercut women’s math performance through raising anxiety about conﬁrming
people's low expectations about women‘s inferior math ability. Based on this research, it
is reasonable to assume that women are more likely confronted with stereotype
expectancies in male-dominated work contexts as opposed to environments where they
are surrounded by other females.

This is also consistent with Relational Demography theory which states that
individuals are likely to compare their own demographic characteristics with those
around them and then determine whether or not their own demographics are similar.
Relational Demography theory suggests that the level of an individual’s demographic
similarity to the demographic composition of those around them drives the individual’s
work-related attitudes and behaviors (Riordan & Shore, 1997). Although Relational
Demography research has not speciﬁcally addressed the issue of self-ratings, the theory
suggests that individuals will be more conﬁdent about their own performance when
performing in situations where they share demographic characteristics with those around
them.

Hypothesis 10a: Industzy will moderate the relationship between leader gender
and self-ratings. In male—dominated industries, female leaders will give themselves lower
agentic and communal development ratings (i. e. needing more agentic and communal
development) than will male leaders. Infemale-dominated industries, no diﬂerences are

expected between male and female leaders with regard to self-ratings of agentic and
communal development needs.

 

 

48

Hypothesis 10b: Functional area will moderate the relationship between leader
gender and self-ratings. In male-daminatedﬁmctional areas, female leaders will give
themselves lower agentic and communal development ratings (i. e. needing more agentic
and communal development) than will male leaders. Infemale-dominatedfiinctiorial
areas, no diﬂerences are expected between male and female leaders with regard to self-
ratings of agentic and communal development needs.

 

Hypothesis I 0c: The proportion of male raters will moderate the relationship
between leader gender and self-ratings. When the mportion of male raters is high,
female leaders will give themselves lower agentic and communal development ratings
(i. e. needing more agentic and communal development) than will male leaders. When the
proportion of male raters is low, no differences are expected between male and female
leaders with regard to self-ratings of agentic and communal development needs.

 

In addition to context variables, the current study has proposed that leader
development ratings provided to males and females will vary partly as a function of the
types of items (i.e. communally or agentically-oriented) contained in the evaluation
instrument. Again, in generalizing these hypotheses to self-ratings, we propose that
females will provide higher self-ratings on communal-type skills than will men.

There is some research to support the claim that women see themselves as having
a higher level of communal leadership skills than do men. For example, in a study of
middle managers' self-reports of how they inﬂuence their bosses, females reported that
they acted more out of organizational interest than self-interest, considered others'
viewpoints, and focused on the interpersonal aspects as well as the task aspects of the
inﬂuence episode (Lauterbach & Weiner, 1996). In another study, both male and female
leaders reported behaving more agentically toward their subordinates relative to their
bosses. However, women reported using a more communal style overall, regardless of
their own organizational level in relation to those they interacted with (Moskowitz, Suh,

& Desaulniers, 1994 ).

49

Hypothesis I I .‘ Female leader self—ratings will be higher for communal skills (i. e.
needing less development in communal skill areas) than male leader self-ratings of
communal skills.

We also propose that in situations where females rate themselves highly in agentic
skills, others will see these leaders as being less effective. This proposition is consistent
with Role Congruity Theory in that women who see themselves as displaying agentic

skills are likely to be perceived by others as using a leadership style inconsistent with the

stereotypical female gender role.

 

Hypothesis [2: Leader gender will moderate the relationship between others’
(combined boss, peer, and subordinate) e ectiveness ratings and agentic development
self-ratings such that the two sets of ratings will be more strongly, positively related for
male than for female leaders. In other words, self ratings indicating a lesser need for
development of agentic skills will be more strongly related to others ' ratings indicating
greater effectiveness for men than for women.

Finally, there is another hypothesis that can be made about the relationship
between self and others’ ratings. London and Wohlers (1991) found that the self-
perceptions of female managers were more in agreement with their subordinates’
perceptions than were those of male managers. This is consistent with the notion in
hypotheses 9 (i.e. effectiveness ratings for female leaders will be positively related to
subordinate communal development scores) and l 1 (i.e. female leader self-ratings will be
higher for communal skills than male leader self-ratings of communal skills) such that
female leaders and subordinates will have more commonality regarding the skills they

rate highly than will male leaders and subordinates.

Hypothesis [3: Self-subordinate rater agreement on communal and agentic
development ratings will be higher for female leaders than for male leaders.

Leader Level. For purposes of this study, leader level refers to the ratee’s position

 

in the organization. Although researchers are often interested in those holding the highest

50

positions in organizations, they often deﬁne leadership more broadly. As mentioned in
the introduction, most research studies examine middle managers because they are
greater in number and more accessible. Although many studies only examine those in the
middle ranks of organizations, many have speculated about the generalizability of
ﬁndings to high-level leaders. Eagly and Karau (2002), for instance, proposed that the
incongruity between the female gender role and the leader role is likely to be more
extreme at higher levels ofleadership. The ﬁndings from a study by Lyness and Judiesch
(1999) are consistent with Eagly and Karau’s proposition. Their study included
managerial employees in a large ﬁnancial services organization in which women at
higher levels of management were less likely to be promoted when compared to men.
Adding to the complexity of the leader level issue, Eagly and Karau (2002)
describe middle managers as having a particularly multifaceted role relative to ﬁrst-line
managers or top leaders. They, and others, claim that the role of middle managers is
socially complex and requires stereotypically feminine interpersonal skills (Paolillo,
1981). Given the more socially complex characteristics of the middle management role,
Eagly and Karau (2002) suggest that the gender-leader role incongruity might be lower
for middle managers than for ﬁrst level or executive level leaders. Therefore, they
conclude that bias against female leaders is especially likely above and below the middle
management level and that women should fare well, relative to men, in middle-level
leadership positions as opposed to ﬁrst line or executive positions. In this study, we are
mainly concerned with the differences between leaders at the middle management and
executive levels. For that reason, the following hypothesis addresses differences between

these two groups.

51

Hypothesis I 4: Leader level will moderate the relationship between leader gender
and others ’(combined boss, peer, and subordinate) ratings. At the executive level, female
leaders will receive lower agentic and communal development ratings (i. e. needing more
development in agentic and communal leadership skills) and lower effectiveness ratings
(i. e. performing less eﬂectively) than male leaders. At the middle-management level, no
diﬂerences are expected between male and female leaders with regard to others’ ratings.

METHODS

Desion

 

The hypotheses presented in the last section were studied using data from an
archival set of responses to a 360 leadership evaluation: the Denison Leadership
Development Survey (DLDS; Denison & Neale, 1996). The responses on this survey
were provided in a single time period. The decision was made to use ﬁeld data as
opposed to lab data in this study for many reasons.

First, multiple researchers have highlighted the need for further ﬁeld research in
the area of gender and leadership evaluation (Bartol, 1999; Eagly & Karau, 2002). Given
that the use of 360 assessments is becoming more and more prevalent in practice and
research, this seems an appropriate forum for investigating the relationship between
gender and leader evaluation.

Second, the ratings provided in this data set were of real leaders by actual peers,
subordinates, and bosses. Although data gathered in a lab would allow for more control,
it would be difﬁcult to simulate a 360 assessment in the lab while maintaining external
validity. This data set gave us an opportunity to use data from a real
evaluation/development instrument currently being used in many organizations rather
than creating a contrived one for a lab study. In addition, ﬁeld 360 ratings allowed us to
assess self, top-down, and bottom-up ratings as opposed to just bottom-up ratings that
could be provided in a lab study.

Third, in addition to variation in types of raters, this data set contained variation in
the contexts within which the data were collected, as well as variation in the types of

ratings provided to individuals. The contexts in this data set included the industries and

functional areas in which the leaders worked, as well as the proportion of male raters for
each leader. The types of ratings included development and leadership effectiveness.
Further, the development items in the DLDS spanned both traditional and
transformational leadership skills.

Finally, this particular data set allowed us to assess ratings provided to female
executive leaders as opposed to strictly middle management, as most studies have done.
Lab studies are usually restricted to examining bottom-up, “subordinate” ratings of paper
leaders or team peers. In addition, as Bartol (1999) noted, the number of studies
conducted in ﬁeld settings has been small and have been largely concentrated on
positions near the bottom of the hierarchy where women have experienced more success
in being promoted.

Participants

The data were collected during the years 1996-2002 from 206 different American
companies]. The sample included organizational members from 51 various industries
including automotive, manufacturing, telecommunications, pharmaceutical, and others.
Subjects in this data set were all members of companies that paid Denison Consulting
Company for their services. Although it is impossible to say if all individual participants
were required to ﬁll out the survey or if they volunteered, it is the case that all surveys
were completed for developmental purposes only. Survey results were provided to the
leaders only. Therefore, survey results were not provided to the leaders’ bosses, peers, or
subordinates and they were not used for performance evaluation purposes.

The data set includes ratings of 3,938 leaders who were rated by 29,566

subordinates, peers, and bosses. It should be noted, however, that not all of these leaders

54

and raters were included in every analysis. For example, some leaders did not provide
self-ratings and some leaders and raters did not report their gender, therefore, these
leaders and raters were not included in analyses where these factors needed to be known.
Of those leaders who reported their gender, 988 were women and 2702 were men. This
means that there were approximately three times more male than female leaders in this
data set.

Table 1 shows the breakdown of the numbers of raters for male and female
leaders. As the table shows, female subordinate and peer raters made up 52% of female
leaders’ raters whereas female subordinate and peer raters accounted for only 24% of
male leaders’ raters. Alternatively, male subordinates and peers made up 63% of male
leaders’ raters whereas male subordinates and peers accounted for only 34% of female
leaders raters. In addition, female leaders were four times more likely to have a female
boss rater than were male leaders. It is important to restate here that leaders chose their
own raters. These numbers suggest that women leaders were more likely to chose
women raters than were male leaders.

Despite differences between the types of raters, both male and female leaders had
a similar average number of raters. As Table 2 shows, male and female leaders both had
an average of approximately 8-9 total raters, 4 subordinate raters, 3-4 peers, and 1-2 boss
raters.

In addition to gender, leaders and raters ﬁlled out a variety of demographic
information including their age, education, years with organization, annual salary, and
ethnic background (see Appendix A). The demographic characteristics of the participants

who provided demographic information are presented in Table 3. The results are

 

’ Eight companies were excluded from the study because they were not American orgrmizations.

55

categorized into the following groups: all leaders, male leaders, female leaders, all
bosses, male bosses, female bosses, all peers, male peers, female peers, subordinates,
male subordinates, and female subordinates.

There were little to no differences between groups or genders in the area of
ethnicity. Most of the participants were white/Caucasian (84-90%). Males reported
being slightly older than females within each group (leaders, bosses, peers, and
subordinates). This may be due in part to the relatively large percentage of male
compared to female leaders, bosses, and peers in the 50-59 age range. Across groups,
most leaders, peers, and subordinates were between the ages of 30 and 49 whereas most
bosses were between the ages of 40 and 59.

Overall, a higher percentage of males reported having higher education levels
than females within each group (leaders, bosses, peers, and subordinates). Across
groups, most leaders, bosses, peers, and male subordinates reported having either a
bachelor’s or master’s degree. Most female subordinates reported having some college
or a bachelor’s degree. Similarly, a higher percentage of males reported a higher number
of years spent in their organizations than females within each group. Across groups,
subordinates (male mean=5.48, 4-10 years; female mean=5. 15, 4-10 years) reported
slightly less years in the organization than peers (male mean=6.03, 6-10 years; female
mean=5.68, 4-10 years) and leaders (male mean=5.97, 4-10 years; female mean=5.56, 4-
10 years) who reported slightly less years than bosses (male mean=6.40, 6-15 years;
female mean=5.94, 4-10 years).

The demographic with the largest differences between groups was salary. Salary

was measured on an eight-point scale ranging from l=$25,000 or less to, 8=$200,001

56

plus (see Appendix A). Within each group (leaders, bosses, peers, subordinates), males
reported higher salaries than females. The salary discrepancies between men and women
were highest for bosses (male mean=6.42, or approximately $100,000 to $200,000;
female mean=5.38, or approximately $75,000 to $150,000) and subordinates (male
mean=4.58, or approximately $50,000 to $100,000; female mean=3.42, or approximately
$35,000 to $75,000). Across groups, female subordinates report being paid the least
(mean=3.42, or approximately $35,000-$75,000) and male bosses reported being paid the
most (mean=642, or approximately $100,000 to $200,000). Female leaders and peers
reported being paid about the same as male subordinates (female leader mean=4.78,
female peer mean=4.55, male subordinate mean=458; or approximately $50,000 to
$100,000), and male leaders and peers reported being paid about the same as female
bosses (male leader mean=5.36, male peer mean=5.4l, female boss mean=5.38; or
approximately $75,000 to $150,000).

Looking strictly at the leaders who were rated in this data set, male leaders had
signiﬁcantly (p<0.05) higher salaries than female leaders even after controlling for age,
education, tenure, ethnicity, functional area, and organizational level (male leader:
N=1277, mean=5.48, SD=1.40; female leader: N=536, mean=4.86, SD=1.40; R2=0.42).
Female leaders reported an average salary of $50,000 to $100,000 whereas male leaders
reported an average salary of $75,000 to $150,000.

Measures

Denison Leadership Development Survey. The instrument used to gather the data

 

used in this study is a 360 degree leadership assessment called the Denison Leadership

Development Survey (DLDS; Denison & Neale, 1996). It allowed each leader to make a

57

self assessment of both their developmental areas and overall effectiveness as a leader.

In addition, each leader chose up to 25 bosses, peers, and direct reports to rate them in the
same areas. Responses were conﬁdential in the sense that no names were attached to the
feedback reports that were provided to the leaders. In fact, raters had to agree that their
ratings could be shown in the case that they were the only rater within a boss,
subordinate, or peer category. Surveys were completed on-line and took approximately
20-30 minutes to complete. For more details about the administration and reporting of
the DLDS, see the excerpt taken from the DLDS facilitator guide (Denison & Neale,
2001) in Appendix B.

The DLDS was developed based on Quinn’s (1988) Competing Values
Framework (CVF). Others have used this framework effectively to explore leadership
and organizational culture issues (Buenger, Daft, Conlon, & Austin, 1996; Denison,
Hooijberg, & Quinn, 1995; Hart & Quinn, 1993; Hooijberg, 1996; Hooijberg & Choi,
2000). Quinn’s (1988) CVF provided a useful model of leadership for the current study
because it addresses both the traditional and transformational roles of a leader. The
tenets of CVF are that organizations are complex and dynamic, therefore managers and
leaders have to fulﬁll many competing expectations; they have to express both ﬂexibility
and control, and maintain both an internal and external focus.

Essentially, then, both the CVF and the DLDS are based on a quadrant type model
where the skills reﬂecting the leaders’ ﬂexibility and control are crossed with the skills
reﬂecting their focus on internal and external issues (see Figure 1). The ﬁrst quadrant
includes skills that represent both an external focus and ﬂexible leadership style. In the

DLDS, this quadrant is called the Adaptability dimension. The second quadrant includes

58

skills that represent an external focus and stable leadership style. This is called the
Mission dimension in the DLDS. The third quadrant includes skills that represent an
internal focus and ﬂexible leadership style. This is called the Involvement dimension in
the DLDS. Finally, the fourth quadrant includes skills that represent both an internal
focus and stable leadership style. This dimension is called Consistency in the DLDS. In
addition, within each of these four main dimensions (Adaptability, Mission, Involvement,
and Consistency), the DLDS speciﬁes three subdimensions as shown in Table 4.

As Table 4 shows, the Adaptability dimension contains skills dealing with
organizational learning and change, and customer focus. The Mission dimension
includes skills dealing with providing followers with the strategies, goals, and vision for
the organization. The Involvement dimension contains skills dealing with the people side
of leadership: empowering others, building teamwork, and developing others. Finally,
the Consistency dimension includes skills dealing with coordinating and integrating
organizational systems, resolving conﬂicts, and deﬁning and enforcing organizational
core values. The excerpt from the DLDS facilitator guide (Denison & Neale, 2001) in
Appendix C provides a more detailed description of each dimension and sub-dimension.
Also, the speciﬁc items within each sub-dimension of the DLDS are listed in Table 52.
The items in this survey are all stated positively such that high ratings indicate
proﬁciency in a skill (i.e. does not need further development). The rating responses are
in a Likert-type format and range from l=strongly disagree to 7=strongly agree.

For purposes of this study, we were primarily concerned with which set of items

dealt most with agentic-type and communal-type leadership skills. Conceptually, there

 

2 For a report on the psychometric properties of the DLDS. sec C 110 (2000) and Hooijberg and Denison
(2002).

59

are some items that appeared to represent either agentic or communal skills. As
described in the introduction, agentic leadership styles are usually associated with
characteristics such as decisiveness, assertiveness, achievement orientation, conﬁdence,
ambitiousness, and an autocratic style. This implies that leaders with an agentic style of
leadership are concerned with providing their followers with direction, establishing high
standards of performance, and independently assessing the needs of the organization.
Alternatively, a communal style of leadership is generally associated with characteristics
such as nurturing others, caring for the welfare of others, being sensitive to individuals’
needs, and taking a service orientation.

In order to determine which items in the DLDS had the best conceptual ﬁt with
the notions of agentic and communal leadership skills, subject matter experts were given
deﬁnitions of agentic and communal skills and asked to rate each item of the DLDS on a
scale of l (agentic) to 5 (communal). The subject matter experts were
Industrial/Organizational Psychology graduate students who were all trained in item
analysis. Table 6 shows the item means, standard deviations, percent of subject matter
experts who rated the items above 3 and below 3, and if the item was chosen for inclusion
in the study as a communally or agentically-oriented item.

In all, 22 subject matter experts rated the items. One person’s ratings were
dropped due to missing ratings and outliers leaving 21 usable ratings. The items were
then chosen based on two criteria. First, the item had to have a majority of people
agreeing that it ﬁt on one side of the scale or the other. In other words, 50% or more of
subject matter experts had to rate the item as a l or 2 in order for it to be considered

agentic, and 50% or more had to rate it as a 4 or 5 for the item to be considered

60

communal. Second, the mean subject matter expert rating had to be equal to or above
3.50 or equal to or below 2.50 in order for the item to be considered communal or
agentic, respectively. Using these criteria, 22 agentic and 32 communal items were
chosen for use in this study (see Appendix D for a list of the communal and agentic
items). Reliabilities of the communal and agentic scales were then computed across
raters in the actual data set. Both scales produced high reliability coefﬁcients (communal
(1:085, agentic (1:090).

The agentic and communal scales were signiﬁcantly, highly correlated (r=O.83,
corrected r=0.95). The nature of the DLDS instrument was such that if a leader was rated
highly (needing less development) on one set of items, he or she was also likely to be
rated highly on most other items. Although every effort was made to create conceptually
distinct scales, the strong statistical relationship between the two sets of development
items made it more difﬁcult to ﬁnd results for those hypotheses suggesting different
effects for agentic and communal skills types. Normally, the strong relationship between
these scales would have helped to make a case for combining them, thereby dismissing
some of the hypotheses presented in this study. However, for purposes of this study, the
development scales were treated as conceptually distinct where hypothesized to be 503.

Leadership Effectiveness Items. There were seven leadership effectiveness items.

 

These items were asked at the end of the DLDS and addressed the overall effectiveness
of the ratee as a leader. These items had the same Likert-type response scale as the
development items (1=strongly disagree to 7=strongly agree; i.e. higher ratings indicated

higher leadership effectiveness). The effectiveness items were not tied to any

 

3 Analyses for the set of combined agentic and communal items were also run for each hypothesis. but did
not produce results dilTercnt from those reported in the Results section.

61

performance-related incentives such as pay raises or promotions. Therefore, the
effectiveness items used in this survey were not used in a traditional performance
appraisal sense. The leadership effectiveness items can be seen in Appendix E. The
reliability ofthe effectiveness items was assessed across groups (a=0.95). The
effectiveness items correlated highly with the agentic and communal scales (r=O.81 and
r=0.82 respectively). These correlations did not differ between the sets of ratings
provided to male and female leaders. The correlation between the development and
effectiveness items put limitations on the likelihood of ﬁnding evidence for hypotheses
suggesting a difference between these types of items. However, as with the development
scales, the set of effectiveness items was treated as a conceptually distinct scale based on
item content.

Organizational Level. The organizational level (or leader position) categories on

 

the Denison Leadership Develop Survey can be seen in Appendix F. Each leader was
responsible for indicating his or her own level. The seven levels used in the survey were
compiled into three levels: line, middle, and executive. Those who reported 2 (line-
management, N=695) were grouped into a line management category given that they
supervise non-management personnel. Those who reported 3 (middle management,
N=1286) were grouped into the middle management category given that they supervise
line managers. Finally, those who reported 4 (senior management, N=753), 5
(executive/senior vice president, N=322), and 6 (CEO/President, N=128) were grouped
into the senior/executive management category given that they supervise middle
managers. Those ratees who reported an organizational level of 1 (non-management,

N=175) or 7 (owner, N=3 7) were not coded given that we can not be sure these

individuals are in leadership roles in their organizations. Table 7 shows the number of
male and female leaders in each major category. As can be seen in the table, 27% of
female leaders were at the line management level and 32% were at the executive level.
Alternatively, 20% of male leaders were at the line management level and 40% were at
the executive level.

Functional area. Each leader was responsible for indicating his or her own
functional area (see Appendix G). Each of the functional areas used in the DLDS were
coded according to how male (or female) dominated it is. The functional areas used in
the DLDS were matched up with corresponding occupational areas in the Standard
Occupational Classiﬁcation (SOC) system used by the US. Bureau of Labor Statistics
(2001b; http://wwwbls.gov/eps/g)saatl1.pdf). The SOC system is used by all Federal
statistical agencies to classify workers into occupational categories for the purpose of
collecting, calculating, or disseminating data. In this system, all workers are classiﬁed
into one of over 820 occupations according to their occupational deﬁnition. To facilitate
classiﬁcation, occupations are combined to form 23 major groups, 96 minor groups, and
449 broad occupations.

Although the occupations used by the US. Bureau of Labor Statistics are not
perfectly representative of the functional areas used in the DLDS, this was the closest
classiﬁcation system available which also provided statistics on the gender make up of
the workforce within each occupation. Therefore, each DLDS functional area was
located in the SOC system and the corresponding percentage of females working within
that occupation was coded in the current data set. This was performed by two subject

matter experts at Denison Consulting. These subject matter experts were chosen based

63

on their familiarity with the DLDS functional area options. In cases where the experts
disagreed (for 3 of the 10 functional areas, or 30% of cases), they discussed the
categorization until they reached consensus. Consensus was reached for all functional
areas. Each DLDS functional area and its corresponding DOC classiﬁcation are also
provided in Appendix G.

Table 8 shows the number of leaders in each functional area. It is important to
note here that individuals could only be included when they provided the functional area
in which they worked. As the table shows, most leaders worked within the sales and
marketing (% female leaders=20, % male leaders=22), administration (% female
leaders=20, % male leaders=l 7), and professional (% female leaders=21, % male
leaders=15) functional areas. This is true for both male and female leaders. A higher
percentage of female leaders worked in the human resources (% female leaders=l 1, %
male leaders=3) and professional (% female leaders=21, % male leaders=15) ﬁinctional
areas whereas a higher percentage of male leaders worked in the manufacturing (% male
leaders=10, % female leaders=3) and engineering (% male leaders=8, % female
leaders=3) functional areas.

Table 9 shows the number of leaders that worked within male and female
dominated ﬁmctional areas. Not surprisingly, more female leaders in this data set worked
in functional areas typically consisting of a high percentage of women. Similarly, more
male leaders from this data set worked in functional areas typically consisting of a high
percentage of men.

Industry Type. The industry type of each company was not provided in this data

set, but was coded by Denison Consulting subject matter experts. The industry categories

64

used were taken from the Standard Industrial Classification (SIC) system that is used
throughout the federal government to group establishments into industries (US. Bureau
of Labor Statistics, 2001c). Two subject matter experts from Denison Consulting were
interviewed and asked what type of business each company in the data set was engaged
in. The subject matter experts were chosen based on their knowledge of the companies in
the data set. These individuals were asked to place each company from the data set into
the appropriate SIC category. In cases where the experts disagreed (for 21 of the 206
companies included in the study, or 11% of cases), they discussed the categorization until
they reached consensus. Consensus was reached for all companies. The 2001 US.
Bureau of Labor Statistics information regarding the percentage of females working in
each industry can be found at http://wwwbls.gov/cps/cpsaat18.pdf.

There were 51 total industries represented in this data set. Table 10 shows the
number of leaders in each major industry area. It is important to note here that one subset
of companies (20 total) in this data was not coded for industry and therefore not included
in any analyses which examined industry. Those companies were consulting ﬁrms which
used the DLDS with their own clients. Although the personal information provided by
the individuals within these consulting companies was still useful, we had no way of
tracking what industries these individuals worked in. In other words, we did not have the
company names from which these individuals actually worked, rather we had the
company name of the consulting ﬁrm that delivered them the survey. Therefore, rather
than grouping all of these individuals into a consulting service industry which would
have misrepresented their true industry affiliations, these individuals were not included in

the industry analyses (878 leaders total: 275 female leaders and 603 male leaders). Of the

leaders who were included, there were male leaders in 48 ofthe 51 industries and female
leaders in 41 ofthe 51 industries.

As can be seen in Table 10, 30% of male leaders worked in the manufacturing
industry while 29% of female leaders work in the services industry. A higher percentage
of female leaders worked in the transportation, communications, and other public utilities
(% female leaders=9, % male leaders=5), ﬁnance, insurance, and real estate (% female
leaders=l7, % male leaders=10); and services (% female leaders=29, % male leaders=17)
industries whereas a higher percentage of male leaders worked in the manufacturing (%
male leaders-=30, % female leaders—=17), wholesale and retail trade (% male leaders=15,
% female leaders=l 1), and construction (% male leaders=l l, % female leaders=6)
industries.

Table 1 1 shows the number of leaders that worked within male and female
dominated industries. As expected, more female leaders in this data set worked in
industries typically consisting of a high percentage of women. Similarly, more male
leaders from this data set worked in industries typically consisting of a high percentage of
men. The industry percent of females and functional area percent of females were
correlated; however, the correlation coefﬁcients were low: r=0.17 for male leaders and
r=0.01 for female leaders.

Proportion of ﬁle Raters. This variable represents the percentage of boss, peer,
and subordinate raters (not including self) who were male. This variable was computed
simply by taking the number of male raters and dividing by the total number of raters for

each leader. Table 12 shows that about half of female leaders (48%) had less than 40%

66

male raters (mean % male raters=4l%). Alternatively, most of the male leaders (64%)
had over 60% male raters (mean % male raters=66%).

The proportion of male raters and the percentage of women in industries were
negatively correlated (r=-0.31 for male leaders, r=-0.39 for female leaders) indicating that
as the percent of women working within an industry increased, the proportions of male
raters decreased. The proportion of male raters was also correlated, albeit to a small
degree, with the ﬁinctional area percent of females (r=-0.14 for male leaders, r=-0.13 for
female leaders.

Analyses

With the exception of Hypotheses 10c, 1 1, and 13 which were tested using
Analysis of Variance techniques, most of the hypotheses in this study required
Hierarchical Linear Modeling (HLM) for testing.

HLM. This data set was composed of numerous levels of nested data which made
it suitable for Hierarchical Linear Modeling (HLM) analysis (Raudenbush & Bryk, 2002;
Snijders & Bosker, 1999). HLM is a statistical method that accounts for the
dependencies that occur in the hierarchical nature of nested data (Hofmann, 1997). In the
current study, individual level ratings were nested within each leader, and each leader
was nested within various industries and functional areas. While HLM is similar to
ordinary least squares (OLS) analysis, it has a few advantages in the analysis of cross-
level data. First, because HLM techniques model both the individual and group level
variances, it can assess within and between group differences independently from one
another (Hofmann, 1997). Second, HLM can generally handle missing data and

unbalanced designs (Hofmann, 1997).

67

Multi-rater feedback studies typically aggregate individual level ratings to the
leader level. Those that involve higher level context variables typically disaggregate
those variables to the leader level. Most studies then use some variant of analysis of
variance (multiple ANOVAs or MANOVAs) to study group differences in ratings (see
Lyness & Heilman, 2002, Shore & Thornton, 1986). Many have commented on the
problems with aggregating individual level data to a higher group variable (see Klein &
Kozlowski, 2000). In fact, in a study exploring gender and race differences in
interviewer ratings, Sacco, Scheu, Ryan, and Schmitt (in press) showed that using
analysis of variance techniques can increase the Type I error rate (ﬁnding an effect when
it does not really exist) in nested data structures. They found more signiﬁcant ﬁndings
when using MANOVA procedures as compared to accounting for the clustering in the
data through using HLM analyses. Therefore, they suggest using HLM for clustered data
sets such as the one being used in this study to reduce the risk of Type I error.

HLM also has a few limitations that will need to be addressed in this study. First,
HLM has limited use as a multivariate procedure and at this point, is almost always used
in a univariate paradigm. Therefore, only one dependent variable can be examined at a
time. The hypotheses in this study were designed with this in mind. Also, HLM is
limited in the number of levels and interactions it can reasonably assess. Technically,
this data set was composed of four levels: Individual ratings (level-1) were clustered
within each leader (level-2), each leader was clustered within companies (level—3), and
companies were clustered within industries (level-4). Unfortunately, anything over a
three-level HLM model is extremely rare and beyond the capabilities of most analysis

programs. In addition, this study was conceptually concerned with the contextual aspects

68

of each company that could explain differences between ratings provided to male and
female leaders. This data set did not contain any information about each company other
than the industry it belonged to. Therefore, even if differences were found at the
company level, little theoretical rationale could be provided for the result given the data
available. This study, therefore, disaggregated the level-4 variables (industry and
functional area) to level-3 (company).

The hypotheses offered in this study which required HLM analyses were tested
via four general models (see Figures 2-5). First, Hypotheses l, 2, Sa-b, 6a-b, 7a-b, and
8a—b all proposed a moderation effect of a level-3 context variable (industry or functional
area) on the relationship between a level-2 variable (leader gender) and a level-l set of
ratings (see Figure 2). Second, Hypotheses 3, 5c, 6e, 7e, 8e, 12, and 14 all proposed a
moderation effect of a level-2 variable (usually proportion of male raters) on the
relationship between a level-2 variable (usually leader gender) and a level-1 set of ratings
(see Figure 3). Third, Hypotheses 4 and 10a-b proposed a moderation effect of a level-2
variable (leader gender for example) on the relationship between two level-1 variables
(for example others’ development ratings and others’ effectiveness ratings; see Figure 4).
Finally, Hypothesis 9 proposed a 2-level, 3-way interaction. Rater type (level-1) was
expected to moderate the relationship between agentic ratings (level-l) and effectiveness
ratings (level—1) and also the relationship between communal ratings (level-1) and
effectiveness ratings (level-1). Leader gender (level-2) was then expected to moderate
the effect of rater type. The formulas for each of these four models are also presented in

Figures 2-5.

69

Hypothesis testing in HLM involves the same procedure regardless of the number
of levels or interactions in each model. It involves testing successively more complex
models, with the statistical signiﬁcance of parameters in initial models as a requirement
for ﬁnding signiﬁcance in later models (Sacco, Scheu, Ryan, & Schmitt, in press).
Therefore, the ﬁrst step in HLM hypothesis testing is evaluating the null model (see
Figure 6). The null model contains no predictors and simply shows whether or not there
is a signiﬁcant amount of variance in the outcome variables that can be explained by the
grouping variables. The variance components of the null model error terms can be used
to calculate the intra-class correlation coefﬁcients or ICCs. The ICC represents the
proportion of between group differences due to each grouping variable.

Based on whether or not the between group variances are signiﬁcant, the next step
in HLM hypothesis testing involves adding more predictors to the model, one by one.
Each model is them compared to the subsequent model to determine if the predictor
helped to explain more variance in the outcome variable. Although R2 is typically used
to represent the variance explained in these models, some have argued that R2 is not an
appropriate measure for use in multi-level models (Raudenbush & Bryk, 2002).
However, to date, no other measure has been offered and therefore R2 is still used.

HLM tests cross-level relationships via a regression model, and provides
information about both intercepts and slopes. Although the moderation effects in which
we were most interested in the current study were represented in the slope coefficients
(intercept and slope variance coefficients are labeled in Figures 2-5 for each of the HLM
models being tested in this study), an issue regarding the interpretation of the intercept

coefﬁcients should be pointed out. The interpretation of intercept coefﬁcients can be

70

affected by centering procedures. Although signiﬁcance tests in HLM are not affected by
the method of centering used, centering is a procedure that impacts the interpretability of
the intercept results by altering the scaling of the variables.

There are two forms of centering (in addition to leaving the data in its raw form):
group mean and grand mean. Hofmann and Gavin (1998) note that grand mean centering
produces equivalent models to not centering at all (raw metric), but affects the
interpretation of the intercept coefﬁcients. Group mean centering, alternatively, is not
equivalent to either grand mean centered or raw metric models. Although Hofmann and
Gavin (1998) point out that there is no “right or wrong” method of centering, some
methods are more suited to answering different types of questions.

Kreft, DeLeeuw, and Aiken ( 1995) and Raudenbush and Bryk (2002) recommend
using grand mean centering for two reasons. First, they note that grand mean centering
usually results in the reduction of covariance between the intercepts and slopes, which
can reduce potential problems with multicollinearity (see Hofman & Gavin, 1998). In
addition, Raudenbush and Bryk (2002) note that grand mean centering is appropriate for
dichotomous variables (i.e. gender and leader level as used in this study) in cases where
there are a discrepant number of individuals in each group (which was the case in this
study given that there were more male than female leaders, and more middle managers
than executives). Therefore, all variables in this study were grand mean centered.
Although the intercept values represented different things in the different HLM models
being used in this study, when grand mean centered, it is important to remember that the
intercept coefﬁcients for each grouping variable represented an adjusted group mean as

opposed to a raw group mean.

71

RESULTS

Overview

The speciﬁc analyses for each hypothesis proposed in this study are summarized
in Table 13. First, some basic descriptive information is presented. Following that, some
tests of main effects are described. Although not hypothesized, these main effects tests
were performed to assess some of the trends in the descriptives data most relevant to the
hypotheses. Then, the results of the 3-level and 2-level moderation hypotheses tests are
reported. The relevant hypotheses within each set of results are mentioned where
appropriate. The ﬁnal outcome of each hypothesis test is also summarized in Table 13.

General Descriptives

 

The means and standard deviations for the agentic and communal scales are
provided in Table 14. The data in this table are broken down by rater group, rater gender,
and leader gender. For female leaders, the agentic scale means ranged from 5.42
(SD=0.82) to 5.67 (SD=0.86) and the communal scale means ranged from 5.53
(SD=0.79) to 5.83 (SD=0.52). Female leader self ratings were higher than the ratings
from the other rater groups for the communal scale (self mean=5.83, SD=0.52), followed
by the male subordinate mean rating (mean=5.73, SD=0.81). For the agentic scale, male
subordinates provided female leaders with the highest mean rating (mean=5.67,
SD=0.86).

For male leaders, the agentic scale means ranged from 5.43 (SD=0.84) to 5.66
(SD=0.64) and the communal scale means ranged from 5.45 (SD=0.80) to 5.78
(SD=0.54). Male leader rated themselves higher than did any other rater group on both

scales (agentic mean=5.66, SD=0.64; communal mean=5.78, SD=0.54) followed by

72

female subordinates (agentic mean=5.57, SD=0.95; communal mean=5.56, SD=0.95).
For both male and female leaders, male peers provided the lowest ratings than any other
rater group for both the communal (female leader mean=5.53, SD=0.79; male leader
mean=5.45, SD=0.80) and agentic scales (female leader mean=5.42, SD=0.82; male
leader mean=5.43, SD=0.84).

Across leader groups, female leaders rated themselves slightly higher than male
leaders on the communal scale (female mean=5.83, SD=0.52; male mean=5.78,
SD=0.54) whereas male leaders rated themselves slightly higher on the agentic scale
(male mean=5.66, SD=0.64; female mean=5.63, SD=0.68). All other rater groups rated
female leaders higher than male leaders on both scales with the exception of male peers
who rated male and female leaders similarly on the agentic scale (female leader
mean=5.42, SD=0.82, male leader mean=5.43, SD=0.84).

The most apparent differences between male and female leader ratings came from
male subordinates and female peers. These two groups rated female leaders more highly
than male leaders on both the agentic and communal scales. For the agentic scale, the
mean rating from male subordinates was 5.53 (SD=0.91) for male leaders and 5.67
(SD=0.86) for female leaders. Again for the agentic scale, the mean rating from female
peers was 5.49 (SD=0.88) for male leaders and 5.63 (SD=0.80) for female leaders. For
the communal scale, the mean rating from male subordinates was 5.53 (SD=0.86) for
male leaders and 5.73 (SD=0.81) for female leaders. Again for the communal scale, the
mean rating from female peers was 5.50 (SD=0.88) for male leaders and 5.69 (SD=0.78)
for female leaders. The lowest apparent differences between male and female leader

ratings came from male bosses. Male bosses rated male and female leaders similarly on

73

 

both the agentic (male leader mean=5.56, SD=0.77‘, female leader mean=5.57, SD=0.77)
and communal (male leader mean=5.60, SD=0.71; female leader mean=5.63, SD=0.71)
scales.

The standard deviations for the agentic ratings ranged from 0.68 to 0.92 for
female leaders and from 0.64 to 0.95 for male leaders. The standard deviations for the
communal ratings ranged from 0.52 to 0.89 for female leaders and from 0.54 to 0.95 for
male leaders. Across male and female leaders, the standard deviations are fairly
consistent for each rater group with the female subordinates showing the most amount of
variance (female leader agentic SD=0.92, female leader communal SD=0.89, male leader
agentic SD=0.95, male leader communal SD=0.95) and leader self ratings showing the
least variance (female leader agentic SD=0.68, female leader communal SD=0.52, male
leader agentic SD=0.64, male leader communal SD=0.54).

Overall, both male and female leaders rated themselves higher on the communal
scale than the agentic scale. In addition, the other rater groups tended to rate women
higher on the communal scale as opposed to the agentic scale. Although this was also
true in a few cases for male leaders, the agentic and communal scale means were more
similar for male leaders than for female leaders across most of the rater groups.

Table 14 also shows the means and standard deviations of the effectiveness
ratings provided to male and female leaders by each rater group. The results shown on
the table are consistent with the pattern of results for the communal and agentic scales.
For female leaders, subordinates provided the highest mean ratings (female subordinate
mean=5.90, SD=1.14', male subordinate mean=5.90, SD=1.06) and male peers provided

the lowest mean rating (mean=5.4l, SD=1.09). For male leaders, female subordinates

74

provided the highest mean rating (mean=5.78, SD=1.22) and male peers provided the
lowest mean rating (mean=5.37, SD=1. 13).

Across leader groups, every rater group produced a higher mean rating for female
leaders than male leaders except for the male bosses. The male bosses rated male and
female leaders quite similarly, though the male leader mean rating was slightly higher
(female leader mean=5.62, SD=0.97; male leader mean=5.64, SD=0.98). The biggest
differences between male and female leader ratings came from male subordinates and
female peers. These two groups rated female leaders (male subordinate mean=5.90,
SD=1.06; female peer mean=5.63, SD=1.09) more highly than male leaders (male
subordinate mean=5.72, SD=1.13; female peer mean=5.47, SD=1.17).

The standard deviations for the effectiveness ratings were higher than those for
the communal and agentic ratings, ranging from 0.97 to 1.14 for female leaders and 0.98
to 1.22 for male leaders. However, across male and female leaders, the standard
deviations were fairly consistent for each rater group with the female subordinates
showing the most amount of variance (female leader SD=1. 14, male leader SD=1.22) and
the male bosses showing the least (female leader SD=0.97, male leader SD=0.98).

Main Effects

 

Prior to testing the hypotheses, a few tests of main effects were conducted to
determine the signiﬁcance of some of the relationships noted in the descriptives section
that were relevant to the relationships proposed in the study. One-way multivariate
analysis of variance (MANOVA) and 2-level, main effect HLM analyses were conducted
to examine the main effects of leader gender, leader level, and rater sex on agentic,

communal, and effectiveness ratings.

75

Leader Gender Main Effects. Leader gender was signiﬁcantly related to others’
(subordinates, peers, and bosses) communal, agentic, and effectiveness ratings such that
female leaders were rated more highly (i.e. more effective and needing less development)
than male leaders on every dimension (see Table 15). In addition, female leaders’ self
ratings were signiﬁcantly higher than male leaders’ self ratings of communal skills (i.e.
needing less communal skill development). There were no signiﬁcant differences in self
ratings of agentic skills (see Table 16).

As Tables 15 and 16 show, the R2 values for the main effects of leader gender
range from 0.001 to 0.004. These values are quite low, suggesting that although
signiﬁcant, leader gender described a very small amount of the variance in each set of
ratings.

Leader Level Main Effects. Organizational level was signiﬁcantly related to
others’ (subordinates, peers, and bosses) ratings for the agentic scale only (see Table 15).
Others’ rated executives’ as needing less agentic development than middle managers’. In
addition, although not statistically signiﬁcant, others’ rated executives as needing more
communal development than middle managers. As for self ratings, executives rated
themselves as needing signiﬁcantly less communal and agentic development than did
middle managers (see Table 16).

As with the leader gender main effects, the variance in ratings explained by leader
level was very low (R2 values ranged from 0.0001 to 0.016). This indicates that leader
gender described a very small amount of variance in each set of ratings.

Rater Gender Main Effects. Rater gender was signiﬁcantly related to others’

 

(subordinates, peers, and bosses) communal, agentic, and effectiveness ratings. Female

76

raters rated leaders higher than male raters did on each scale. This indicates that female
raters considered leaders as being more effective and needing less development than did
male raters. The variance in ratings explained by rater gender was low (R2 = 0.004 to
0.005).

HLM Null Models

 

For all hypotheses requiring the use of HLM procedures, the null models were
assessed in order to determine whether or not there was a signiﬁcant amount of variance
in the dependent variables that could be explained by the grouping variables. In the 2-
level models, each ICC represented the proportion of variance in the outcome that was
between groups (in most cases, leader gender). In the level-3 models, the ICC for the
level-2 variance component represented the proportion of variance among the level-2
variables within the level-3 variables (for example, the proportion of variance among
leader genders within industries), while the ICC for the level—3 variance component
represented the proportion of variance among the level-3 variables (industry or functional
area). Table 17 shows the null models for each outcome variable for each hypothesis. As
can be seen in Table 17, of the 70 between group variance components calculated, all
were signiﬁcant except for four level—2 components and one level-3 component. For
Hypothesis 7a, the between gender variance within industry was not signiﬁcant for
female bosses’ communal ratings. For Hypothesis 7b, the between gender variance
within functional area was not signiﬁcant for either female bosses’ communal rating or
for female bosses agentic ratings. In addition, in Hypothesis 7b the variance among
functional areas was not signiﬁcant. Finally, for Hypothesis 7c, the variance among

leader genders was not signiﬁcant for female bosses’ communal ratings.

77

Normally, if the between group variance coefﬁcients are not signiﬁcant, it is a sign that
the grouping variables are unlikely to explain variance in the outcome variables and
therefore the model does not require further testing. However, in this study, all steps for
all hypotheses were completed up to the ﬁnal models for sake of completion. The ﬁnal
models (in their complete form) for all hypotheses requiring HLM analyses are presented
in Tables 18-21.
HLM Model 1: 3-Level, 2-Way Interactions

This ﬁrst set of hypotheses dealt with the moderation effect of context variables
(industry and functional area) on the relationship between leader gender and others’
ratings (see Figure 2). In these hypotheses, others’ individual ratings were level-1
variables. These ratings were all clustered within the second-level variable: leader. The
leaders, in turn, were clustered within the third-level context variables. Therefore, 3-level
HLM analysis was used for each of these hypotheses. Table 18 shows the results for all
of the 3-level, 2-way interaction hypotheses.

Hypotheses 1. 5a. 6a. 7a, and 8a. Hypotheses 1, 5a, 6a, 7a, and 8a all proposed

 

that industry would moderate the relationship between leader gender and others’ ratings.
There were four signiﬁcant interactions resulting from this set of analyses, but three of
the four relationships were not in the direction expected. First, for Hypothesis 1, industry
moderated the relationship between leader gender and others’ development ratings
(R2=0.008 1) such that others’ rated female leaders more positively than male leaders (i.e.
needing less development), especially in industries composed of higher percentages of
women. Although both men and women were rated higher in female dominated

industries, women were rated signiﬁcantly higher than men. Second, the same industry

78

moderation effect was found for others’ effectiveness ratings in Hypothesis 1: others’
rated female leaders as more effective than male leaders, especially in industries
composed of higher percentages of women (R2=0.0069). Third, when looking at
communal development items speciﬁcally (Hypothesis 5a), the same result was found
(R2=0.0067). Others’ rated female leaders as needing less communal development than
male leaders, especially in female dominated industries. All three of these results were
opposite of what was expected. Males were hypothesized to receive higher ratings,
especially in male-dominated settings.

The fourth and ﬁnal signiﬁcant relationship found for this set of hypotheses came
from the tests for Hypothesis 8a. Here again, it was found that women were rated higher,
especially in female-dominated industries. However, in the case of Hypothesis 8a, this
was hypothesized to be true. Female peers rated female leaders more highly than male
leaders on agentic skills (i.e. needing less agentic development), especially in female-
dominated industries (R2=0.0139). No effect was found for communal development
ratings in Hypothesis 8a.

Hypotheses 1L5b, 6b, 7b, and 8b. Hypotheses 1, 5b, 6b, 7b, and 8b all proposed
that functional area would moderate the relationship between leader gender and others’
ratings. There were no signiﬁcant moderation effects found for functional area.

HLM Model 2: 2-Level. 2-Waﬂnteractions

This set of hypotheses all proposed that a level-2 variable (usually proportion of

male raters) would moderate the relationship between a level-2 variable (usually leader

gender) and a level-1 outcome variable (usually others’ ratings). The model for this set

79

of hypotheses can be seen in Figure 3. The results for this set of hypotheses are reported
in Table 19.

Hypotheses 3. 5c, 6c, 7c, and 8c. Hypotheses 3, 5c, 6c, 7c, and 8c all predicted
that the proportion of male raters would moderate the relationship between leader gender
and others’ ratings. The percentage of male raters for each leader was used to represent
the proportion of male raters in this study. There was one signiﬁcant ﬁnding in this set of
hypotheses. As predicted in Hypothesis 8c, female peer effectiveness ratings were higher
for women than for men, and were especially high when the proportion of male raters
was low (R2=0.0059).

Hypothesis 12. Hypothesis 12 predicted that the relationship between self agentic

 

ratings and others’ effectiveness ratings would be moderated by leader gender. In this
model, both leader gender and self—ratings were level-2 variables. Support was not found
for this hypothesis.

Hypothesis 14. Hypothesis 14 predicted that the relationship between leader

 

gender and others’ ratings would be moderated by leaders’ organizational level. Both
leader gender and leader level in this case were level-2 variables. Support was not found
for this model. Although not included in the hypotheses, an analysis was also run to test
for a leader gender by leader level interaction on self-ratings. As in the case with others’
ratings, support was not found for this interaction with regard to development or

effectiveness ratings.

80

HLM Model 3: 2-Level, 2—Way Interactions

These hypotheses all proposed that a level-2 variable (for example, leader gender)
would moderate the relationship between two level-1 variables (such as two sets of
ratings). The general model for this set of hypotheses can be seen in Figure 4. The
results for this set of hypotheses are reported in Table 20.

Hypothesis 4. This hypothesis predicted that leader gender would moderate the

 

relationship between two sets of ratings (others’ effectiveness and development ratings).
Here, each set of others’ ratings were treated as level-l variables and leader gender was
treated as a level-2 moderator. The moderation effect of leader gender was not
signiﬁcant.

Hypotheses lOa-b. Hypothesis 10 predicted that context (industry and functional
area) would have a direct effect on female leaders’ self-ratings. Given that there was
only one self-rating for each leader, self-ratings were modeled at the leader level. In
previous hypotheses, the leader level was described as aleve1-2 variable because the
dependent variables in previous hypotheses were at the individual rater level (level-1). In
Hypotheses lOa-b, however, the leader level was the lowest level of analysis and
therefore became level-1. Similarly, the context variables that were modeled as level-3
variables in previous hypotheses were modeled as level-2 variables in the analyses for
Hypotheses lOa-b. In short, because leader self-ratings and not individual-level others’
ratings were modeled in Hypothesis 10, self-ratings were considered a level-1 variable
and context variables were considered to be level-2.

There was a signiﬁcant moderation effect of industry on the relationship between

leader gender and self agentic ratings. As hypothesized, women rated themselves lower

81

than men (i.e. needing more development) on agentic skills in male dominated industries.
In addition, although no differences were expected between male and female self-ratings
in female-dominated industries, the data showed that females rated themselves higher
(i.e. needing less agentic development) than men in female dominated industries
(R2=0.0292).

HLM Model 4: 2-Level, 3-Way Interaction

There was only one hypothesis which used this model: Hypothesis 9. Hypothesis
9 was a 2-level, 3-way interaction and can be seen in Figure 5.

Hypothesis 9. Hypothesis 9 stated that rater type (boss or subordinate) and leader
gender would interact in their influence on the relationship between agentic and
effectiveness ratings, and on the relationship between communal and effectiveness
ratings. It was expected that the relationship between agentic and effectiveness ratings
would be higher for bosses than for subordinates, more so for men than for women. In
addition, the relationship between communal and effectiveness ratings was expected to be
higher for subordinates than for bosses, more so for men than for women. This
hypothesis was not supported (see Table 21).

Analysis of Variance Tests

Three hypotheses (10c, 1 l, and 13) were tested using Analysis of Variance
analyses. The results for these hypotheses are reported in Table 22.

Hypothesis 10c. Hypothesis 10 predicted that the proportion of male raters would
moderate the relationship between leader gender and self ratings. Given that there was
only one self-rating for each leader, self-ratings were modeled at the leader level. In

previous hypotheses, the leader level was described as a level-2 variable because the

dependent variables in previous hypotheses were at the individual rater level (level-l). In
Hypothesis 10c, however, the leader level was the lowest level of analysis and therefore
became level-l. Similarly, the proportion of male raters was a leader level variable.
Given that all of the variables in this hypothesis were at the same level, HLM was not
required. Rather, a 2-Way MAN OVA was performed. No signiﬁcant interactions were
found for this hypothesis.

Hypothesis 1 l. Hypothesis 1 I predicted that there would be a direct effect of

 

leader gender on leader communal development self-ratings. In this instance, both leader
gender and leader self-ratings were modeled at the leader level. Therefore, a one-way
ANOVA was sufﬁcient to test this hypothesis. The main effect of leader gender on
communal self-ratings was signiﬁcant in the direction hypothesized. Women had higher
self ratings of communal skills (needing less communal development) than did men
(R2=0.002). However, analyses of main effects showed that women also rated
themselves higher on agentic skills (needing less agentic development).

Hypothesis 13. Hypothesis 13 predicted that there would be a difference between
male and female leaders with respect to self-subordinate rater agreement on communal
and agentic development scores. Contrary to the other hypotheses, Hypothesis 13
required that individual level data be aggregated. Speciﬁcally, a self-subordinate rater
agreement index was created for each leader. Although HLM could have been used here
to examine the correlational relationship between self and subordinate ratings, it could
not be used to examine the level of agreement between these sets of ratings. Therefore,
in order to create a measure of self-subordinate agreement, subordinate scores were

averaged for each leader. Then, a difference score was created by subtracting the

83

subordinate average score from the leader self—rating. The absolute value of the
difference score was used as the dependent variable for this hypothesis. Therefore, this
hypothesis test assessed the direct effect of leader gender on self-subordinate agreement.
Both of these variables were modeled at the same level given that they both existed at the
leader level. Since there were no cross-level relationships hypothesized here, a one-way
MANOVA was sufﬁcient to test Hypothesis 13. There was a signiﬁcant main effect of
leader gender on self—subordinate rater agreement on agentic ratings. The level of
agreement for agentic ratings was higher (i.e. difference score was lower) between male
leaders and subordinates than between female leaders and subordinates (R2: 0.002). This
was contrary to the hypothesis that the level of self-subordinate agreement would be

higher for female leaders than for male leaders.

84

DISCUSSION

There are few empirical explanations for the disparity in the numbers of men and
women in leadership positions. Some have pointed to the pipeline explanation, which
states that there are simply not enough women with the appropriate education and
background to ﬁll top leadership roles. However, given that women are receiving
equivalent degrees at the same rate as men and are equally represented at lower levels in
organizations, the pipeline explanation is losing steam (Carli & Eagly, 2001; Wellington,
Kropf, & Gerkovich, 2003). In addition, some researchers have noted that even when
women enter management positions at the same time as men, they progress more slowly
and fail to move as far up the ladder as their male colleagues (Heilman, 1995; Morrison,
White, & Van Velsor, 1992).

The continuing lack of female representation in the executive suite has led some
researchers to examine the role of gender bias and stereotypes in organizational
leadership. Eagly and colleagues (see Eagly & Karau, 2002) have performed a sizeable
amount of work in this area, and have proposed Role Congruity Theory to explain how
gender stereotypes affect women striving for leadership roles in organizations.
Generally, Role Congruity Theory states that stereotypes about women are at odds with
stereotypes about leaders. For instance, women are typically associated with feminine or
communal characteristics such as sensitivity and nurturing, whereas leaders are typically
associated with a more masculine or agentic set of characteristics including assertiveness
and ambition.

Role Congruity Theory also asserts that the organizational context within which

individuals perform may serve to either weaken or enforce the incongruence of gender

85

and leader stereotypes. Speciﬁcally, Eagly and colleagues (see Eagly & Karau, 2002)
have proposed that variables in the work context that make a female’s gender more
salient will result in more bias concerning that woman’s ability to lead. For example,
they suggest that when women work in male-dominated settings, their gender becomes a
novel and distinguishing characteristic that increases the likelihood that they will
experience difﬁculty in attaining leadership roles.

Given the possibility that stereotypes affect the advancement of women in
organizations, it is important to understand how and when these stereotypes are
manifested. Some have proposed that stereotypes reveal themselves through gender bias
in performance evaluations. Unfortunately, the ﬁndings are mixed with regard to the
relationship between gender and performance evaluations. One potential reason for this
is that studies involving this relationship are often performed in lab settings, few of which
include measures of leadership skills. In addition, many of the ﬁeld studies that do exist
were performed in a single organization thereby precluding tests of the possible between-
organization context effects suggested in Role Congruity Theory.

Others have suggested examining the differential leadership development
opportunities provided to men and women in organizations. There are fewer empirical
studies of gender differences in development opportunities than there are of gender
differences in performance evaluation, however, evidence suggests that in some cases
women receive fewer (or lesser quality) development experiences than their male
counterparts.

Although gender differences in performance evaluation and development

opportunities have both been cited as barriers for women hoping for leadership roles, past

86

studies have not examined possible gender differences in ratings provided for
developmental purposes. Given the increasing prevalence of multi-rater development
tools, it is important to understand how what we know about how possible gender
differences in performance evaluations translate into a multi-rater appraisal of leadership
development needs. It would also be helpful to know if there are differences in the types
of leadership skills that leaders are perceived as needing as a function of their gender. In
other words, are women rated as having adequate communal skills and inadequate
agentic type skills? Finally, within the framework of Role Congruity Theory, we can ask
how context variables (such as industry, ﬁinctional area, and proportion of male raters)
affect the relationship between gender and ratings of leader development needs. The
purpose of this study was to investigate the role of gender in the assessment of leadership
development needs using Role Congruity Theory as a foundation. The broader goal of
this research was to contribute to the understanding of how gender plays a role in the
advancement of women in organizations.
Findings

A ﬁeld data set was used in this study to test possible gender differences in
development needs and effectiveness ratings. The data set was large, containing leader
ratings from multiple rater sources in over 200 companies, over 50 different industries,
and nine ﬁanctional areas. There were many beneﬁts to using ﬁeld data for testing the
relationships proposed in this study. First, the ratings provided in this data set were of
real leaders by actual peers, subordinates, and bosses. This allowed us to assess self, top-
down, and bottom-up ratings as opposed to just bottom-up ratings that could be provided

in a lab study. Second, this data set contained variation in the contexts within which the

87

leaders performed as well as variation in the types of ratings (e. g. communal and agentic
development, and effectiveness) provided to individuals. Third, the data set allowed us to
assess ratings provided to leaders at different organizational levels which was an
advantage over past studies which have tended to examine effects at the middle
management level only.

However, despite the size and depth of this particular data set, few signiﬁcant
results were found. In addition, it is important to note that even for the few relationships
in this study that proved to be statistically signiﬁcant, the amounts of variance explained
by the predictor variables were very low. In fact, there was a small amount of between
group (i.e. gender) variance to be explained prior to adding predictors.

It is not uncommon to ﬁnd statistically signiﬁcant relationships that account for a
small amount of variance in a study with large sample sizes. Eagly and Karau (2002)
note that small effect sizes are a relatively common phenomenon in this ﬁeld of research,
and they have claimed that small biases can produce large consequences when repeated
over individuals and occasions. In fact, Martell, Lane, and Emrich (19%) performed
computer simulations which showed that over time, a small bias against women of 1% of
the variance in initial performance ratings produced senior management levels with only
35% women. In addition, they showed that a 5% initial bias produced only 29% senior
women. Martell & Robison-Cox (2002) subsequently found that introducing a small bias
against women at each yearly review readily creates the disproportionately small
numbers of female senior executives that are typical of corporations. Although these
simulation studies do not address the effect of ﬁrst introducing bias at higher levels, (i.e.

after women have already progressed through the middle management ranks), they help

88

tojustify how the chances of women rising to high-level positions in organizations are
greatly reduced even when slight prejudice is consistently acted on.

Despite the propensity of small effect sizes in this type of research, the effects
found in the current study were so small as to be negligible. The largest effects found in
this study explain approximately 0.01% of the variance in ratings whereas the amounts
found in many studies have ranged between 1% to 5% variance at the lowest (Eagly &
Karau, 2002). Therefore, even when comparing the ﬁndings of this study relative to
others which have found small effects, the effects in this study are slight. For this reason,
it is difﬁcult to argue that the few signiﬁcant effects found in this study provide evidence
of gender bias in ratings. However, some of the trends in the data are notable and are
described below.

Main Effects. Although there were no hypotheses predicting main effects of

 

gender on ratings, the data showed that women were rated higher than men by others on
both development dimensions as well as on overall effectiveness. In addition, women
rated themselves signiﬁcantly higher on communal skills than did men. Therefore, the
data suggested a slight propensity for female leaders to be perceived by themselves and
by others as needing less development and being more effective than male leaders This
was an unexpected ﬁnding which served to negate many of the hypotheses offered in this
study given that most of the hypotheses predicted that men would receive higher ratings
than would women.

In addition, there was also a signiﬁcant effect of rater gender on communal,
agentic, and effectiveness ratings such that females provided higher ratings than males.

This ﬁnding did not affect the hypotheses to a great extent given that rater groups were

89

usually not compared with one another. Because HLM can only handle univariate
analyses, the ratings from various groups were either grouped together or ratings from a
speciﬁc group (for instance, “female peers”) were assessed in independent analyses.
Once again, it is important to note that the variances explained by leader gender and rater
gender were small.

Descriptive Data. Although signiﬁcance tests were not performed for all of the
descriptive data presented, there were some interesting trends in the data. One example is
the type of raters who provided the higher ratings to leaders within the sets of male and
female leader ratings. Perhaps not surprisingly, both male and female leader self ratings
were higher than ratings from any other set of raters. However more surprisingly, male
subordinates provided the highest ratings than any other group to female leaders whereas
female subordinates provided the highest ratings to male leaders, albeit the effects were
trivial. This pattern of comparisons is not consistent with the theory of Relational
Demography. Relational Demography Theory would suggest that raters are likely to
provide higher ratings to demographically similar as opposed to demographically
dissimilar ratees.

Another interesting trend in the descriptive data was that female leaders reported
being paid signiﬁcantly less than male leaders even after controlling for age, tenure,
education, organizational level, functional area, and ethnicity. This was the one ﬁnding
in the study that had a notable effect (R2=0.42). This ﬁnding is consistent with those
from other studies showing that even when women are in positions comparable to men,
and have equivalent experience and education, they are paid less (Renner, Rives, &

Bowlin, 2002).

90

Context Effects. A large number of hypotheses in this study predicted that

 

context would play a moderating role in the relationship between gender and some given
set of ratings such that when the leaders’ gender was salient (i.e. leader gender was the
minority), rating differences would be greatest. The types of context variables examined
were industry type, functional area, and proportion of male raters. The information of
interest regarding these context variables was the degree to which they were female or
male dominated.

The set of hypothesis which tests for an industry moderation effect produced four
signiﬁcant interactions that all said the same thing: individuals rated female leaders more
positively than male leaders, especially in female-dominated industries. This was
opposite of what was expected. It was hypothesized that men would receive higher
ratings, especially in male-dominated contexts. Although the effect of this relationship
was very small, and more information is needed to draw any conclusions about rating
bias, the trend in the data is more consistent with a pro-female bias in ratings as opposed
to the pro-male bias that was hypothesized.

Alternatively, the slightly higher ratings of female leaders could be due to true
ability differences. Women may be more competent given that they became leaders
despite working in male dominated contexts, or perhaps because women need to be more
competent in general to reach the same leadership positions as less competent men.
Unfortunately, there are many alternative explanations that could not be controlled for in
this study. It could be that individuals responsible for rating women were

overcompensating in their rating so as not to appear biased. It could also be that

91

individuals rated women based on a different standard than that on which they rated men.
In short, too little is known to draw any strong conclusions about this set of ﬁndings.

The type of effect found to be signiﬁcant with industry was not replicated with
functional area. In fact, this variable failed to produce any effects on the relationship
between leader gender and ratings. Part of the issue here could have been that the
functional area Options on the DLDS are not completely parallel to the occupations they
were paired with from the US. Bureau of Labor Statistics occupational classiﬁcation
system. Although subject matter experts reached consensus on where the functional area
options ﬁt into the classiﬁcation system, these classiﬁcations were very broad.
Therefore, the extent to which the percentages of women working within each functional
area are accurately represented is unclear.

In addition, it is possible that industry and functional area interact such that the
male domination of a ﬁanctional area only has an effect to the extent which that function
is being performed within a male-dominated industry. In other words, a woman working
in human resources in a female-dominated industry may receive different types of ratings
than a woman working in human resources in a male-dominated industry. Although not
tested in this study, this might help to explain the lack of effect of functional area.

Finally, there was one signiﬁcant ﬁnding with regard to the proportion of male
raters context variable. This context variable affected the relationship between leader
i gender and ratings in the same way industry did, which is not surprising given that the
percentage of women working in each industry was moderately correlated with the
proportion of male raters for each leader. Findings showed that females were rated

higher than males, especially when the proportion of male raters was low (i.e. the

92

proportion of female raters was high). Again, it is difﬁcult to draw any conclusions about
why this occurred, but it is clear that this did not provide support for the gender bias in
ratings that was hypothesized.

Kanter (1977) theorized that once a critical mass of the proportion of women in
the workplace is reached, they will no longer be seen as the minority and therefore their
status would be better appreciated. This suggests that perhaps the context effects in this
study would only occur in the most male-dominated of settings. In other words, only in
highly male-dominated settings would females receive lower ratings than men. The full
range of data were analyzed in this study, however, further analyses of only the most
male and female dominated settings may have shown more gender differences in ratings.

(See Lortie-Lussier & Rinfret, 2002 for conflicting ﬁnding).

 

Rater Effects. The only signiﬁcant rater effect found in this study was with
regard to self ratings. As hypothesized, female leaders rated themselves lower on agentic
skills (i.e. needing more agentic development) than did males in male-dominated
industries. However, whereas no gender differences were expected in female-dominated
industries, ﬁndings showed that females rated themselves higher (i.e. needing less agentic
development) than men in female-dominated industries. This might suggest that women
are more confident in their level of agentic skills in female-dominated industries, and
interestingly, that men are less conﬁdent about their agentic skills in female dominated
industries. Alternatively, women may not believe that agentic skills are as necessary in
female dominated industries, therefore they did not focus on their development needs in
that skill area. The ﬁnding with regard to men in female-dominated industries is not so

straight forward.

One possibility is that the roles of male leaders in female dominated industries
differ from those of women. Men may be more likely to be responsible for traditional
leadership duties (e. g. providing strategic direction) whereas women may be more
responsible for communal duties (e. g. mentoring and developing others). If this is the
case, then perhaps focusing on agentic development is more adaptive for men given the
types of leadership tasks they perform.

Another possibility is that male leaders’ traditional approaches to leadership have
failed in female dominated contexts. Yet, rather than considering a shift in leadership
style congruent with the context within which they are performing, they instead focus on
improving traditionally masculine leadership skills. This implies that men are less likely
to recognize the situations in which communal leadership skills are needed, instead
believing that agentic leadership will be effective across situations if only developed
more ﬁilly.

Limitations

Although the data used in this study provided a unique opportunity to examine
possible gender differences in development and effectiveness ratings in light of Role
Congruity Theory, it had some limitations that precluded drawing conclusions about the
role of rating bias in the glass ceiling effect.

Nature of Rating Dimensions. Although every effort was made to create two,
conceptually distinct development scales, those used in this study were highly correlated
with one another. In addition, the development and effectiveness items were also highly
correlated. This essentially created a unidimensional measure of leadership.

Unfortunately, this prevented us from making conclusions about how perceptions of

94

development needs related to overall leadership effectiveness. In this particular data set,
good leaders received high ratings across development and effectiveness items which
raises questions about whether or not responses from this type of instrument would be
useful in future studies attempting to compare development and performance ratings.
Raters need to be able to make a distinction in the types of skills being rated. Ideally,
these data would have included responses to a second, distinct measure of effectiveness
in order to better compare perceptions of performance with perceptions of development
needs.

Although the effectiveness items required raters to evaluate the overall
effectiveness of each leader, these items were collected in a manner consistent with the
development items. In other words, the items assessing overall leadership effectiveness
used the same rating scale as the development items, were given on the same survey at
the same time as the development items, were reported back to the leaders in the same
manner as the development ratings (conﬁdentially), and were not tied to any
performance-related incentives such as pay raises or promotions. In addition, the same
respondents provided information about both leadership development needs and
effectiveness. These factors suggest a common method bias in the data set.

Some meta-analyses indicate that the concern about inﬂated correlations due to
common method bias may not be as large a problem as Campbell (1982) noted. For
instance, Crampton and Wagner (1994) found that the “percept-percept inﬂation may be
more the exception than the rule in micro-research on organizations” (p.72). Doty and
Glick (1998) also state that the biasing effect of common method variance to distort true

correlations among constructs usually does not invalidate our theoretical interpretations

95

and research conclusions. Yet, Crampton & Wagner (1994) also found that correlations
among leadership behaviors are more susceptible to common methods bias than other
areas of research. Therefore, using caution in interpreting results is warranted. Given the
strong relationship between scales in this study, conclusions can really only be drawn
about individuals’ perceptions of leadership performance in general. We cannot make
conclusions about differences in distinct skill and effectiveness ratings.

Leader Characteristics. Another limitation of this study was the lack of
information regarding how leaders came to be in their industries and ﬁinctional areas.
The women working in the male dominated industries or ﬁanctional areas may have
different characteristics from women working in the female dominated industries or
functional areas. The possibility exists that women working in male dominated contexts
self-selected into those jobs because they are more similar to the men in those contexts.

It is also possible that the women who were hired into these male-dominated contexts
necessarily had to be high performers to be selected in the ﬁrst place, which would help
explain why these women were rated as highly as men in male-dominated industries. For
instance, if the women working in male dominated industries have the agentic skills
required within that setting, and are also perceived as having good communal skills as a
function of being women, then perhaps they receive higher ratings than men because they
not only possess traditional leadership characteristics, but also display interpersonal and
supportive leadership.

ﬁndards of Performance. The tool used in this data set was subjective, and
therefore, the data did not contain information regarding what or who raters were

comparing each leader against. It is also important to emphasize that leaders were

96

 

responsible for choosing their own raters. It is also clear from the descriptive data that
women had more female raters than did men, suggesting that women were more likely to
choose female raters and men were more likely to choose male raters. If leaders chose
raters who they thought would rate them more positively, then the variance in ratings was
likely reduced as compared to a situation in which raters were chosen at random.

Shifting standards theory (Biernat & Fuegen, 2001; Biernat, Manis, & Nelson,
1991; Biernat & Vescio, 2002) states that in the case of subjective measures as used in
this study, raters compare each gender to others within that same gender group.
Therefore, a good female leader is not the same as a good male leader, rather, they are
both good only relative to others in their own gender groups. Those adhering to this
theory would say that gender stereotypes will only show through when you can be sure
that individuals are rating leaders of both genders against the same standard of
performance. This suggests that the types of rating forms used for research should
contain relative comparison information such as a behaviorally anchored response scale.

We also had no information regarding the speciﬁc types of leadership tasks that
leaders were performing. It is possible that if the women in these organizations are
performing less complex, less challenging, or lower-proﬁle leadership tasks than their
male counterparts, then despite comparable ratings, they will still be seen as less qualiﬁed
for higher level roles during the promotion process.
Future Research

Although this study did not provide evidence of gender bias in ratings of
development needs or effectiveness, inferences could not be made concerning whether or

not bias exists in some other aspect of leader development or promotion processes. In

97

other words, gender stereotyping may play a role in the glass ceiling phenomenon;
however, results of this study suggest that those stereotypes are not manifested in the
types of 360 degree leadership development ratings used here. In addition, little support
was found for the principles of Role Congruity Theory. The context variables used in
this study did not have the anticipated moderation effects on the relationship between
leader gender and ratings. However, this is not to imply that Role Congruity Theory
would not hold in a study investigating different types of context variables. The idea that
organizational context inﬂuences the development of female leaders, thereby affecting
women’s progression up the organizational ladder, is still a worthy subject of
investigation. Future research should investigate organizational context from a different
perspective. For example, the gender composition of work groups or departments could
be examined. Also, it would be helpful to measure the level of perceived gender
stereotypes within organizations. Although this might be difﬁcult in a ﬁeld setting, a lab
study could be designed in which the degree of agentic and communal contextual
characteristics are manipulated, allowing researchers to have more control over context
effects.

Promotion Criteria. Although investigating performance ratings and development
needs can tell us part of the glass ceiling story, it is important to better understand the
whole story. In other words, what is the full set of both formal and informal criteria used
when making promotion decisions? Although ratings are typically used, and are an
important part of most promotion systems, ratings alone do not predict who will be
promoted. If the various components of promotion processes were better understood,

then perhaps there would be more avenues through which to investigate the role of

98

stereotypes as they relate to leadership advancement. In the current study, the ratings
provided were not directly linked to promotion decisions. Rather, they were linked to the
feedback individuals received about where they need to improve their skills. Therefore,
we are unable to make any direct inferences about how rating bias (or lack thereof) helps
us to understand the glass ceiling effect.

Although we have female leaders representing all organizational levels, we had no
progression information. In other words, we had no information about how long it took
leaders to advance into their current roles or how long they have been in those roles. We
also could not assume that the raters who provided information in this study would have
provided the same types of information had this tool been used in making promotion
decisions.

The case of Price Waterhouse vs. Hopkins is relevant here. Essentially, Hopkins
was denied promotion despite having ratings that were equivalent to her male peers who
did get promoted. She was partly denied promotion because her bosses met informally
and decided that she lacked the appropriate communal skills. Although part of the
promotion process, it was likely unknown to employees that meetings such as this played
such as large role in promotion decisions. This case provides anecdotal support that a
woman who ﬁlls a leader role may elicit negative reactions even while receiving positive
performance evaluations (Eagly & Karau, 2002), and that these negative reactions can
manifest in other ways (possibly through informal mechanisms) that effect the
advancement of women.

Development Opportunities. Given that no evidence of gender bias in ratings

 

resulted from this study, more questions can be asked about where and when bias may

99

play a part in the progression of women into leadership roles. In this study, men and
women were given similar development feedback. However, the types of learning and
performance opportunities men and women are given in order to improve their skills may
still differ.

The results of a recent survey given to 825 Fortune 1000 CEO’S and women
executives (vice president level and above) showed that both men and women think that
the largest barrier to the advancement of women is lack of line experience (deﬁned in this
case as those with proﬁt and loss experience) (Wellington, Kropf, & Gerkovich, 2003).
The second largest barrier cited by women in the survey is exclusion from informal
networks. Interestingly, exclusion from informal networks was not rated highly by men
as a barrier for women. This suggests that men have a difﬁcult time recognizing when
women are not provided with the same informal opportunities that they are. Men may
also be underestimating, or taking for granted the importance of informal opportunities.
Third, stereotypes about women’s roles and abilities were cited by both men and women
to be a likely barrier to the advancement of women. Despite the lack of ﬁndings in the
current study, the fact that individuals are aware of leadership stereotypes and admit that
that they may present a barrier for women lends credence to the idea that these biases do
exist and warrant further research.

This anecdotal evidence supports other research suggesting that the
developmental opportunities afforded women are different, and often detrimentally so,
from those provided to men. For instance Lyness (2002) and others (see Morrison & Von

Glinow, 1990; Ragins & Cotton, 1991) have noted that women are less likely than men to

100

receive mentoring opportunities, challenging assignments, performance feedback, and
access to informal social networks within their organizations.

McCauley (2001) notes that the leadership development process has three
components: developmental experiences, the ability to learn, and an organizational
context that supports development. Opportunities to learn are provided via
developmental experiences, and the context of the organization can affect the quality and
quantity of those opportunities. McCauley also states that people learn the most under
high challenge situations. If the assertions of research are true, and women are not
receiving these opportunities or challenges, then perhaps they will be viewed as less
qualiﬁed than men for leadership roles.

In addition to opportunities to learn, Ford and colleagues (Ford, Quinones, Sego,
& Sorra, 1992, Quinones, Ford, Sego, & Smith, 1995) have pointed out the importance of
opportunities to perform to enable individuals to transfer the skills they have learned.
They specify three dimensions of opportunity to perform: breadth, activity level, and type
of tasks performed. They performed a study using an Air Force sample, and found that
airmen received differential opportunities to perform. They also found that context
variables affected whether or not opportunities were provided. Speciﬁcally, a
supervisor’s perceptions of the airman’s capability, skills, and likability affected the
likelihood of that airman receiving opportunities to perform trained tasks on the job
(Ford, et a1, 1992). In the context of gender bias, this may mean that regardless of ability,
if the stereotype that females are incapable of performing important leadership roles is

believed, then females are unlikely to receive opportunities to controvert the stereotype.

lOl

In future studies, it would be helpful to know what types of organizational context
variables make development opportunities more or less similar for men and women. For
instance, it would be helpful to have information about the organizational culture with
regard to leader development and promotion choice points. This may help us to assess
how the opportunities offered to men and women may differ in such a way as to create
differential qualiﬁcations for higher level roles.

ChangingNature of Leadership. Although many have commented that the
number of female executives is disproportionately low, the number of female middle
managers is on the rise. Some have attributed this to the changing nature of leadership
roles. Diekman and Eagly (2000), for instance, introduce the idea of dynamic
stereotypes. The Theory of Dynamic Stereotypes suggests that as the social roles of
groups change, so do stereotypes about those groups. Given the notion provided by
Social Role Theory that the role behavior of group members shapes the group stereotype,
groups should have dynamic stereotypes to the extent that their typical social roles are
perceived to change over time.

Applied to gender and leadership, this theory makes two predictions about
perceived change. First, perceivers should think that gender differences are diminishing
because of increasing similarity in the roles of men and women (both in society and at
work). Second, the female stereotype should be particularly prone to change given the
large scale changes that have occurred in women’s social roles over the past 50 years.
Based on the theory of dynamic stereotypes, one could make the argument that roles are
converging and as higher level leadership roles begin to take on a more communal

characterization, women will be more likely to be promoted into these roles.

102

It would be interesting to study the ways in which leader roles may already be
changing toward more androgynous deﬁnitions (Schein, 2001). Given the landscape of
today’s organizations (e. g. rapidly changing, globally competitive, high-technology), it
would be worthy of future research to understand how leadership roles are adjusting.
Whereas the traditional autocratic and directive style of management may have worked
well under more hierarchical organizational structures, it may be the case that a new style
(perhaps one more consistent with a participatory, delegatory, and open-communication
structure) is better suited to modern organizations. The implications of a changing ideal
for leadership may have direct implications on the advancement of women into such roles
Eagly & Karau, 2002).

Furthermore, as globalization continues, organizations are likely to be faced with
the cultural impact of gender stereotypes with regard to leadership. From a research
perspective, it will be important to investigate how gender stereotypes affect the
advancement of women in cultures where stereotypes are more or less prevalent than they
are in the United States. Additionally, with regard to practical implications, female
leaders working overseas could beneﬁt from a better understanding of how cultural
contexts interact with stereotypes about their leadership abilities. This may be most
_ important in cases where women in important leadership roles are placed within countries
where female leaders and communal approaches to leadership are not well accepted.
Conclusion

In sum, this study did not provide evidence of gender bias in leadership
development and effectiveness ratings. In addition, the results of this study showed little

support for the moderating effect of context variables on the relationship between leader

103

gender and ratings as suggested by Role Congruity Theory. There were some limitations
to this study that could possibly have impacted our ability to ﬁnd evidence of rating bias,
and have precluded our ability to speak directly to the glass ceiling phenomenon.
However, it is important to note that the lack of ﬁndings with regard to gender
differences in a data set of this size and quality tells us that perhaps bias is not
manifesting itself in ratings, but perhaps in some other stage of the
development/advancement process. The results of this study might also suggest that
gender bias simply does not exist in this context.

It will be important for future studies to begin where this study leﬁ off, and
examine points at which gender stereotypes may play a role after development needs are
assessed. Future research should focus on furthering our understanding of the criteria
used for promotion decisions and the potential impact of gender bias on those decision
processes. In addition, future research should address the possible gender differences in
employee development, speciﬁcally in the areas of opportunities to learn and perform.
Finally, it is important to expand our investigation of various contextual characteristics
that serve to either weaken or reinforce the incongruity between gender and leader

stereotypes.

104

REFERENCES

Alexander, L. D. (1979). The effect level in the hierarchy and functional area
have on the extent Mintzberg's roles are required by managerial jobs. Academy of
Management Proceedings. 39, 186-189.

_Allport, G. W. (1954). The nature of prejudice. Reading, MA: Addison-Wesley.

Antonioni, D. (1996). Designing an effective 360-degree appraisal feedback
process. Organizational Dynamics, 25, 24-3 8.

Atwater, L. (1998). The advantages and pitfalls of self-assessment in
organizations. In J. W. Smither (Ed), Performance appraisal: State of the art in prgctice
(pp. 331-369). San Francisco, CA: Jossey-Bass.

 

Atwater, L., Ostroff, C., Yammarino, F. J ., & F leenor, J. W. (1998). Self-other
agreement: Does it really matter? Personnel Psychology. 51, 577-598.

Atwater, L., Roush, P., & Fischthal, A. (1995). The inﬂuence of upward feedback
on self and follower ratings of leadership. Personnel Psychology. 48. 35—59.

Atwater, L. E., & Waldman, D. (1998). Accountability in 360-degree feedback: Is
it time to take the 360-degree feedback method to its next step? HRMagazine, 43, 96-
102.

Bakan, D. (1966). The duality of human existence: An essay on psychology and
religion. Chicago: Rand McNally.

Banaji, M. R., & Hardin, C. D. (1996). Automatic stereotyping. Psychological
Science 7, 136-141.

 

Banaji, M. R., Hardin, C. D., & Rothman, A. J. (1993). Implicit stereotyping in
person judgment. Journal of Personglity_and Social Psychology, 65, 272-281.

 

Bartol, K. M. (1999). Gender inﬂuences on performance evaluations. In G. N.
Powell (Ed), Handbook of gender & work (pp. 165-178). Thousand Oaks, CA: Sage.

 

Bartol, K. M., & Butterﬁeld, D. A. (1976). Sex effects in evaluating leaders.
Journ_al of Applied Psychology. 61, 446-454.

Bass, B. M. (1981). Individual capability, team response, and productivity. In E.

A. Fleishman & M. D. Dunnette (Eds), Human performance and productivity. Hillsdale,
NJ: Lawrence Erlbaum Associates.

105

‘ Bass, B. M. (1990). Bass & Stogdill's handbook of leadership: Theory, research,
and managerial applications (3rd ed). New York: Free Press.

Becker, J ., Ayman, R. A., & Korabik, K. (2002). Discrepencies in
self/subordinates' perceptions of leadership behavior: leader's gender, organizational
context, and leader's self-monitoring. Group & Organizational Management. 27. 226-244.

Bennett, S. K. (1982). Student perceptions of and expectations for male and
female instructors: Evidence relating to the question of gender bias in teaching
evaluation. Journal of Educational Psychology. 74. 170-179.

Biddle, B. J. (1979). Role theory: Expectations, identities, and behaviors. New
York: Academic Press.

Biernat, M., & Fuegen, K. (2001). Shifting standards and the evaluation of
competence: Complexity of gender-based judgment and decision making. Journal of
Social Issues. 57, 707-724.

Biernat, M., Manis, M., & Nelson, T. E. (1991). Stereotypes and standards of
judgment. Journal of Personality and Social Psychology, 60. 485-499.

Biernat, M., & Vescio, T. K. (2002). She swings, she hits, she's great, she's
benched: Implications of gender-based shifting standards for judgment and behavior.
Personality and Social Psychology Bulletin, 28, 66-77.

Bigoness, W. J. (1976). Effect of applicant’s sex, race, and performance on
employers’ performance ratings: Some additional ﬁndings. Journgl of Applied
Psychology, 61. 80-84.

Blair, 1. V., & Banaji, M. R. (1996). Automatic and controlled processes in
stereotype priming. Journal of Personality and Social Psychology. 70. 1142-1163.

Brenner, O. C., Tomkiewicz, J., & Schein, V. E. (1989). The relationship between
sex role stereotypes and requisite management characteristics revisited. Academy of
Management Journal, 32. 662-669.

Bowen, C., Swim, J. K., & Jacobs, R. R. (2000). Evaluating gender biases on
actual job performance of real people: A meta-analysis. Journal of Applied Social
Psychology. 30. 2194-2215.

Buenger, V., Daft, R. L., Conlon, E. J ., & Austin, J. (1996). Competing values in

organizations: Contextual inﬂuences and structural consequences. Organization Science.
2, 557-576.

106

Butterﬁeld, D. A., & Grinnell, J. P. (1999). “Re-viewing" gender, leadership, and
managerial behavior: Do three decades of research tell us anything? In G. N. Powell (Ed)
Handbook of gender & work. (pp. 223-23 8). Thousand Oaks, CA: Sage.

Butterﬁeld, D. A., & Powell, G. N. (1981). Effects of group performance, leader
sex, and rater sex on ratings of leader behavior. Organizational Behavior and Human
Performance, 28, 129-141.

 

 

Campbell, J. P. (1982). Editorial: Some remarks from the outgoing editor. Journal
of Applied Psychology. 67, 691-700.

Carbonell, J. L. (1984). Sex roles and leadership revisited. Journal of Applied
Psychology, 69, 44-49.

 

Carli, L. L., & Eagly, A. H. (2001). Gender, hierarchy, and leadership: An
introduction. Journal of Social Issues, 57, 629-636.

Catalyst. (2000). Census of women corporate ofﬁcers and top earners. New York:
Catalyst.

 

Cho, H. J. (2000). Diagnosing leadership development: Validating a model and
method. Ann Arbor, MI: Denison Consulting.

Cleveland, J. N., Stockdale, M., & Murphy, K. R. (2000). Women and men in
grganizations: Sex and gender issues at work. Mahwah, NJ: Lawrence Erlbaum.

Copeland, C. L., Driskell, J. E., & Salas, E. (1995). Gender and reactions to
dominance. Journal of Social Behavior and PersonalityLIO, 53-68.

Crampton, S. M., & Wagner 111, J. A. (1994). Percept-percept inﬂation in
microorganizational research: An investigation of prevalence and effect. Journal of
Applied Psychology, 79, 67-76.

Davison, H. K., & Burke, M. J. (2000). Sex discrimination in simulated
employment contexts: A meta-analytic investigation. Journ_al of Vocaticm Behgyior. 56,
225-248.

Deaux, K. (1979). Self-evaluations of male and female managers. Sex Roles. 5,
571-580.

Deming, W. E. (1986). Out of the crisis. Cambridge, MA: Massachusetts Institute
of Technology, Center for Advanced Engineering Study.

 

Denison, D. R., Hooijberg, R., & Quinn, R. E. (1995). Paradox and performance:
Toward a theory of behavioral complexity in managerial leadership. Organization
Science 6, 524-540.

 

107

Denison, D. R. & Neale, W. S. (1996). The Denison leadership development
survey. Ann Arbor, MI: Aviat.

Denison, D. R., & Neale, W. S. (2001). Denison leadership development survey:
Facilitator guide. Denison Consulting, LLC: Ann Arbor, MI.

 

Denmark, F. L. (1993). Women, leadership, and empowerment. Psychology of
Women Quarterly, 17, 343-356.

Diekman, A. B., & Eagly, A. H. (2000). Stereotypes as dynamic constructs:
Women and men of the past, present, and future. Personalityﬁand Social Psychology
Bulletin, 26, 1171-1188.

Dipboye, R. L. (1985). Some neglected variables in research on discrimination in
appraisals. Academy of Management Review, 10, 116-127.

Dobbins, G. H., & Platz, S. J. (1986). Sex differences in leadership: How real are
they? Academy of Management Review, 1 1, 1 18-127.

Dodge, K. A., Gilroy, F. D., & Fenzel, L. M. (1995). Requisite management
characteristics revisited: Two decades later. Journal of Social Behavior and Personality,
m, 253-264.

Doty, D. H., & Glick, W. H. (1998). Common methods bias: Does common
methods variance really bias results? Organizational Research Methods. 1, 374-406.

 

Eagly, A. H. (1987). Sex differences in social behavior: A social-role
interpretation. Hillsdale, NJ: Erlbaum.

Eagly, A. H, & Johnson, B. T. (1990). Gender and leadership style: A meta-
analysis. Psychological Bulletin, 108, 233-256.

Eagly, A. H., & Karau, S. J. (2002). Role congruity theory of prejudice toward
female leaders. Psychological Review, 109, 573-598.

 

Eagly, A. H., Karau, S. J ., & Makhijani, M. G. (1995). Gender and effectiveness
of leaders: A meta-analysis. Psychologgml Bulletin, 117, 125-145.

 

Eagly, A. H, Makhijani, M. G., & Klonsky, B. G. (1992). Gender and the
evaluation of leaders: A meta-analysis. Psychological Bulletin, 111, 3-22.

Eagly, A. H., & Steffen, V. J. (1984). Gender stereotypes stem from the
distribution of women and men into social roles. Journal of Personality and Social
Psychology, 46, 73 5-754.

 

108

Eagly, A. H., Wood, W., & Diekman, A. B. (2000). Social role theory of sex
differences and similarities: A current appraisal. In T. Eckes & H. M. Trautner (Eds),
The developmental social psychology of gender (pp. 123-174). Mahwah, NJ: Erlbaum.

Esses, V. M., Haddock, G., & Zanna, M. P. (1993). Values, stereotypes, and
emotions as determinants of intergroup attitudes. In D. M. Mackie & D. L. Hamilton

(Eds), Affectgcognition, and stereotyping: Interactive processes in group perception (pp.
137-166). San Diego, CA: Academic Press.

Fiske, S. T., Bersoff, D. N., Borgida, E., Deaux, K., & Heilman, M. E. (1991).
Social science research on trial: The use of sex stereotyping research in Price Waterhouse
vs. Hopkins. American Psychologist. 46, 1049-1060.

Fiske, A. P., Haslam, N., & F iske, S. T. (1991). Conﬁising one person with
another: What errors reveal about the elementary forms of social relations. Journal of
Personality and Social Psychology, 60, 656-674.

Ford, J. K., Quinones, M. A., Sego, D. J., & Sorra, J. S. (1992). Factors affecting
the opportunity to perform trained tasks on the job. Personnel Psychology, 45, 511-527.

 

Frank, F. D., & Drucker, J. (1977). The inﬂuence of evaluatee’s sex on
evaluations of a response on a managerial selection instrument. Sex Roles. 3, 59-64.

 

Freedman, S. M., & Phillips, J. S. (1988). The changing nature of research on
women at work. Journal of Management, 14, 231-251.

Garvin, D. A. (1993). Building a learning organization. Harvard Business Review,
7_1_, 78-91.

Geis, F. L. (1993). Self-ﬁilﬁlling prophecies: A social psychological view of
gender. In A. E. Beall & R. J. Stemberg (Eds), The psychology of gender (pp. 9-54).
New York: Guilford Press.

 

Giannantonio, C. M., Olian, J. D., & Carroll, S. J. (1995). An experimental study
of gender and situational effects in a performance evaluation of a manager. Psychological
Reports, 76, 1004-1006.

Glick, P., Zion, C., & Nelson, C. (1988). What mediates sex discrimination in
hiring decisions? Jougal of Persopalityaand SociaLPsycholcagy. 55, 178-186.

Goldberg, P. (1968). Are women prejudiced against women? Transaction, 5, 316-
322.

Gomez-Mejia, L. R., McCann, J. E., & Page, R. C. (1985). The structure of
managerial behaviors and rewards. Industrial Relations, 24, 147-154.

109

Greguras, G. 1., & Robie, C. (1998). A new look at within-source interrater
reliability of 360-degree feedback ratings. Journal of Applied Psychology, 83, 960-968.

Griffeth, R. W., & Bedeian, A. G. (1989). Employee performance evaluations:
Effects of ratee age, rater age, and ratee gender. Journal of Organizational Behavior. 10,
83-90.

Gutek, B. A., & Morasch, B. (1982). Sex-ratios, sex-role spillover, and sexual
harassment of women at work. Journal of Social Issuesﬁ8, 55-74.

 

Hackman, J. R., & Wageman, R. (1995). Total quality management: Empirical,
conceptual, and practical issues. Administrative Science Quarterly. 40, 309-342.

Hall, F. S., & Hall, D. T. (1976). Effects of job incumbents’ race and sex on
evaluations of managerial performance. Academy of Management Journal, 19, 476-481.

Hamner, W. C., Kim, J. S., Baird, L., & Bigoness, W. J. (1974). Race and sex as
determinants of ratings by potential employers in a simulated work-sampling task.
Journal of Applied Psychology, 59, 705-711.

Hart, S. L., & Quinn, R. E. (1993). Roles executives play: CEOs, behavioral
complexity, and ﬁrm performance. Human Relations, 46, 543-574.

Hazucha, J. F., Hezlett, S. A., & Schneider, R. L. (1993). The impact of 360-
degree feedback on management skill development. Human Resource Management. 32,
325-352.

Heilman, M. E. (1983). Sex bias in work settings: The lack of ﬁt model. In B.
Staw and L. Cummings (Eds), Research in organizational behavior (Vol. 5). Greenwich,
CT: JAI.

 

Heilman, M. E. (1995). Sex stereotypes and their effects in the workplace: What
we know and what we don’t know. Journal of Social Behavior and Personality, 10, 3-26.

Heilman, M. E. (2001). Description and prescription: How gender stereotypes
prevent women’s ascent up the organizational ladder. Journal of Social Issues. 57, 657-
674.

Heilman, M. E., Block, C., & Martell, R. (1995). Sex stereotypes: Do they
inﬂuence perceptions of managers? Jouraal of Social Behavior and Persppality, 10, 23 7-
252.

Heilman, M. E., Block, C. J., Martell, R. F., & Simon, M. C. (1989). Has anything

changed? Current characterizations of men, women, and managers. Journal of Applied
Psychology. 74, 935-942.

110

Hellriegel, D., Jackson, S. E., & Slocum, J. W., Jr. (2002). Management: A
competencL-based approach. Cincinnati, OH: South-Western.

 

Hofmann, D. A. (1997). An overview of the logic and rationale of hierarchical
linear models. Journal of Management, 23, 723-744.

 

Hofmann, D. A., & Gavin, M. B. (1998). Centering decisions in hierarchical
linear models: Implications for research in organizations. Journal of Management. 24,
623-641.

Hooijberg, R. (1996). A multidirectional approach toward leadership: An
extension of the concept of behavioral complexity. Human Relations. 49, 917-946.

Hooijberg, R., & Choi, J. (2000). Which leadership roles matter to whom? In
examination of rater effects on perceptions of effectiveness. Leadership Quarterly. 11,
341-364.

Hooijberg, R., & Denison, D. R. (2002). What makes leaders effective? A
stakeholder approach to leadership effectiveness. Unpublished manuscript.

 

Howell, J. M., & Avolio, B. J. (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, J. M., & Higgins, C. A. (1990). Champions of technological innovation.
Administrative Science Quarterly, 35, 317-341.

Hunt, 1. G., Boal, K. B., & Sorenson, R. L. (1990). Top management leadership:
Inside the black box. Leadership Quarterly, 1, 41-65.

Izraeli, D. N., & Izraeli, D. (1985). Sex effects in evaluating leaders: A replication
study. Journal of Applied Psychology, 70, 540-546.

Jacobson, M. B., & Effertz, J. (1974). Sex roles and leadership: Perceptions of the
leader and the led. O_rganizational Behavior and Human Performance. 12, 383-396.

Jago, A. G., & Vroom, V. H. (1982). Sex differences in the incidence and
evaluation of participative leader behavior. Journal of Applied Psychology, 67, 776-783.

Johnson, J. W., & Ferstl, K. L. (1999). The effects of interrater and self-other
agreement on performance improvement following upward feedback. Personnel
Psychology, 52. 272-303.

Juran, J. M. (1988). Juran on planning for quality. New York: Free Press.

 

lll

Kanter, R. M. (1977). Men and women of the corporation. New York: Basic
Books.

Kaufmann, G., Isaksen, S. G., & Lauer, K. (1996). Testing the "glass ceiling"
effect on gender differences in upper level management: The case of innovator
orientation. European Journal of Work and Organizational Psychology, 5, 29-41.

Kirnan, J. P., Farley, J. A., Geisinger, K. F. (1989). The relationship between
recruiting source, applicant quality, and hire performance: An analysis by sex, ethnicity,
and age. Personnel Psychology, 42, 293-308.

Klein, K. J., & Kozlowski, S. W. J. (Eds). (2000). Multilevel theory. research.
and methods in organizations: Foundations. extensions, and new directions. San
Francisco, CA: Jossey-Bass.

 

Knight, P. A., & Saal, F. E. (1984). Effects of gender differences and selection
agent expertise on leader inﬂuence and performance evaluations. Organizational
Behaviorgand Human Performance, 34, 225-243.

Korabik, K. (1990). Androgyny and leadership style. Jourpai of Business Ethics,
9, 9-18.

Korabik, K., Baril, G. L., & Watson, C. (1993). Managers' conﬂict management
style and leadership effectiveness: The moderating effects of gender. Sex Roles, 29, 405-
420.

Kreft, I. G. G., DeLeeuw, J ., & Aiken, L. S. (1995). The effect of different forms
of centering in Hierarchical Linear Models. Multivariate Behavioral Research. 30. 1-21.

Lauterbach, K. E., & Weiner, B. J. (1996). Dynamics of upward inﬂuence: How
male and female managers get their way. Leadership Quarteiy. 7. 87-107.

Lee, D. M., & Alvarez, K. M. (1977). Effects of sex on descriptions and
evaluations of supervisory behavior in a simulated industrial setting. Journal of Applied
Psychology. 62, 405-410.

Lee, J., & Hoon, T. H. (1993). Business students' perceptions of women in
management-The case in Singapore. Management Education and Development. 24, 415-
429.

Leuptow, L. B., Garovich, L., & Leuptow, M. B. (1995). The persistence of
gender stereotypes in the face of changing sex roles: Evidence contrary to the

sociocultural model. Ethology and Sociobiology. 16. 509-530.

London, M., & Beatty, R. W. (1993). 360-degree feedback as a competitive
advantage. Human Resource Management, 32, 352-373.

112

London, M., & Poplawski, J. R. (1976). Effects of information on stereotype

development in performance appraisal and interview context. Journal of Applied
Psychology. 61. 199-205.

 

London, M., & Smither, J. W. (1995). Can multi-source feedback change
perceptions of goal accomplishment, self-evaluations, and performance-related
outcomes? Theory-based applications and directions for ﬁiture research. Personnel

Psychology, 48, 803-841.

London, M., & Stumpf, S. A. (1983). Effects of candidate characteristics on I
management promotion decisions: An experimental study. Personnel Psychology, 36,
241-259.

 

London, M., & Wohlers, A. J. (1991). Agreement between subordinate and self-
ratings in upward feedback. Personnel PsychologyL44, 375-390.

Lord, R. G, Foti, R. J ., & de Vader, C. L. (1984). A test of leadership
categorization theory: Internal structure, information processing, and leadership
perceptions. Organizational Behavior and Human Decision Processes. 34. 343-378.

Lortie-Lussier, M., & Rinfret, N. (2002). The proportion of women managers:
Where is the critical mass? Journal of Applied Socialisychology, 32, 1974—1991.

Luthar, H. K. (1996). Gender differences in evaluation of performance and
leadership ability: Autocratic vs. democratic managers. Sex Roles, 35, 337-361.

Lyness, K. S. (2002). Finding the key to the executive suite: Challenges for
women and people of color. In R. Silzer (Ed). The 218’ century executive: Innovative
practices for building leadership at the top. San Francisco: Jossey Bass. (p. 229-273).

Lyness, K. S., & Heilman, M. E. (2002). Gender ﬁt and rating; of upper-level line
and staff managers. Paper presented at the 17th Annual Conference of the Society for
Industrial and Organizational Psychology, Toronto, Ontario, Canada.

Lyness, K. S., & Judiesch, M. K. (1999). Are women more likely to be hired or
promoted into management positions? Journal of Vocational Behavior. 54, 158-173.

Maher, K. J. (1997). Gender-related stereotypes of transformational and
transactional leadership. Sex Roles, 37, 209-225.

Martell, R. F ., & Desmet, A. L. (2001). A diagnostic-ratio approach to measuring
beliefs about the leadership abilities of male and female managers. Journal of Applied

Psychology, 86, 1223-1231.

113

Martell, R. F., Lane, D. M., & Emrich, C. E. (1996). Male-female differences: A
computer simulation. American Psychologist, 51, 157-158.

Martell, R. F., Parker, C., Emrich, C. G., & Crawford, M. S. (1998). Sex
stereotyping in the executive suite: "Much ado about something." Journal of Social
Behavior and Personality, 13, 127-138.

 

Martell, R. F ., & Robison-Cox, J. (April, 2002). Gender bias and organizatioﬁnal
mobility: A computer simulation. In R. F. Martell (Chair), Symposium presented at the
Society of Industrial-Organizational Psychology Annual Convention, Toronto, Canada.

 

Massengill, D., & di Marco, N. (1979). Sex-role stereotypes and requisite
management characteristics: A current replication. Sex Roles, 5, 561-570.

 

Maume, D. J., Jr. (1999). Glass ceilings and glass escalators: Occupational
segregation and race and sex differences in managerial promotions. Work and
Occupations, 1999, 483-509.

Maurer, T. J ., Raju, N. S., & Collins, W. C. (1998). Peer and subordinate
performance appraisal measurement equivalence. Journal of Applied Psychology, 83,
693-702.

McCauley, C. D. (2001). Leader training and development. In S] Zaccaro and R.
J. Klimosky (Eds) The nature of organizationgeadership: Understandingperformaknce
imperatives confronting today’s leaders. San Francisco: Jossey-Bass. (p. 347-383).

..Megargee, E. I. (1969). Inﬂuence of sex roles on the manifestation of leadership.
Journal of Applied Psyacholgy, 53, 377-382.

Mobley, W. H. (1982). Supervisor and employee race and sex effects on
performance appraisals: A ﬁeld study of adverse impact and generalizability. Academy of
Management Journal, 25, 598-606.

Morrison, A. M., & Von Glinow, M. A. (1990). Women and minorities in
management. American Psychologist, 45, 200-208.

 

Morrison, A. M., White, R. P., Van Velsor, E., & the Center for Creative
Leadership. (1992). Breakingpthe glass ceiling: Can women reach the top of America’s
Largest corporations? (updated ed.). Reading, MA: Addison-Wesley.

Moskowitz, D. S., Suh, E. J., & Desaulniers, J. (1994). Situational inﬂuences on
gender differences in agency and communion. Journal of Personalityand SOQELI
Psycholagy, 66, 753-761.

114

Mount, M. K., Judge, T. A., Scullen, S. E, Sytsma, M. R., & Hezlett, S. A.
(1998). Trait, rater and level effects in 360-degree performance ratings. Personnel
Psycholggy. 51, 557-576.

 

Newport, F. (2001, February 21). Americans see women as emotional and
affectionate, men as more aggressive. Retrieved from the Gallup Poll News Service Web
site: http://wwwgallup.com/poll/releases/prO10221.asp.

Nieva, V. F ., & Gutek, B. A. (1980). Sex effects on evaluation. Academy of
Management Review, 5, 267-276.

 

Nyquist, L. V., & Spence, J. T. (1986). Effects of dispositional dominance and

sex-role expectations on leadership behaviors. Journal of Personality and Social
Psychology, 50, 87-93.

 

Osborn, R. N., & Vicars, W. M. (1976). Sex stereotypes: An artifact in leader
behavior and subordinate satisfaction analysis? Academy of Management Journal, 19.
439-449.

 

Ott, E. M. (1989). Effects of the male-female ratio at work: Policewomen and
male nurses. Psychology of Women Quarterly, 13, 41-57.

Paolillo, J. G. (1981). Manager's self assessments of managerial roles: The
inﬂuence of hierarchical level. Journal of Management, 7, 43-52.

 

Pavett, C. M., & Lau, A. W. (1983). Managerial work: The inﬂuence of
hierarchical level and ﬁinctional specialty. Academy of Management Jormal. 26, 170-
177.

Pazy, A. (1986). The persistence of pro-male bias despite identical information

regarding causes of success. Organizational Behavior and Human Decision Processes. 38.
366-3 77.

Peiperl, M. A. (2001). Getting 360 feedback right. Harvard Business Review,
January 2001, 142-147.

Pfeffer, J ., & Salancik, G. R. (1975). Determinants of supervisory behavior: A
role set analysis. Human Relations. 28, 139-154.

 

Powell, G. N. (1993). Women and men in managament. Newbury Park, CA:
Sage. . ,

Powell, G. N., & Butterﬁeld, D. A. (1979). The "good manager": Masculine or
androgynous? Academy of Management Journal, 22, 395-403.

115

Powell, G. N., & Butterﬁeld, D. A. (1989). The "good manager": Did androgyny
fare better in the 19805? Group and Organization Studies, 14, 216-233.

Powell, G. N., & Butterﬁeld, D. A. (1994). Investigating the "glass ceiling"
phenomenon: An empirical study of actual promotions to top management. Academy of
Management Journal, 37, 68-86.

 

 

Powell, G. N., Butterﬁeld D. A., & Parent, J. D. (2002). Gender and managerial
stereotypes: Have the times changed? Journal of Management,28, 177-193.

Pulakos, E. D., & Wexley, K. N. (1983). The relationship among perceptual
similarity, sex and performance ratings in manager-subordinate dyads. Academy of
Management Journal, 26, 129-139.

 

Pulakos, E. D., White, L. A., Oppler, S. H, & Borman, W. C. (1989).
Examination of race and sex effects on performance ratings. Journal of Applied
Psychology, 74, 770-780.

 

Quinn, R. E. ( 1988). Beyond rational management: Masteripg the paradoxes and
competing demands of high performance. San Francisco, CA: Jossey-Bass.

Quinones, M. A., Ford, J. K., Sego, D. J., & Smith, E. M. (1995). The effects of
individual and transfer environment characteristics on the opportunity to perform trained
tasks. Training Research Journal, 1, 29-48.

Ragins, B. R. (1991). Gender effects in subordinate evaluations of leaders: Real or
artifact? Journal of Organizational Behavior. 12, 259-268.

Ragins, B. R., & Cotton, J. (1991). Easier said than done: Gender differences in
perceived barriers to gaining a mentor. Academy of Management Journal, 34, 939-951.

 

Ragins, B. R., & Sundstrom, E. (1989). Gender and power in organizations.
Psychological Bulletin, 105, 51-88.

Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models:
Applications and data analysis methods (2nd Ed). Thousand Oaks, CA: Sage.

Reilly, R. R., Smither, J. W., & Vasilopoulos, N. L. (1996). A longitudinal study
of upward feedback. Personnel Psycholagy. 49. 599-612.

Renner, C., Rives, J. M., & Bowlin, W. F. (2002). The signiﬁcance of gender in
explaining senior executive pay variations: An exploratory study. Journal of Managerial
Issues 14, 331-345.

 

Rice, R. W., Instone, D., & Adams, J. (1984). Leader sex, leader success, and
leadership process: Two ﬁeld studies. Journal of Applied Psychology, 69, 12-31.

116

Ridgeway, CL. (1992). Gender, interaction, and inequality. New York: Springer-
Verlag.

Riordan, C. M., & Shore, L. M. (1997). Demographic diversity and employee
attitudes: An empirical examination of relational demography within work units. Journal
of Applied Psychology, 82, 342-358.

 

Robbins, T. L., & DeNisi, A. S. (1993). Moderators of sex bias in the
performance appraisal process: A cognitive analysis. Journal of Management, 19. 113-
126.

Rosen, B., & Jerdee, T. H. (1974). Effects of applicant's sex and difficulty of job
on evaluations of candidates for managerial positions. Journal of Applied Psycholo 59,
51 1-512.

Rudman, L. A. (1998). Self-promotion as a risk factor for women: The costs and
beneﬁts of counterstereotypical impression management. Journal of Personality a_nd
Social Psychology, 74, 629-645.

 

Rudman, L. A., & Glick, P. (2001). Prescriptive gender stereotypes and backlash
toward agentic women. Journal of Social Issues, 57, 743-762.

Sacco, J. M., Scheu, C. R., Ryan, A. M., & Schmitt, N. (in press). An .
investigation of race and sex similarity effects in interviews: A multi-level approach to
relational demography. Journal of Applied Psychology.

Sarbin, T. R., & Allen, V. L. (1968). Role theory. In G. Lindzey & E. Aronson
(Eds), Handbook of social psychology (2nd ed, Vol. 1, pp. 488-567). Reading, MA:
Addison-Wesley.

Scandura, T. A. (1991). Breakingthe glass ceilingin the 1990’s (Technical
Report). Washington, DC: US. Department of Labor, Women’s Bureau.

Schein, V. E. (1973). The relationship between sex role stereotypes and requisite
management characteristics. Jorﬂal of Applied Psychology, 57, 95-100.

 

Schein, V. E. (1975). Relationships between sex role stereotypes and requisite

management characteristics among female managers. Journal of Applied Psycholggx._6_0,
340-344.

Schein, V. E. (2001). A global look at psychological barriers to women's progress
in management. Journal of Social Issues. 57, 675-688.

117

Schein, VE, Mueller, R., & Jacobson, C. (1989). The relationship between sex
role stereotypes and requisite management characteristics among college students. $33
Roles 20, 103-110.

 

Schein, v.13, Mueller, R., Lituchy, r, & Liu, J. (1996). Think manager-think
male: A global phenomenon? Journal of OLganizational Behavior, 17, 33-41.

Senge, P. M. (1990). The ﬁfth discipline: The art and practice ofthe learning
organization. New York: Doubleday.

Shore, L. M., & Thornton, G. C. (1986). Effects of gender on self and supervisory
ratings. Academy of Management Journal, 29, 115-129.

Shore, T. H. (1992). Subtle gender bias in the assessment of managerial potential.
Sex Roles. 27, 499-515.

Simmons, W. W. (2001, January 11). When it comes to choosing a boss,
Americans still prefer men. Retrieved from the Gallup Poll News Service Web site:
http://www.gallup.com/poll/releases/prO101 1 1.asp.

Skrypnek, B. J ., & Snyder, M. (1982). On the self-perpetuating nature of
stereotypes about women and men. Journal of Experimental Social Psychology. 18, 277-
291.

Snijders, T. A. B., & Bosker, R. J. (1999). Multilevel analysis: An introduction to
basic and advanced multilevel modeling. Thousand Oaks, CA: Sage.

Snyder, M. (1981). On the self-perpetuating nature of social stereotypes. In D. L.

Hamilton (Ed), @gnitive processes in stereotypingand intergroup belavior (pp. 183-
212). Hillsdale, NJ: Erlbaum.

Spencer, S. J ., Steele, C. M., & Quinn, D. M. (1999). Stereotype threat and
women's math performance. Journal of Experimental Social Psychology, 35, 4-28.

Stangor, C., Lynch, L., Duan, C., & Glass, B. (1992). Categorization of
individuals on the basis of multiple social features. Journal of Personality @d Social

Psychology, 62, 207-218.

Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual
identity and performance. American Psychologist. 52, 613-629.

Stumpf, S. A., & London, M. (1981). Capturing rater policies in evaluating
candidates for promotion. Academy of Management Journal. 24, 752-766.

118

Stroh, L. K., Brett, J. M., & Reilly, A. H. (1992). All the right stuff: A comparison
of female and male managers' career progression. Journal of Applied Psychology, 77,
251-260.

Thompson, D. E, & Thompson, T. A. (1985). Task-based performance appraisal
for blue-collar jobs: Evaluation of race and sex effects. Journal of Applied Psychology,
10, 747-753.

Tornow, W. W. (1993). Perceptions or reality: Is multi-perspective measurement
a means or an end? Human Resource Managment, 32, 221-230.

Tsui, A. S. (1984). A role set analysis of managerial reputation. Organizational
Behavior and Human Performanceé4, 64-96.

Tsui, A. S., & Gutek, B. A. (1984). A role set analysis of gender differences in
performance, affective relationships, and career success of industrial middle managers.
Academy of Management Journal 27, 619-635.

Tsui, A. S., & Ohlott, P. (1988). Multiple assessment of managerial effectiveness:
Interrater agreement and consensus in effectiveness models. Personnel Psycholo 41,
779-803.

Tsui, A. S., & O’Reilly, C. A. (1989). Beyond simple demographic effects: The
importance of relational demography in superior-subordinate dyads. Academy of
Management Journal, 32, 402-423.

Tsui, A. S., Porter, L. W., & Egan, T. D. (2002). When both similarities and
dissimilarities matter: Extending the concept of relational demography. Human Relations,
ﬁ, 899-929.

US. Bureau of the Census. (2000). Current population reports: Educational
attainment in the United States: March 2000 (Table 1: Educational attainment of the
population 15 years and over, by age, sex, race, and Hispanic origin). Retrieved from
http ://www. census.gov/population/socdemo/education/pZO-536/tab01 .txt.

US. Bureau of Labor Statistics. (2001a). News: The employment situation: July
2001 (Table A-1: Employment status of the civilian population by sex and age).
Retrieved from http://www.bls.gov/news.release/pdf/empsitpdf.

US. Bureau of Labor Statistics. (2001b). Demographic characteristics of the labor
force (Table 1 1: Employed persons by detailed occupation, sex, race, and Hispanic
origin). Retrieved from http://www.bls.gov/cps/cpsaat11.pdf.

US. Bureau of Labor Statistics. (2001c). Demographic characteristics of the labor

force (Table 18: Employed persons by detailed industry, sex, race, and Hispanic origin).
Retrieved from http://www.bls.gov/cps/cpsaat18.pdf.

119

Van Der Leeden, R., & Busing, F. M. T. A. (1994). First iteration versus igls/rigls
estimates in two-level models: A monte carlo study with ML3. Preprint PRM 94-03,
Psychometrics and Research Methodology, Leiden, Netherlands.

van Knippenberg, A., van Twuyver, M., & Pepels, J. (1994). Factors affecting
social categorization processes in memory. British Journal of Social Psychology, 33, 419-
431.

 

Vroom, V., & Yetton, P. (1973). Leadership and decision making. Pittsburgh, PA:
University of Pittsburgh Press.

Wellington, 8, Kropf, M. B., & Gerkovich, P. R. (2003). What’s holding women
back: As barriers shift, lack of line experience has become a chief obstacle. Harvard
Business Review, 81, 18-19.

Williams, C. L. (1992). The glass escalator: Hidden advantages for men in the
"female" professions. Social Problems. 39, 253-267.

Williams, C. L. (1995). Still a man's world: Men who do "women's work."
Berkeley: University of California Press.

 

Woehr, D. J ., & Roch, S. G. (1996). Context effects in performance evaluation:
The impact of ratee sex and performance level on performance ratings and behavioral
recall. Organizational Behavior and Human Decision Processes, 66, 31-41.

 

Wohlers, A. J ., Hall, M. J ., & London, M. (1993). Subordinates rating managers:
Organizational and demographic correlates of self/ subordinate agreement. Journal of
Occupational and Organizational Psychology, 66, 263-275.

Yammarino, F. J ., & Dubinsky, A. J. (1988). Employee responses: Gender or job
related differences? Journal of Vocational Behavior, 32, 366-383.

Yoder, J. D., & Sinnett, L. M. (1985). Is it all in the numbers? A case study of
tokenism. Psychology of Women Quarterly, 9, 413-418.

Yukl, G., & Van Fleet, D. D. (1992). Theory and research on leadership in
organizations. In M. D. Dunnette & L. M. Hough (Eds), Handbook of industrial and
organizational psychology (2nd ed., vol. 3) (pp. 147-198). Palo Alto, CA: Consulting
Psychologists Press.

120

Table 1. Number of Raters for Male and Female Leaders

 

 

 

 

 

 

 

 

 

 

% Raters o
23:33; for # 1:3?” /° 21“" # Raters % Rate rs
Rater Female for Male for Male
Male Female Female
and Male Leaders Leaders
Leaders Leaders Leaders
Leaders
Female
Subordinates 4,259 19 1,799 31 2,368 15
Male Subordinates 5,630 26 768 13 4,722 30
Female Peers 2,686 12 1,218 21 1,421 9
Male Peers 6.591 30 1,239 21 5,189 33
Female Bosses 469 2 228 4 232 1
Male Bosses 2,343 11 512 9 1,768 11
TOTAL RATERS 21, 978 100 5,764 100 15,700 100

 

 

 

 

121

 

 

 

Table 2. Average Number of Raters Per Leader

 

 

 

 

 

 

 

 

Rater Group Female Leaders Male Leaders
Average Number of Total Raters 8.82 8.52
Average Number of Subordinate Raters 3.93 4.02
Average Number of Peer Raters 3.76 3.77
Average Number of Boss Raters 1.27 1.30

 

122

 

Table 3. Leader and Rater Demographics

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Demographics All Male Female All Male Female
Leaders Leaders Leaders Bosses Bosses Bosses
Age # % # % # % # % # % # %
Under 20 0 0 0 0 0 0 1 1 O 0 0
20-29 147 4 100 4 47 5 41 20 1 19 4
30-39 1123 33 780 31 338 37 566 20 454 20 102 22
40-49 1473 43 1075 43 393 43 1221 43 973 42 222 48
50-59 659 19 535 21 123 14 875 31 752 32 112 24
60 or over 46 1 37 1 9 1 126 116 5 9 2
Education # % # % # % # % # % # %
High school 132 4 94 4 38 4 80 61 3 17 4
Some college 387 11 270 11 117 13 212 175 8 33 7
Associate’s 166 5 1 l9 5 47 5 113 83 4 30 7
Bachelor’s 1052 31 769 31 275 30 794 28 654 29 128 28
Some graduate work 409 12 301 12 105 12 328 12 266 12 50 11
Master’ 8 944 28 695 28 247 27 931 33 752 33 164 36
Doctoral 284 8 2 l9 9 65 7 307 1 1 278 12 22 5
Other 50 1 32 1 18 2 30 23 l 7 2
Years # % # % # % # % # % # %
Less than 6 months 93 3 63 3 29 3 55 32 l 22 5
6 months-1 year 118 3 85 3 32 3 61 52 2 8 2
1-2 years 279 8 198 8 78 9 134 96 4 35 8
2-4 years 464 13 298 12 I65 18 280 10 221 10 50 11
4-6 years 373 11 257 10 114 12 309 11 264 11 40 9
6-10 years 526 15 372 15 154 17 366 13 277 12 80 18
10-15 years 596 17 455 18 138 15 531 19 437 19 82 18
More than 15 years 994 29 785 31 206 22 1065 38 917 40 139 30
Salary # % # % # % # % # % # %
$25,000 or less 12 0 7 0 5 l 22 12 1 9 3
$25,001-35,000 38 l 20 1 18 2 23 8 l 15 5
$35,001-50,000 158 6 85 4 73 10 59 40 3 19 6
$50,001 -75,000 763 29 51 1 27 251 34 131 87 6 40 12
$75,001 - 100,000 634 24 439 23 195 26 305 16 216 14 86 26
$100.001-150,000 598 22 471 24 122 17 ' 475 25 386 25 81 24
$150,001-200,000 213 8 181 9 32 4 395 21 . 351 23 38 11
$200,001 or over 252 9 211 11 40 5 507 26 456 29 43 13
Ethnicity # % # % # % # i % # % # %
Asian 117 4 81 4 36 5 51 41 2 8 2
African American 64 2 38 2 25 3 33 23 l 10 3
Hispanic 74 3 54 3 20 3 78 68 4 8 2
White/Caucasian 2281 86 1640 87 635 86 1948 89 1584 88 333 90
Other 102 4 77 4 25 3 90 81 5 9 2

 

 

123

 

Table 3 (continued). Leader and Rater Demographics

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Demographic All Male Female All Male Female
Peers Peers Peers Subord Subord Subord
_Age # % # % # % # % # % # %
Under 20 6 0 1 0 3 0 11 0 7 0 2 0
20-29 424 5 239 4 179 7 121 1 12 525 10 655 16
30-39 2878 31 1932 30 858 33 3515 35 1934 35 1455 35
40-49 3890 42 2646 41 1141 44 3347 34 1910 35 1333 32
50-59 1937 21 1500 23 394 15 1679 17 1010 18 631 15
60 or over 178 2 143 2 34 l 170 2 109 2 57 1
Education # % # % # % # % # % # %
High school 403 4 210 3 179 7 745 8 345 6 374 9
Some college 1013 11 625 10 361 14 1756 18 804 15 911 22
Associate’s 507 6 361 6 134 5 977 10 465 9 479 12
Bachelor’s 2948 32 2086 33 775 30 3199 32 1873 34 1213 30
Some graduate work 930 10 648 10 247 10 857 9 474 9 338 8
Master’s 2478 27 1743 27 665 26 1725 17 1092 20 580 14
Doctoral 847 9 647 10 183 7 424 4 313 6 101 2
Other 115 1 71 1 42 2 183 2 81 l 91 2
Years # % # % # % # % # % # %
Less than 6 months 116 1 72 1 36 1 333 3 185 3 142 3
6 months-1 year 270 3 187 3 74 3 565 6 283 5 265 7
1-2 years 679 7 441 7 217 8 1005 10 506 9 479 12
2-4 years 1343 15 844 13 465 18 1823 19 908 17 841 21
4-6 years 1073 12 693 11 348 14 1217 12 683 13 501 12
6-10 years 1445 16 978 15 432 17 1491 15 819 15 631 15
10-15 years 1667 18 1213 19 412 16 1334 14 773 14 523 13
More than 15 years 2566 28 1910 30 587 23 1987 20 1244 23 693 17
Salary # % # % # % # % # % # %
$25,000 or less 1 11 2 43 1 63 3 418 6 109 3 288 9
' $25,001-35,000 218 3 90 2 112 6 766 10 197 5 549 17
$35,001-50,000 487 7 248 5 234 12 1394 19 489 12 875 27
$50,001-75,000 1478 22 870 19 579 30 2313 31 1386 34 884 28
$75,001-100,000 1531 23 1071 24 431 22 1192 16 836 20 330 10
$100,001-150,000 1592 24 1227 27 339 18 916 12 719 17 181 6
$150,001-200,000 561 9 457 10 98 5 255 3 215 5 33 1
$200,001 or over 597 9 509 11 74 4 231 3 167 4 52 2
Ethnicity # % # % # % # % # % # %
Asian 264 4 201 4 56 3 344 4 218 5 l 17 4
African American 167 2 91 2 73 4 245 3 90 2 154 5
Hispanic 197 3 153 3 42 2 249 3 I43 3 104 3
White/Caucasian 6253 88 4340 88 1749 88 6478 85 3576 85 2709 84
Other 236 3 149 3 77 4 332 4 164 4 153 5

 

 

124

 

Table 4. Denison Leadership Development Survey Skills, Dimensions, and
Subdimensions (Denison & Neale, 1996)

 

 

 

 

 

Stable

Type of Skill Dimension Subdimension
External Focus Adaptability Promotes Organizational Learning
Flexible ,

Emphasrzes Customer Focus
Creates Change
External Focus Mission Deﬁnes Strategic Direction and Intent

 

Deﬁnes Goals and Objectives

 

Creates Shared Vision

 

Internal Focus
Flexible

Involvement

Empowers People

 

Builds Team Orientation

 

Develops Organizational Capability

 

Internal Focus
Stable

 

 

Consistency

Manages Coordination and Integration

 

Works to Reach Agreement

 

 

Deﬁnes Core Values

 

125

 

Table 5. Denison Leadership Development Survey Items (Denison & Neale, 1996)

 

Dimension

Sub-Dimension

Items

 

Involvement

(Internal Focus
Flexible)

Empowers People

5"”?

@899.“

Secs that decisions are made at the lowest possible level.

Shares information so that everyone gets the information s/he needs.
Creates an environment where everyone feels that his/her effort can
make a difference.

Involves everyone in shaping the plans and decisions that affect them.
Ensures that the necessary resources are available to do the job.
Conveys conﬁdence in people’s competence to do their job.
Encourages others to take responsibility.

Delegates authority so that others can do their work more effectively.

 

Builds Team
Orientation

.0

Builds effective teams that get the job done.

. Encourages effective teamwork by others.
11.
12.
13.
14.
15.
16.

Knows how to use a team approach to solve problems.
Knows when to use a team approach to solve problems.
Fosters teamwork within the work unit.

Knows how to design work so that it can be done by a team.
Values the contributions of the people s/he works with.
Acknowledges and celebrates team accomplishments.

 

Develops
Organizational
Capability

17.

18.
19.
20.

21.
22.

23.
24.

Builds the capabilities of employees into an important source for
competitive advantage.

Knows how to utilize the diversity of the workforce.

Coaches others in the development of their skills.

Is sensitive and responsive to diversity issues when dealing with
others.

Helps direct reports create realistic development plans and create
opportunities for them.

Uses rewards and recognition to motivate good performance.
Develops his/her own people so that they are ready for promotion.
Builds employee skills so that the organization always has good
“bench strength”.

 

Consistency

(Internal Focus
Stable)

 

Deﬁnes Core
Values

25.
26.
27.
28.
29.

30.
31.
32.

Does the “right thing” even when it is not popular.

“Practices” what s/he “preaches”.

Has an ethical code that guides his/her behavior.

Helps deﬁne the organization’s culture, values, and ethical standards.
Helps employees learn to apply the organization’s values when
dealing with others.

Lives up to promises and commitments.

Has earned the conﬁdence and trust of others.

Clearlyarticulates a set of fundamental beliefs that are not negotiable.

 

Works to Reach
Agreement

33.
34.

35.

36.
37.
38.
39.
40.

Helps people to reaCh consensus, even on difﬁcult issues.

Works to find alternatives that will beneﬁt all when confronted with a
disagreement.

Helps people in his/her organization be effective at reaching
agreement on key issues.

Incorporates diverse points of view when making decisions.

Promotes constructive discussion among people with conﬂicting ideas.
Is willing to compromise when necessary in order to reach agreement.
Works toward win/win solutions when disagreements occur.
Reconciles differences bLseeking to clarify and understand others.

 

 

Manages
Coordination and
Integration

 

41.
42.
43.

45.
46.

47.
48.

Works hard to foster the alignment of goals across all functional areas.
Builds coordination across departmental boundaries.

Uses informal networks to get things done.

Builds relationships with key people in other functions and levels.
Helps create an environment that facilitates coordination of projects
across functional units.

Makes certain that things do not “fall between the cracks”.

Builds support for ideas through contracts with other departments.
Establishes mechanisms that facilitate effective cross-functional
communication.

 

126

 

Table 5 (continued). Denison Leadership Development Survey Items (Denison & Neale,

1996)

 

Dimension

Sub-Dimension

Items

 

Adaptability

(External
Focus
Flexible)

Creates Change

49. Continuously looks for new and better ways to do work.

50. Encourages creative thinking.

51. Challenges the way that things have always been done and looks for a better
way.

52. Champions change that goes beyond the scope of his/her job.

53. Challenges organizational practices that are nonproductive.

54. Foresees problems before they arise.

55. Serves as a model that creates change in other parts of the organization.

56. Generates innovative ideas and solutions to problems.

 

Emphasizes
Customer Focus

57. Encourages direct contact with customers.

58. Responds quickly and effectively to customer feedback.

59. Ensures that employees have a deep understanding of customer wants and
needs.

60. Uses customer comments and recommendations to change organizational
practices.

61. Actively seeks feedback from customers.

62. Continuously tries to improve service to customers.

63. Incorporates customer input into the planning process.

64. Recognizes the need to respond quickly to customer concerns.

 

Promotes
Organizational
Learning

65. Deals constructively with failures and mistakes.

66. Views failures as an opportunity for learning and improvement.

67. Creates a working environment in which learning is an important objective.
68. Openly accepts criticism without being defensive.

69. Works well under conditions of ambiguity and uncertainty.

70. Knows the strengths and weaknesses of the competition.

71. Encourages others to learn about the best practices in the industry.

72. Helps others to understand “the big picture”.

 

Mission

(External
Focus
Stable)

 

Deﬁnes
Strategic
Direction and
Intent

73. Provides employees with a clear mission that gives meaning and direction to
their work.

74. Implements strategies by developing clear goals, objectives, and tactics.

75. Focuses on long-term strategies, rather than quick ﬁx “band-aid” solutions.

76. Effectively allocates resources in line with strategic priorities.

77. Helps deﬁne strategies and tactics that keep his/her organization competitive.

78. Has a clear strategy for the future of his/her own part of the organization.

79. Is able to meet short-terrn demands without losing sight of the long-term
strategy.

80. Communicates a clear and compelling rationale for the business strategy.

 

Defines Goals
and Objectives

81. Sets clear goals that are ambitious, but realistic.

82. Holds individuals and teams accountable for achieving goals and objectives.

83. Provides clear directions and priorities for employees.

84. Establishes high standards of performance.

85. Involves employees in the goal-setting process so goals and objectives are
understood and shared.

86. Tracks progress against stated goals.

87. Effectively communicates the goals and objectives of the organization.

88. Aligns goals and objectives with the strategy and vision.

 

 

Creates Shared
Vision

 

89. Helps create a shared vision of what this organization will be like in the
future.

90. Communicates the organizational vision to his/her employees.

91. Uses the vision to create excitement and motivation for employees.

92. Realizes short-term goals without compromising long-term vision.

93. Organizes work so that everyone sees the connection between the vision and
daily activities.

94. Translates the vision into reality in a way that helps guide individual action.

95. Inspires others with his/her vision of the future.

96. Engages others in ways that ensure buy-in and commitment.

 

127

 

 

 

 

 

 

 

-- mm 9. N5 .5 35 5.595055 55050.5: 500m 5555 0.55355 50555~E5m50 05 555 0m 5555050. 00.355550 5555m N
U - .Ne . ~05. .- no.5. .Om..m . ...50550E05m 50.5 55.05 055. .005. 55.5 3 Bacon 5.30 5055.55 w5050>0Q mN.
: . .mm . .mm . . 2.5.5 No.5 . - . 0055555055055 500» 055>550E 05 5.055.503.0055 555.5300 .505 N.N
1 . . cm. . . N. 00.0 mm.” E05 50.5. 50555550550 05505.0. 555 55555 5505505035 055555505 0.55.050 5050505 500555 55505.5 5N .
U . 5.0. .5 echo 5m .5. 55050 .553. $55505 5053 505055 55555035 0.5 03050550... 555 03555505 55 ON
.0. . co m. Rd NE 5.555.050. .5505 .50 550E5050>05 05 E 550.50 5.050500 05.
-- 5555 .. 5.5. mad ova 00.50.555.03. 05 50 3505035.. 05 0555.55 .05. 30.55 .3055 M55
-- -9 R - WN5 SN .0w5555>55 0>555505E00 50.5.8558 55550555 55 .0555 $050—$500 .50. 50555555553 05. 5555mm 5.1.5
0 mm 5 who 05.5. 5550555055an005 552 005550500 55 5050530553.. 55
Q , om . .o - 5.56 m: .553. 95.503 0.55. 055.005 05. ..50 0550555555500 05 50.555>. ,m5 .
0. .No . .05. 5.55.5 50m. .5505 .5 .3 0.505 .05. 550 55 .555 00. 050.3 55,505 0.5 3055, 030.5% .3.
U 8.5 . .o : owd ”W5. . .5E5 05.5.0.3 0.5 55553 50355505. 5.505505. 2..
U - - 5.5.5. ..m. - {.0.-o mma - ..mE03055 0.205 .5 55.055.05.55 5.50.5. mm? 9.5.5.0553. .5355! .N.5. ..
O . NW m . . «who 5.55m «55.05.15.055.0.>5ow.o5.50w05.5%5 E59 5.05.5 05 305.5. 53055. .5 5-
...U. . 5.0.5.. , . o. ..- .. «To 05.5.. . .. . 5.5.050 ..3. 5.0355505. 0350050. mom555005m .55 ..
--. 5...». . . 0N. . 3.5 555 . . ..0505. 9.0.5. 05. 5.0m 5555. E55505 05500.50 055555 a .
1 555. mm N52 3 .m 550550.050 050E 05503 550.5 05 550 .5050 555 05.. 3550555 5055M050Q m
U . N5. mm .. NW5. SM. .. . - . :355550505005 00.555 05.505.55.50. muwgowEm :5. .
U 00 0 who 1.5...503505 05. 05 0050505800 0 055005 55 00.55.055.558 5.6.5.500 5 .
... «N S. . 5.5.0 No.m .505 05 05 8 03555.55 05 500550005 b53000: 05 5555 0.055555 . m
U .3 m .. N55 5 5.5. .5505 505.555 555 05055005 55.5 55555 05 M55555 .5 .050b0>0 5020355 .5.
U . .oo5 . o 05.0 5.5. 00505095555 5 005555550. 5.50.50 5.0555: 55.5 .0.—005 050b0>0,0.5.053 50.55.50.550 55 35.50.50 m.
U :85 . o 3.0 50.5. 55005 05.5 50555E50E5 05. 550m 050.3050 55 0550555550555 355m N
O 5.55 m 3.5 E .5. 50.505 0355505 550305 05 55 055E 0555 0.5052005 555 000W 5
555.5550 .50 ..0
_ 25555.0 swamw. 5mg”; 50 E .555 55.55 a

 

 

 

 

 

 

 

 

$55555 5055 55055505059 5555EEoU 555 05550m< .5 035.5.

128

 

 

 

 

 

 

 

 

U 00 3 55.0 $6 .50.5.55.0555EEoo 55505055500050 035500.50 0555555055 555 mEm555500E 0055555555 555.
. ..- . . . . WN. ,- .. Iwn. .-- 05.0. 55...N . .-. .. -. .. “0550555505 .5050 553 55.05.5500. 5550555 5.5055 555.505050-05555m. . 5.5.
-- - . 5.5.. .. .555. . - 055.0 .00..N .. - - .. .. .. .- .. ...- . .. . .....5050550: 05 50.033.50.35 555.5... 50.5- .050 555.5 5.55 55.5500 00.05.52. 0.5..
-- . Nw . .mN. . 550.0 05 ..m .5555 555.05.505.35. 050500 05005.85 .50 50555555000. 0055555055 555 550E5055350 .55 0550.50 05505.5 we
-- 0.,N... 55.5... 0N5 5.0..N . 05030.5 55.55 05.05.0555 5.0.50 ..55 0550.05 3.005. 5.3 055505055505 555555 3. .
-- . .5555. . . .mm. . 5N5. Noam . . . . . ..0505 0.5555. 5.0m 05 .505503505 555550.555. 50.5. 9.
U. . .No. . . .. .55.. .. 5.0.0. .0.m..m ...005555105. 5555055555505 500.505 50555555000 0.5555m. N5.
. .2 . 30.5 - - 5.5.3 8.5.. . . .58.... 55.90255 55 .522. 25.05 50.535.550.55. 05.5.55 2.50555 .0553 .5.
U 005 0 050.0 0N... 305.3 5:0. 0.5—5.5.05 0 505.0 5555050555 555 @5550 05 M52000 .35 305050.555 5055050005 05.
to . . - .00.. 0.5. . 0.0.5 05 .5. . . 550.00 5505500555550 5053 55.05.555.00. 5,535.53. 5.5305. 005.03 0m.
. 0. {mm .. . .m. . N00 05 5.. .EOE005m5 505.05 0.5 5055.0 55 355000.005 5053 05550555500 05 $555533. 5 .55m .
:0 00. m. 05.50 05 .5. ...m.5055 555055500. 553 05500.5 m50E5 5055050055 03505555500. 0050E05n5. R...
.U . m0- ..0. 5.5.0. 5.N.5. . . 550555005 w55055.E 505.3 3.05.3 50 0.55505 00.50.35 505550555055. 05.
U 55 .m N550 N055 . 5050.55 .3005 50 550500555 m5550505 55. 05005.50 00. 5055555550 505.055 55 055005 55505.5 mm.
U . 00.5.. .0. 50,0. 05.5. . ..550E005w5055 5 .553 505505.500. 5053 555 55.5.0505 :53 .055 0.035.555.0555 555.5 .05. 005.503.. 5.5 .
.0 00 m. 8.0 5.0.5. 505505 55505.55. 50 50.30 050500.500. 5.0505 0.5 05500.5 55505.5 .3.
< 0 mm mm .0 m: 035505.05 505 055 5555 0.505505 55550555555 50 500 5 50555505555 3555050 Nm
. ..- . 0m 05 . 00.5 mm.m . 5.50.5.0 .50 50.5.5 555 005059500 05 505550.553..5m1
-- . m0. . 05 3.0 M53“ . .55505555E00 555 000550.55. 05 .55 003.55 0m
U R . 0- 550.0 05.5 3.5050 5.5.3. 555505 5053 M0555> 0 505555555550 05 355.55. 05 5.5.505 .0003055E0 05505.5. . 5N.
-- 0N 5.5 - 25.0 5.5 .m. 5055555555 55.05.50 .555 0055.3 0555.550 5 55.055.55.55»? 05 0550.5 5550mm 0N
U . 0m. . m. . . 3.0 N05 . . - . 50.53.5505 555.555 005555. 5.55 0.500 55050 .55 055.5 N.N .
:1 .. 5N. 0N. 5.0.0 00.5 . . . , . . . . : . . . .- . . 500505055... 0.50 .5553 5005505535.... .0N.
-- 0N Nm 5..N.5 mmN 5555505 505 55 55 5053. 5030 2555.5 5555 05 500.9 5N.
555EE0 .50 .50
5 50505050 hamw. 50%;”; 50 2 555 55.55 0

 

 

 

 

 

 

$5555.35 E055 550E5050>0Q 5555EE00 555 05550m< 55055555005 0 0555.5.

129

 

 

 

 

 

 

 

 

 

 

 

 

-. 5555 2 8.5 3 .m . 055550555 55 05.. 55505055 05 0.5050 5055 N5.
. .1 ...-- . . . Wm: . zmv, . 5.0.5 5.»..N . . 50.55555 .05 E 0.0055055 5.0.05 .05 55°55 55.0.5 9 05.050 .w0mm5sou55m 5.»
- 1 -< x - .o: . -05. cm a .35 . .. .. .p . ...550555505558 05 we 000.00.555.03 55 0.550550 05 0.3055 on
.-.. . . o5 . :25. 5.5.55 Nm.~ ...555555005 5.55 5055555 5 0.5595550 5055.555. :03 0.55503 3
-- ... msm. . .3. . 5.5.5 gum . . 50.305.50.505 5.5055 5.550553 5559055550 0.555005 b55050 we
. . L .3- . . .3. mad. Sam. . 0350030 5555.05.55 5. W5 w505 550.5553 .5 55505555505550 m.51o.3.m 550555050. S
.... . .555..- .. 555. .355 mad 5550555058555 55 5552 .85 5555555055 5 5 5.055555 0.305 ow
. - -.. - . ..5. E ..5 L . 5.5.0 5.9m. . -.. _. . .. .. . .- - . - - , ”05550.5. 55. 00.55.5555 553 55.555555055550558 2550a no
-- 3 o5 oho ovm 5.500558 505550550 05 5505055555 555255505 05 50055 05 0055805 55

. ..U - 5m. : m. .090 :55“ . . i -. .-.. .. ...wmwoﬁmwgwaumugﬁ 5.5555555. 50555055050555.5555.-. mo.
, . .... ...Nw. . . .555. 5.0 3. m. . . 05.05.555.50 8 00500. 056.555.5555 05 505.55. .55—5550555555580 S.
...U . .. . 55.. .o. 05.0 ohm 50.55.505.555 5550.55 055550555 $500.0 b02504. G

. . - .--. . . Kw. . .3. . Sac ovm . . 3.2505555 55555555555590 05.55.550.05 505.555.55.505805 55 5555058 5059.55.50 0.00.9. co .
o 2. . 0 mod ww.m 5.00.55 55. 5.5.3. 505050550 550 M55555 500.5 m 0355.5 0005505550 5555 00555555. an.

.-..-1 .55.»: m. . mod 555% . 55055555505 505.555.5550 .05. 55.03.50.050 5555 555.505.5555 555055.005. mm .
-- n... . 3. Rd 9mm . 0505585055 553 505.558 5.0055 505555855 5%
.- m wm Etc 35.. .mE05o555 05 055255558 55 5055 0553055555 3555505500 om
. 5w. - o . .- .R . .. nod 35 - - - - 55555550 05 555.555.505.55. 5 555.055.55.050. :5 505.955 55 5 m0t0w mm
, . . -.. o .53 . $5 85. .. 0.0.5 50.55 050.5055 5503855. 00805055 5%.
. .. 5.5.. .. _ .o .. mm. 5.00 5.5.5 .. ., . -. . . .503555555555555055 .05 .55 5500555555 5555555550 0050:5555 R
<. ...w - . 5m . mod on.~ .....555551 5.0555. 5 “05580. 05 55.59505 0003555550. 55855555550 mm

. . .555: . m .. . ow. owd mo.m .53 50.55055 55. 55 5.0.0.— 5.5. .0555 5.005 0535.55 .035 5.055555 5555 53. 05 00505555550 3 .
. ..- w .. 555 wed an .. 5555555555. 0.3555050 005555855 ow
. < . m .3 55:5 mmd ..5—503.55 0.5. .053 5050.55 55. 3055 58 00585 5555055555500 95

EEEEo .50 .50 .

_ 505555550 55%;”; swam”; 55m E 5555 055555 55

 

 

55555535 E055 5550Eno50>0Q 5mc=EEoU 55 055550m< 505555555585 5 055.5.

130

 

  

 

 

 

 

 

 

 

-- hm H.H mod 3. m HHHoHHHHHHHHHHHoo HHHHHH 5:99 235 $5 9&3 HHH $38 8 HH HHmH om

.... . H.H . . . mm. ,. . .NHHH . nod. . . . .oHHHHHHH .uHHH Ho :onH.> HoHHHmHHH HHHHB mHuHHHo. wuHHnHmHHH mo

..HV . . w . - , - Om. . In; , m: .. ..HHoHHuHH. HHHHHEHHHHHHHH uHHHHHw .mHHHuHH HHHHHH .33 HH HHH bHHHHuH oHHHH .HHonH> oHHH monHmHHHHHH. H3.

... . ,,,,, mHH .. - 0H, 3.0 amnm .mmHHH>H.HoH..H HHHHHHH H.HHHHH. .HHonH> 05 50.33 5:858 05 38 0:355. HHHHHH 8 VHS! woNHHHmmHO ma

..- . 32.. we . .owd . WON. . .. . ; .H.HoH.mH> .HHHHoH.mHHoH wHHHmHHHHoHHHHHHoo. HHHoHHHHB mHHHow. HHHHoH-,HHoHHm HUNHHHHEH . no

- . H , , 1-3..-. - - mu. rNHHH NH .m . .muoHoHHHHHHu. H8 HHoHHHH>HHo.HHH. HHHHHH HHHoHHHoHHoxu 835 0H :33 05 moms. Ho
.- mm ....mw... .. .HoH Had. ,. - . . , - . , , i ...muoHHoHHHHHHu HoHHHWHHH. oH HH.oH.wH> HHHHHQHHSHHHHHmHo H.H.HH muHmuHHHHHEHHHoU .oo.

.- NW. , mm ooh H 2 .m .o.HHHHHHH 05 5 .93. on HHHB HHoHHHHNHHHamHo mHHHH HHHHHB .Ho HHonH> HHuHHHHHm HH uHHHuHo WHHHoHH am.

< o no and E .N .HHonH> HHHHIHHNMIHHHHHH oHHH HHHHB 828030 HHHHHH mam mHHmHH< mm

a..- H.H... . We: .93 mod . .HHoHHHHNHHHHHwHo .uHH.H Ho mu>HHoquo HHHHHH mHHHom o5 moHHHoHHHHHHHHHHHoo .HHo>HH.oubmH. . 5
<. . , m: 1 No . 85 and, , .mHHHow 3:3 HmHHHHHwHH $.onon HHHHHHHHH. .on .

o 3 . .m. 8.0 w: . _ 8an . 8.

. .98 382095 oHHH 338.030 HHHHHH H.H.uom 8 3.0.85 wHHHHme- Hwow v.6 E $3295 829::

< . o ..oo . no.0 NHHH .uoHHHHEHOHHoHH Ho mHHHHHHHHHHHHm HHwHHH onHmHHanmmH 3H

.4. o oa $6 3: . . .muumoHHHHHHw Ho.H. moHHHHoHHHH HHHHHH mHHoHHooHHHH HSHo 33>on mw
_ ..<. . o- . om. No.0 . .2... H. .Hu>.H.HouH.HHo. HHHHHH mHaom wHHH>oHHHuw How 03858on 952 HHHHHH wHwnHHEHcaH H.H.HHoHH. .mw-

< o co mod 5“ H. . .oHHmHHSH .HHE. mHHoHHHHHHHHHH. oHHH HHHHHH ”Hmow 320 Sam Hm

< H: on 3.0 waH .mmuHmHHm mmuHHHmHHHH 06 H8 oHaHHoHHmH mHHHHHRHHHHoU Ea HmoHo a .muHHHoHHHHHHHHHHHoU om

< . . .o . . . Km. . . .Nwd -HN..~.. kuHHHHHm .HHHHuH..mHHoH 05 ..Ho HHHme wHHHmoH HHHoHHHHB muHHHHHHHuHH .HEuHHHoHHm .62: 9 03m .2 . on

. < .o . on H.H...o SH .., HHoHHHHNHHHHHwHo 2.: Ho :3 55.0 HoHHHHHHH. H0 2.33 05. H.o.H .xmonHHm HmuHo HH mHHHH wm

._ 4. am :3: no.0 «H.H..H, ..o>.HH.H.Hu.HHHH.H8 HHQHHHHNHHHHHwHo HuHHHmHHH muuHH HHHHHH .muHHoS HHHHHH. monoHHHHHm oHHHHuHH .mnHHoHH. . 2..

< . ..o .Hw H50 wwH . . .muHHHHoHHHH .onuHHHHHm HHHH3 uHHHH, HHH 3.8282 muHHHooHHHH mHuwHHuubm . on

<. .2. 4 5w Nwd .omd. . . . é ...mHHoHH.H.HH0m H.HHHHHHHHHHHHHW. xH.H HHoHHHHH 5.2.: 5.58 monoHHHHHm HHHHoH-m.H.HnHH HHo. womzoonH .. WE

c . o om 8.0 . . S. H. .338 HHHHHH 3360.30 mHHHom 8.20 wHHHnHoHuHHoHH .3 wuHmBHHHHw H.HHHoHHHoHHHHHHH ..Hﬁ.

< . a.” - .3 NH. ..H .wHHN .. .. HHS») HHoHHH...oH HHoHHouHHHH Ea. mHHHHHHHoE 323$: HHonmHHHH :83 m HHHHB muwonHHHHHo wuVSonnH m.»

HaHHHHEEoU m .8 H. N .8 H
3.53. BEE. 8.2.x. am Hz .5: 3.3 H

 

 

 

 

 

 

 

meHHHwJH HHHoHH HHHoEHHoHo>oQ HHHHHHHEEoO Ea oHHHHom< .AHHoHHHHHHHHooV o 2an

131

Table 7. Number ofLeaders in Each Organizational Level

 

# Female

% Female

 

 

 

 

 

 

 

 

 

 

 

Organizational # Female % Female # Male % Male
and Male and Male
Level Leaders Leaders Leaders Leaders
Leaders Leaders
Line Management 695 22 224 27 470 20
Middle Management 1286 40 344 42 938 40
Executive 1203 38 263 32 934 40
TOTAL RATERS 3184 100 821 100 2342 100

 

L»)
[Q

 

Table 8. Number of Leaders in Each Functional Area

 

 

 

 

 

 

 

 

 

 

 

 

 

0
. # Female A Female # Female % Female # Male °/o Male
Functional Area and Male and Male
Leaders Leaders Leaders Leaders
Leaders Leaders
Finance and
Accounting 298 9 80 9 2 16 9
Engineering 222 7 22 3 199 8
Manufacturing and
Production 266 8 28 3 238 10
Research and
Development 177 5 49 6 128 5
Sales and
Marketing 689 21 170 20 513 22
Purchasing 141 4 32 4 109 5
Human Resources 167 5 94 ll 73 3
Administration 581 18 168 20 411 17
Support Staﬁr 149 5 32 4 117 5
Professional Staff 539 17 181 21 355 15
TOTAL 3229 100 856 100 2359 100

 

 

 

 

 

 

 

H
b.)
b.)

 

Table 9. Number of Leaders in Functional Area Percent-Female Ranges

 

% Females in

# Female

% Female

 

 

 

 

 

0 0
Each Functional and Male and Male # Female A Female # Male A Male
0 Leaders Leaders Leaders Leaders
Area Leaders Leaders

Above 60% 316 10 126 15 190 8
Between 40%-60% 2248 70 631 74 1604 68
Below 40% 665 21 99 12 565 24
TOTAL 3229 100 856 100 2359 100

 

 

 

 

 

 

 

“Complete range is lO.4%—78.7%

34

 

Table 10. Number of Leaders in Each Major Industry

 

 

 

 

 

 

 

 

 

 

# Female % Female # Female % Female # Male % Male
Industry and Male and Male
Leaders Leaders Leaders Leaders
Leaders Leaders
Construction 306 10 46 6 240 l 1
Manufacturing 798 27 121 17 640 30
Transportation,
Communications,
and Other Public
Utilities 180 6 63 9 l 15 5
Wholesale and
Retail Trade 409 14 80 1 1 31 l 15
Finance, Insurance,
and Real Estate 336 11 125 17 206 10
Services 614 20 212 29 359 17
Public
Administration 353 12 74 10 273 13
TOTAL 2996 100 72 l 100 2144 100

 

 

 

 

 

 

 

135

 

Table l 1. Number of Leaders in Industry Percent—Female Ranges

 

 

 

 

 

 

0
% Females in $26313: 1:15:23]: # Female % Female # Male % Male
Each Industry“ Leaders Leaders Leaders Leaders
Leaders Leaders

Above 60% 751 25 325 45 408 19
Between 40%-60% 763 25 191 26 510 24
Below 40% 1482 49 205 28 1226 57
TOTAL 2996 100 721 100 2144 100

 

 

 

 

 

 

 

“Complete range is 9.7%—92.0%

 

Table 12. Number of Leaders in Proportion of Male Raters Ranges

 

# Female

"/0 Female

 

 

 

 

 

% of Male and Male and Male # Female % Female # Male % Male
Raters” Leaders Leaders Leaders Leaders
Leaders Leaders
Above 60% 2067 54 269 27 1727 64
Between 40%-60% 861 23 246 25 586 22
Below 40% 882 23 473 48 389 14
TOTAL 3810 100 988 100 2702 100

 

 

 

 

 

 

 

“Complete range is 0.0%-l()().0“o

137

 

dint? _o>o_ 523— a ﬂ 2822: 98 832 .on 958 28 Eco a 265 £ng :8 Sumo. .«o 88 05 E 35.8
..23 mm :68 Boat? 357332 3505 so «Soto m: mam—owe: 5:3 033.5» :95. a up on? .36 z 5533* .mmcuﬂ Bug 3 :08 332...? E>24§E>€E
.332 :o ﬂoobo m: mam—258 $9298 5 E 8632: 5:3 03%? 295— a on Q8 “venom ~023— .£a:8xo com .338 7828862 :23 no.“ 3 £335

Lo :32 $032 05 :23 so mﬁncoaoc b9» >9: 03%? :80 mo 3%: 05 35 8oz £85092 come He Etoﬂoqsm E @2865 a 03%? some .3 _o>o_ 23. a

 

 

 

 

 

 

 

 

 

 

 

«$2.25 ummzuéﬁmﬁ.» EB Euﬁmﬁgwn 26.38 12656

2 3.8%? 5.3. ashamN $656523 $63 $953 V2338

Eu nootmxukxﬁ o: .893 35.8§\W&u£§o§éa£&

3mmch 5 .2th3 £65 :65 3956me .a.& 9556me .9:

mmocozuuobo ”MES; 3,6523%» x633 EB «Emﬁmsgmh 3:32:58

.9850 use carom: v.62 M233: .3» anus Emﬁmﬁmﬁh

$298 $33 memo? E: 2th£ £35K .898

. _ mmauﬂ Bzocotﬁgﬁg 5 “$53; $35?an 33

votoannm coauEBE .33.-~ E2: 2036 took anon thEoQ. 9850 ES xmhzum $363 :6?ch

82 332 m ”21:.“ .9850 333 REESE ”ﬂotsam 3934 35:9:33 65 63.569: 3.: 32463 ”NE

“wanna anunuéﬁwkwm EB EmEmoNoﬁw 2336

..Emﬁo S FEM? 5.; 2th3 63561336 365 2333

38838 98 36:29153 o: .mmeﬁE hluElEEloh

Smmﬁwﬂ g5 .9th3 £3: :65 \bwaaotmxu 3.3 M£Exo\xm&

mmoco>nuoto .u ..t $2.5: ammzoxﬁomﬁm L638 38 azmﬁmﬁmah

$2550 223858 38 336%» 6.62 35:33: .mé 9.355

EmEQSogh 2598 $32 menu? 53 uxmhuﬂ mBEwsx

_ 3.58 663355 $3.33 5 6355 «meanwhxocﬁ.

cutansm 530885 unafm EoEmoEE—u has 2me .9.on hmEnEoQ. .9856 ES amnnmm 16368

82 ﬂak: m “24$ .muoﬁO loam—65 ”gnoccom Sumo: 595g Qwﬁzozsux ms Saxohoﬁ 33 gm .. NE
338% 83.23. .“ohunwcmw 2398.82 .mwohmumhuwy 3353mm

 

 

£36m new woman—~54 .momoﬁogm 32m MO 3885 .2 2an

138

 

 

88855 885565

-855 5 86588.8 885 85 5 88838 5 $58.5
855 58 8555. .8885 88888.5 2R 5.858538
55 5.85 S 88558 .3 «855.5853. 58 ..5qu

 

 

 

 

 

 

 

 

 

 

 

 

 

Sum—58 .8265 55.5565 .9856 595 888 5.85. 55558

88..an canon—85 >83-~ 8:38:80 5 5858885 8.85 @582 8.: «@555 5858885
82 £96— m ”242 .3050 3.3385 388:8 .8984 553558 8.35 8988.5 5.: 98535 8558k ”amt

. . .859: 8\

85 85 8\u88>58§m 88me M55855 $55.5 5

5858.5 $38.5. 885 85 2.2: 5858885 8.0885 8.3.5

8 353.55 358; .58.: .856 5 88.585 8558\8\

85 98585 55 85888.5 A8>anu 538.5. 8.85 ..5

5.: $58.58 8.8“. 9.: 85 55 85. 8.58.5 8.885858

Saw—58 2 maﬁa 5.8 555558 58 8.58?» 5855885 5988

82.—255m .888th 857m 38:828on :88 2988 852.5853. 58 .88Q .985 582.558» .9858 88385
82 £56— N ”Ed.“ .3050 383% 8984 .9850 55.8555 85 85.8565 3.: 858m. .8533 ﬁt
“$58.5 3.885858 58 5858885

5988 .9858 S 5.8me 5.3» 98585 8558\58 855

88385 588888 88 8888.55 6: .35 a... 8.85.5 855

3% 85 85: 5.8535 855 85 382858

3.5 M55858 .85 8.9595 888588.58 835 558

Sum—58 5.858885 55.558 .38 «58$» 885 M558: 8.:

8853858 $55.5 2858885 2:88 .835 8.86.5 5.3. “.8585

.9850 85855 18.5 5 8.88.5 8558 8.5.88.5 85 85:

3.55.5 885.5853. 58 88% .885 5.5.5580. .4856

_ $.58 58 858% 859.5 88385 55.8555 85

88.8.53 88525 mmaé “sesame—8.68 3 £25 25: 88.8th 5.: 888.5 852.5855. 58 .885. .365 855
82 £28— m ”352 .3050 8 :oquoE “ﬂoccom 8984 \8 83589885 8.: 98:5 8558 58.:on95 5E ”mm

3.33— 39:22 “Moo—”HMNW 288882 umwwumwhoﬂﬁ— 88:89am

 

 

$33M 98 @3359. 8858.22 83m .8 55556 3035803 2 2an

139

 

3.95.55

 

 

 

 

 

 

 

385288.58
.8385 8.52
8% 885568-885
Smmgﬁ 8558mm 5 885858.55 885 85 88 588888 .5 “55.585 8858\
.8885 28.2 858 885 58385 8588885 855 885885 8858\8
58.58588 .8 885885 885 85 38585.8 885 M558\8Q
5858525 Sum—58.. .8 ..5 “55.85 885835858 885 558 55858885
53.5895 >53-~ 5558588 5.85 M5885 8.5 «55.585 5858885 8.5588
82 £8.6— m “$32 .3385 282 Q9385 3.855% 8584 558 8555588 8555 85.38.55 53 88885 88: 85%
888.85 8885 885 8 58.5.8 85
85 .8 588558588 885 85 85 588888 5.5 $5.85
885 558 885.5 58385 558588888 855 58385588
885 5.85 85 58.58588 88 888558558 558 688%
5858825 35:58 .8885 88555885 8858 58.5 8888 5.88 8555588
585895 >557~ 5558588 S 525 285 5.8 5858885 8.85 M58885 .8 ..5 $5.585 ~585R88>85
82 £38— N ”332 .5850 .8 8858885 38858 8553 8555588 8385 83.8885 53 885885 8858K 85¢
.888 88.5855 885.585
-885 5.5 58858588 885 85 85 588888 8.5 $5.585
885 858 88wa 58385 88888.8 855 58385588
885 5.85 88 5885588 88 888588558 558 .8825
5858585 3&58 8.3.85 585.5588» 5858 58.5 888.8 5.88 8555588
52.8855 .83.~ _888588 5.5 55858885 885 M55885 .8 ..5 $5.85 8858885
82 28.62 m ”Eu—I .3850 march“ 35858555 388% 8584 8555588 .8385 83.8885 53 885885 885885 .558.
3.58% 39:85.. ”Hum”— ...838582 «mam—”MW”? 8852;:

 

 

 

 

 

 

 

338m 585 £89285... .8858wa 38:5 .58 858555 58:85:88 2 25%

140

 

.ﬁmae

 

 

 

 

 

 

 

 

 

 

mwo=o>noubo
.383 03:25
.§ 232232823222 2.2 28222295
Sumauﬁ cup—own 9622 ..5 S h33$8 .2 $2.222 3222 222 332%
.333 03th $9.23 202232822 3R 2.23223 3222 3 298328
.3 .2323 3923. 3 333.22%» 9222 MSESEE.
520885 3&258 .u .2 3.2222 unm2m>2um§m .3232 222 2232283323“
32035 .33-~ 3:22:80 .83 M23282 .32 32.222 223222333» 9.22me
“oz 226— m “Sim .8309 Ban—om B23365 Shogun .8984 222 3222228 2332‘ 3326.3 3.: 3.3.3 335% ”amt
_%wae
mmocozuoobo
.833 282
38392.2 .2322 3222.? 23232022
Smmcaﬁ oncoma 32 .3 28222293 9822 we 3 23um88 3 $2.222 3223.
.833 3&2 22a 3222 2333 Ao22km2932 mﬁ .Em223 322$\S
298328 .8 22223 3222 S 3m>2um§m 9622 MSESERN
228885 Emu—58 .m ..2 ....MEBL 2322382.? 22232 222 32322333»
3:255 .334” ESE—:8 3 £88 3Q: .83 M23282 .32 $2.23 22323333» 9.2229»
82 296— m “Sim .3989 032 mo 20.9805 352:...» $33 222 32222228 23332 «2.2622 52 8.3.3 3&3 62%
..meE
mmozgnooto
.333 322
.3922 3220.202: 23222292-3222
Sam—BE eczema 2.2 28220293 9822 3 3 23888 3 «M2222 32223.
.833 332 222 352 2»?ch 202232932 3: «.2323 32223.8
298328 .3 2333 392 3 «buébﬁ? 9292 w2222o\2m&
228825 :meE .u .2 $2.222 2.822323%» 23332 222 s2m2223§m2
3:396 ﬁvefm 322288 .23 M23282 6.2 $2.222 22325323» 322»?
82 333 m ”33m .333 3&2 33.8 Raouocsm 5.822% 8an3 222 32232228 .3232 €32.22. 2.2: 8.362 32: .32:
33555 .8 2.8
8.38m menu—22¢. Eon—33¢ 28.2032 «:0hﬂo230— 3853mm

 

 

338m was $8392 .momoﬁoaxm 33m mo 59:85 30:25:08 2 033.

141

 

 

 

 

 

 

 

3&238
mmocozwoobo
.983 £280..»
.§ 2822,2282-u~22$\ 2.2 2858203
$958 oasowm 96.2 3 S 28886 .3 $2.25 $22 222 235%
.Eooq 22th 23.5.3 AuSiEuﬂh SE .932ch £222 8 2222928
.3 922223 ﬂuEQS 3m>2uu§v 2.822 msgo‘bwm
Snag—u .» .2 $2.23 nam2u>2om§m 222mm: 222 s2m2§£§mh
botanzm 208885 QBA 123888 .33 w2~282 6.2 $2.222 22u2§£u>m2 0222mm
raga 296— m ”21E .Eooq 228m 523325 Reason .533 222 32222228 2332 ~23ka 2.3» .222» SEER ”amt
sass.
mmo=o>zoobo
.mommon £2th
38982.2 322.? 20.2.8292N 22 .3 28220293
Smwanﬂ oncoma 9622 ma 2 23838 .3 $2222 3222 222 $22.22*
.3809 228m 2333 2022mequ SR “2383 £222 8 2922328
.3 ,thcﬂ 3222* S «buéﬁﬁNm 9822 MEESbuQ
sumacﬁ .m ..2 $2.222 3.222229%» 23$: 22c 322239362
3:033 228825 QBA 322528 3 $28 29: 8.3 M22222 ......2 $2.23 228223962 c.22QMc
82 .226— “ ”SS: .333 32th Ho 2220qu 353% 5234 222 3222228 $2ME which :2: .4882 mBEmk cum
sage
$052,883
.838 vacuum
.898 322.52% 2&222222éu22 2.2 28220293.
Emmanﬂ 220mm 222 we 2 2.8838 .3 ”M2222 £222 222 ~32$\
.833 32th 232325 20223532 2E .2th3 £222 8 232228
m2 «$223 m~222m\ S 338%» 9822 w2~2to\2m&
:mwczﬂ .m ..2 $2.222 3.22259wa 2332 222 a2m2§£§§
cow—253m Snow—85 23>»-N 3:22:80 .83 M22282 .2 .2 «M2222 “$233,262 c.22mmc
, 82 £33 m ”Eda .mommon 292$ SSS 32028.5 ”Haw—How Sumo:— 222 32222228 $2ME 33.3922. :2: $982 322$ SE
338% 892232 «woo—”wan”— ESEouo—z .mwnuMHM—Loﬂom mono—.82»:

 

 

 

 

 

 

 

333m 93 .8332 .momoﬁoaxm 355 no 32885 .Goacscoov Q 2an

142

 

.2229: 3x232 222 2&8. 3.322 .3232 3\
232 332.3328, 22\2~2m.~2 3 =2: $2.222 mum2m>2um§m

 

 

 

 

 

 

 

 

 

 

 

3 mmgﬂ 035%“ 222 3232.28 23332 $522332 22 22.2322
.mugmﬁcuonnm 2N .2229: .6.\2S2 2322 3x6“. 9322 63323325.
222 .mommom 2&225 3362 353232 ma 2.2x $2.222 mam2§2um§m
322 9.223%» 23332 $322332 22 35 2232.2»
3 $528 3 $228 .2 2N .nME32 nam2m>2um§u 222 32222228 23332
unocozaoobo S 522% .5234 3222228 222222232 22 22 222 ”$2.32 uuu2m>2um§m 222 u22mM2
28.8295 20208821238 hug—222538 .meaﬁusnnm 239232 32.32.2232 22 20 82823.2 .222 2.2 32222.2 :2:
82 £26— N “24m 25 .mommom E cab “and can .mommom 322% .3223 222 332.3323. .8 36$ 3&5 33% ”at
:mwcdﬁ
30233250
.382 29202
63.3232 963.2 2323 22.2.6293 m5 .3 282222922
3%ch 0220? 9822 3 S 288%» .2 “$2.32 ~35 322 335%
.Eooa 2920.2 23332 292233.03 ER .2323 ~33 8 332.239
.3 “.3223 m32§\ S 3m>2uu§u 322 MEESEK
Emmauﬁ .m ..2 $2.222 umm2u>3~§w 232M222 222 «222230332
223296 2028285 >23-~ 3:22:28 3 ESE 228 mg M2332 6.2 $2.222 2222363262 u22mM2
mamnumm 296— m “215 .9302 222—om mo 202.8205 2 ”Hagen 8934 222 3222228 223.3 32.28.2521 92me ~328k ..omm
Saws?»
mmocgugbo
.282 0:252
.393 322.2922 3322222$3EQ 2.2 28232222
Emmy—.28 occowm 9322 m2 9 33228 .2 $2.222 ~32 222 m32$\
.muooa 22th 23232 282229.283 at 6.3222 ~35 8 332.58
.3 923232 m32$\8 32.23% 9222 w222to\2m&
:mwaaﬁ 3.2 $2.222 9.323229%» kiwi 222 2222283282
cocoaqsm 2020885 $27~ 3:22:28 .22 M22332 .92 ”M2222 223232323 9.22»?
62 £33 m “33$ .2qu 22th 3322 3202055 3.022% .8234 .222 32222228 32ME 3.262% :2: 223% m323k 2th
338% 39:23 hoe—”“22”“— Ecuagoccz amwwmwww“? 8352::

 

 

338m 28 £83242 .momoﬁoazm 2825 no @9285 30:22:03 2 033.

143

 

8:3“. 353558
\o ”MEBKKE 888‘ 835 585 888.5 38:. 323558

 

 

 

 

 

 

 

 

 

 

 

>55; 3 "am—58 5 28588588 .38 M85885 8% “538 253558
62.5996 <>OZ< ESE—Sou how 282 Shoncom .8284 Losx8$m3 ma :5: $583.58. .8888 8~c58k .. ~ NI
. .8885 2858888 553558 88 258mm,? $585.58.
8 3.8%? 5.5: 98888 8358\38 835 88.53 888888
8.8 8888.98 .5 £3 .3 .9885 885% 58.2.8898 85
8S: «.8883 885 2.5: .85 9858883 355558
3 $58 88 8.58% 8.85 $5.88: .86 $585 5858888
280mm bow 55.558 58 8.5838 .832 838.585 83m 33 98588
885.5 53: a... .9885 835.8 8.58qu 85 8:
botanzw 530825 %m>»-~ 3 3.58 S 828 28: 8385.58. 88 888% .8588 58385 $552385
82 <>Oz<2 85:88 how .«o 583on iouaom H.684 85 88.8305 2...: 988.5 8558 8.5888. 2: .8 NE
. $9.85 5858883 553558 88 8.58%»? $5.55
$88 2 38mm: 5.5: 8.8588 8558\88 855 89:85
88888 8.8 8888be o: 58.8 550255§
g5 8.8888 835 :5: «85 9858888
3 $.58 553558 88 8.58%» 8.85 M58885 8.: $525
oncomm bum 88588888 353558 .58 8.5898 .833 8388585
83% 3.: 8.8888 855% 68.8 55888553
Btaqam 58835 .33-N 3 $558 -855 5 .mwtccxubmu 58 858% .8888 «88285
62 296— m ”2.5 858.88 how 388 88:85 8565» 8284 5552585 85 88.8.35 :3 3% ..3 ~h
3885 5858888
55558 88 8.5.8MQ\0 mm=25$88 2 «58me
.3.: .5th8 8858\88 855 88388 388858 8.8
.8888 .8 85 8.58335 8855588558.“: .98th
3 "”958 855 2.5: :8: 9858888 553558 88 8:89» 8.85
oncowm bum $588: .8 .c $55.5 2858888 553558 88 858mm
.838 838.4588 83m :5: 8.8888 835% 8.5885
3:896 :28825 83$" 3 3:28 V8855§8~a5 5 .mMEBLubmm 58 888% .8588
£888 332 N “24$ 85888 bow 359365 Shanon 833 8823 8552585 85 85.835 2.5: g SQ Ex
833:9 one—natu>
3.39m nonran< 28883: 9.88352 83:2.qu 8853mm

 

 

 

 

 

 

 

$3.8m 98 £835? £8230qu Exam mo b.8883 $835803 my 033.

144

 

saga

mung—0500b».
.3050

smmauﬂ
oncoma .Eoﬁo

5M5“? . .238 8 38M»: 5::

Eunuﬁ MNQEQEB ~35 5.253 «$38.98 98 umuﬁtgﬁ
o: <§m~ EmEmNazuséﬁEﬁ us a: .Emhuﬂ £3:

~85 3338.95...» 3.8 MSESKLmQ at “.355 muuzméﬁﬁh.
$33 38 3.2%. Sﬁknuﬂ NSSEES ~28 022mm:

5 EuEQSSSh 985 Mﬁhmmz 6.: amps? Ewﬁmﬁméh
3:35:89 SE 953%» $3: 9:8»: 3: 9:»th

 

 

 

 

 

 

Emu—58 $2.8»: <3§ a 2% :x “.385 «ﬁgﬁxoezm
8:255 5388:: 33$ 35.858 38 £me 5.3 3:358». Evie ES xmnzmm .633:
“oz £32 N ”24E .ﬁoﬁo SE>2 3984 .3 598m 833 2333 35.8.23»: 2: 89585 3: Mg ..EE
3 mmczﬁ
228% so
anoohwm ~28
035938.23
3 manna
- 3:25.80 :0 Suns: £9: .8\ :35 runs»:
Btoaasm $34 2883me ~28 u~UEm\xQ\x§ME ma 5: Swans EmERSmﬁw 92$va
82 <>OZ<§ Bacnzonsmﬁum 282 5 320m .833 ES B§§=ou no 32:8me :33 auﬁhxonEKEm. ..MNE
.2229: .8.\ =3: :2: .8.\
mamzméummﬁm :86me 35835 ”MES: 222:0 S 33?:
ANMFSKE v.65 3 3.: .35.” uwgzmmia EmEQSSSh
.8.\hmu= .583 u MSBEVS nutzux§ma 598.: .850
5 .3th3 £5:ng :85 35: Sbgmnsmx 3%:ti
«swash. was 3 2.2: um:.§x\c $8. SE 2: ES <22
Sum—5E awnhux$mu EmEQcNguh 3:33» «:8 “$38.: amtméﬁmﬁv
@38ng 58825 unafm mmocozuooto S mmsuﬁ $35338. has :me 5.3 3:358» .233 2.823
82 £32 N ”24m .3050 unaccom 833 eczema bum 35.20232 m5 $9532: 5: Sanqumhgu .3:
3.58m nonran< “Mu—”MW” 9.3.2932 .mwﬂMHM—wwwu 88:8an

 

 

 

 

 

 

 

33.8% new £8324: “8853me :35 mo Emﬁﬁsm .Qoscscoov 2 03:

I45

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

woo vow owmﬁ :Mo ooh mm: tho $6 2.: go New mmv :o mow in who hmw o; mowmmvm
3 A New mg Quo oow mmm mwo omw :N oog bow wow who Sum mmm Nwo New wNN MMWAWM
2A 50m hmmv owo m: min woo mew $3 cog 2% ﬂood and mm.m wmmﬁ Nwo Sim wmﬂ M50“
:A new 5: mod om.m m3; woo ova 23 mo; 86 mwoﬁ who mod :2 omo mod 22 0 ”MW“
2.“ Sum no; owo. mmh Nonv 3o mm.m oobv cog ooh Go go mbw won owo new mom moomcﬁwnwsm
mm; oﬁm mmom moo omh 3mm moo hmw oomm 3 A ooh 5N2 amo ooh .3: mod .66 mo: maﬁaEMuM—mm

-. -- -. who win 38 30 cm: 38 -- -- -- N2 9% E woo 3m to _ hum
QM «83% Z QM 23: 2 QM 2am: 2 QM :3: Z QM =3: 2 QM 2cm: 2

mmucgnoobm 8:25:00 oncow< mmocoéootm 3.565800 oacow< 23am
2233 0.52 9.2.3..— 225,..—

 

 

 

Eonmoq 228m 23 032 So 32$:an 9855 can .882 .3 055

146

Table 15. Summary of HLM Descriptive Main Effects of Leader Gender and Leader

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Level
Independent Dependent Y” Level 2 Parameter Estimates
Variables” Variables Coef SE df 'I“ R’
Others’ ..100.-. __5_ 5103 .. _0 _()_(_)87 3311 632.72:
communalmings“ 7m -0.ll96 0.0194 3311 -6.l6* 0.0037
- . 806 0.0091 3310 604.91*
LC de de L2 Other-s _ ....-o-109........ ,,,,, .5. .51-...“ ......o - .--...“
" ”5°" ' agenmmmgs” 701 0.0741 l 0.0200 3310 -3.70* 0.0013
Others' -100. 5.5395 0.0131 4 2997 421.29.:
effectiveness ratings” yo, -0.0837 0 0291 2997 -2.87* 0.0010
Others’ W700 _. .- ...... 2.5.19.8 ......... 29.1.9.5... -..... 3.3.92. -2992: .
communal ratings L‘ 701 -0.0276 0.0210 2305 -1.31 0.0002
' 07 2305 $17.53*
LC dc 1 113 0111th _ “-109 .-.- _..“§ ”5.1.9.9.. ___..0,Q.l_- ..
a ' m agent'cratmgs” r 701 0.0474 0.0214 2305 222* 0.0005
Others' ...-..100--. -...-é-ESL? _ --9.-.0.l.§3. .,.,..__29.7..§-. i..-§5.2;§9i.
effectiveness ratings ” 701 0.0443 0.0316 2076 1.40 0.0001

 

 

 

 

 

 

 

 

" = Leader gender is coded 1 (female) and 2 (male). Leader level is coded 1 (middle management) and 2
(executive). All variables were grand mean centered.

b = 700 is the intercept variance coefﬁcient and Y01 is the slope variance coefﬁcient.
c = p<.05. Values in bold are the slope signiﬁcance tests.

147

 

movo 1..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mooo momoo 283$ Boo
o3 Son 898” Sum moose assist“. .3650

33.2 3.5 _ 3.5 52638.2

mooo Soon Soooooo Eco.
9o soon 9%? arm amaze 35mm .235

2.90.8 8% o 8.3 . . coooomceé

vooo momom 3:28 .309
.1 mod .ﬁwwON oﬁﬂdvlﬁy - .. :1.- il. l illusotmmi owning ECDEEOO .2250
$53 2.3 o 2.3 sonoﬁeom ..

Boo Rom axiom Boy
31.. ovo . Rom $.32.-.- sem mmooeooﬁmﬂsm

Room.-- WWW-m...- io- 3.2 1 -- ozswowoaw» mooowajowwmmmmom.

vooo - Rom eat-moo - - -1:- --1 1-1-1. Grok
omo 88 ohm? 5am amaze Eseeooaom

..o-oi- :o o. Dim; ....-oz.3mow61m,mo.o§1Ego-momm-

oooo 11 -. .33 8.53: .- Eco
-..-...---- --.m.._._o-.- ...-mmmm Sow-moi..- sem aqueooswssm

m2 moo _ moo - 52% $83

Nooo boom omommvs 33
:111: own-poi: .mmom .- ..Q..ow.m1H-.1l. : 11.1-1 -- - 1 is .-o-mow- owccoZoSEEoorom

...-o: 1am- ..o. .--. sow-N. - 1,: .:--.-.cooommmomom-

"M 9m m2 at mm out—om 033...; Eon—anon

 

 

 

 

3250 58¢ 98 430A Snood .5950 533 we mooobm 5344 333539 852:3 .«o $92.23. mo boEEsm .3 03oF

148

.mo.vm M68888 oo§> anew gooﬁng :08 8m 52 8583a 233-20 Bu 2 WEE u o

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

23.58%“: .3 Ea 0: Ba 3:8an8 85.53 95% 5953 05 ..8 Eon—Fa 05 .239: _o>o_-m can 35: 05 .8“ n a
oSoo in? o 88o 23o $.38" 23 oommo 3&5... . a
.uounon 012
88o 1 2m o Sooo Sooo #32; $2 oSmo saumnouoﬁm 3.2. 358.5 3.2.5» as: 6
:8o .52 o oNooo 83o Boo-am: 22 ES .1..._.e_.§eea
.332. 012
82o ..ooamm a. 82o 32o .8on N3 mono 53335366»
..o-non £12
22 .o Loo: a. SS .o SoNo 5.32 no: 32 .o sauﬁoﬁﬁ 3532 as?» .023 3
£2 .o $3.2 a. 88o Mano .38: no: 88o - 3,95: .2258
.332. 012
«a so so so Rom-o 5&82 2 mm mm: .o sass. 33.5.8 222: sea :33 x
.9850 23:0 cow—omen— 3
Rooo .2 .2 o Rooo 38o 3393 38 G: .o 32...... ...:wa as: .322: as?“ 5.3 a
5 o .o #33 om woooo 32o $023 32 2.2 .o 35:21....wa 3335 saga £3 a
so so Q: s: 23o .oo_o:2 moon ammo :33 .wuﬂo 2.55» 5.3 scissohwﬁ v
s: so so so Rooo $282“ moon mono :95: a
.2050 3:82 .8 comb a
s: so g. 3: ago L382 2 mm 2: .o saga 5&223 2:: 520%: s u o 3 m
@350
omooo .82 o ovooo oENo .3382 Q: omzo 35:: “who
. m
«woo-o Room o omooo Ego 12mg 32 $2 .o 35.... Loam as: .222: 25.5 323 N
omwoo 13% 3 mos .o ES .83: 2mm moono : .. 3M0
. E w
o: o .o 333 on mmooo 32o .323 22 mm: .o sass. aoeommﬁ a 5.5 25.; 53 _
00— .=. 8: 00— .E .9. \oz 333:5
3:25 :50 3:35; n :53 3:25 :80 35:55 N .33 mom—rm a .954 333:3 Suzanna 2220qu 2.2.598:— I

 

 

 

 

 

 

 

 

 

coasts, mac-5 Begum Sm menace ”—802 :32 05 mo $38M .2 03$.

149

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Soc-o and: a $8.8 some ﬁns-ON 2.: 38.0 3.952 £52,3-:-
..-..-::» -- -- -- -----..; is: ----...: iii---iisi :32... wast
#000 o ...bc ~N o hvood movmd *VMNNNN mmm~ mmomd Summhwﬂ ”WWW...“ A 338 383055 389:.» 8:34 am
38.: as; - a .v-omo..m:--:sw-mai. _-u~mww~:~--:----mmms--.-Nd~»oi - s: -...--_-:.ww-§:§aaa
: .33: 23:0,.“
was: #3: S 883 28.0 :32: am 3:: gas-535:3.
-. 2:. .---2.-.: .2: ...:i .-.--.- ..-... : .-.- ................ , - .--s: .--: --.::-...--:..:-: :i :_ --.-_-: ---.:.--:-: ..ém 1%»:-
, 28 o .8 com a. 388 as: .32: 3: :28 ...-WW”... ”M.,-Hm“ 35%: 3.023 as: a
28¢ .58: 2 £30 .-wwwwwdé. -.um-..%-w.:..-.--- -.._.$..::-_-..mwmd.-s-..---.------magmaEase-u
.33: 23:0:
s: s: s: a: :33 .222 3: 83 3.335223%
s--- :: -.--2.----. - :-..-:: ................ :iisaomoweo:
«E s: s: s: 3:: .38.. OS 38.8 3.3.22.5». . see: a a? 3 o
2:-.- - _. -2 . ., - is... 5..--..- - --.-.- -- - .. -.%-:.£ wwsm. 22cc SE82: 3 u a A b
e: «E s: a: :88 93$ 03- ommo 0 ages. 25558
.330: 2:50.:
vmmcd .592 a mmood mwomd *8.ch Em wmvmd 3:952 383:3?
-. --.-...: -....- -- :- - ... -..- - - - ..-... - - ----Maméwiﬁ:
88 o ...8 2 a 38 c :8: 2+8 8m 38.0 ...-mmﬁmwnw“ :3; Ease: as?” 2.3-: an
88.0 32 o 88.8, . 28.0.: --o-«gm--. awn-..-- 8.08.0. -- -- gangs-«€959.
.mommon 2::—cum
33¢ ...mﬁmb 0m 2.3.0 owmmd ...wmeom mmm 2.38.0 3mmc:em8=o>_§to
- :- ....... .. - , - .. . - - ..... mesons-“ea.
.88 o *3 3 2 88 o 2: .o 3.23 can 88 o ...Mwmmmmmw 3:26: :53 as: a
32 o 338 2 $88 58.0.- .-a..3~ 8N ., - .38 o .. macazsaees-
.mommon 2:th
s: s: s: 3: 52¢ E .wam 82 83.8 s “was: ...-0522:.
-. : . .- - 1 -..--2-.. yams-0:2
m\: E: m m . . . $5.2 28.-um: m8 :
- . s: - . i :- . \..c.. ...: .1 \=:h©©NO *mmopw . Moo-ﬁg - wow-mo. - ...- .. ...-....Mommom 3:2 -. 2::—.3 common—08:“ Sbnsom 8:34 oo
3: a: s; s: :33 .322 82 22: sass 25258
.mommon 2:2
00— ﬁx .8 8: U2 9% E of \6: 8338:»
3:83:80 3:223 m .964 3:08.380 8:325 N .964 :8 ~ 8.54 8338:» Sousa—on 8:88:82 33:23::— I

 

 

8855 98:0 :838m 8.: @539 8:02 :32 05 .8 333m .Goscsgov E 2an

150

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

s: E: s: a: 33 .2 .NE 88 Same 3.952 32.2%“
.EE ..E E E mad-.-- ---ua...§-1-- 82 as .E ...-5.0 2.312.. 312.51... E
.-mE .......... S: -.mE-111mE-1-1s --.mwwmqo ...... -u.wo..-m.ww-w----i-raw-aw ..... {Wm-910.1,...-1--.--.--.-Magmagammy”-
mE Sc 3: «\G omwmd *Nbﬁnkmﬁ omwm Sand 3 $5.2 mm0:o>m_wu%m 3.0::0m 333 3 amaze ozcomabom S
S: s: s: s: 88.0 1.3: a 38.0 saéewzmwm
WEE-.11.. www.- -m3111.1.m.\1:11 -1 ZVONOHO 1 1*VBVF1di - -. -WWOO..O11 ...-111. ,. . Ewmﬂumus-qamnﬁsoo, .. 33.8 ices—Sm 3 Encom .5304 no—
20m
3: s: s: s: E: .o ..z .wmm on 8:3 3:52 an...”
- ..1- - .. 11.:-.1 - 1.. .- .1 - .1 ...-1-111 ...-:1 :..:...1.1- -. ...-.... -1.- . 11:: .1-- .1- .. .. -..- . .. .- 5 :0 Com .5 a an:
s: s: «a s: E 3.: o 12.2: on 28¢ a as... 155%“ a as s E E 3
s: 3: EEE 8: N33 ...?va 0%“ ”move 25.: sammmmm
amass: «mag—60:0 3:020» .0301H . as: ..mommom
3 . . &
320:6..33 :5 .mommom 3 099 83¢ 3 $53: 3:38:80
3336833.
«0:: .mommom
S: S: S: S: I: m.o *2 .wovm meg coomd : 3.52: 323.08%
: . . _. .. ..caﬁzea
S: e: E: s: 33 18.28 $2 9.2.0 sass. use”: 3222 55» .523 um
.300: 022:0.» 2::: ho comtoaoi 3
.EE- 1 E... .....E. ....E. - . $3- -...E.-_.§. $2.-- - w 03; _ .. .............a......
.050: ...—«Bow
8: EN ... 2.. :8 3x .... ...: E: 322.;
3:33:80 0053...; n .96»— 3:0::%::U 3:323 N .054 1:: H .954 335.5, 0:230:3— 232032 30.52.03: 3

 

 

8:0:3/ 95:0 :00250m :8 9:33. ”3:02 :32 05 m0 338% 3035803 2 033-

151

 

.3000 00:80-00:30 5.0000088 0-000000 05 000 Eon :0 00203 .mo.v00 .1- 0
00:00:00 500:0 00:00.0» 0.003 0030000314. 000005 N 00000 00000000.: 0 00008 00 0000:0w 00000-0 n E

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10.00.001- 01- 1.10.00.00.0-1M0000...- --00-0-. 1.... 0 --.--...0-N000 0N00.0. 000000000.-. 30000220500300
-0-0-0-----0------0.E00.0 0N009 -000--N-.0-0-..0-- - $00.0; E0000 0000.030 .---.---..1-.-E,...-0-0-0..0-w_.0m0-
100.00.110.11--mm-00w0-10-m-00-0.0-010w1 - -E0-0 ---..- .0.-1---. -.m0-0m0-0 0000.0 0000:0000 3000.05.00.00 300000200000 0050.005 00
10000-110 -..-.... 0E0..0.-0- 1000.00.00”- ...00 0-E-0 - 0100000 00000 0000000 0E 03.00000: 3 . s

00.0 0 080.0 0000.0 000 -0. 00000 .-m0.00u0_ ...-...00000000E... ...wmmmmamﬁm-

E00. 0 EN00.0 0000-0. ......00 00N. .0.--- EON00 000-010, -..0000000E- ...-0.000002
...-.000-.,-.E0-----E000?000000.--- 0.. 0- 00-00000 00:000.,-0.-00000.. 0:.
-..-0.00----- ..N0_-.-... ..000‘0..,0.-..-...0-E..0.-0.-0....w 00.0- .00....- N0}... 00000 ;E..0.E.0.0...0 0000000000; {...-00.000.00.02-
--..00...0..--.-,00--.-,--._0.0000,.000000-: 0E 0- .- 0...; 00000 0000.0... ..,._0000E. ...-3.00:. ...... 0 ...-...... .... ...

E00- N0 0000.0 0N00.0. ...00 00 N0 0N000 0.00.00 0.00.000E 00:000-000.: 3 0 s 0
.‘..-.0E..0H.-.,_...-0_......0000..-....000...0..___. .. .0-. :0... , .0000 00000....000.

0E0- N0 0000.0 008.0. .00.. N00- N0 0.0.000 E00000 0000000000 000800.02
:-0-0H0..-.0.0---...-.....0-N.00..0: ...0-0000..-_ 000 .0.-,- E000: M0000 .. --.-000.00% _-_...0002020208 022002002 0050:0003 00
0.00.0- 0 0000.0 E0000- ...0E. 0.0N 0 00000 00000 00000000E ...-so a . - a
-....E.v..,~....- mvwoooo-Rcoo-m 20-30 0.80.0. ..voooﬂo- 000500000.- 3000002005868 0000000005 0000:0000 0 0
000.0- 00 008.0 0000.0- .03: 00 N000-0 00N00. 0_0E000E .2205 a a 0 0.0 0
7000.--, 0.-.-.. 00000-,0000.0..--... 00,.-.0..- 0- 00000 -.-/0000.0... . --.-.0000- ....0..........,.0.0.
..... 00-0..i0, 0.0.00.0--.- 00000..- ...-E0000. 0. - 00000 0.000 E00000.._ -- _- 0.0 ...-0.0.0.... .. ...... .-
200 T0000-00--..0.000..,0... 00. 0 0 , E0000 2000  .._0_0,000.. ...-.00....150010 .. . 0 0 .- -_ E
...EEN- 0 00N00 00E0.0- ....N0 NN.0 0 0:00 0.0.00.0 000E000E ...-.205
.-00.00me- E011 00-000,- 70000.0- 00-0 E00.--- 0N..000 0.0.000 0.0000000. 30000050030000
é-cmm-Hmv}... ...NwV. ..... 1..®--mOwO--w-m-N1.OO- .- WNF. .00.-- -1 . WV. Nm-w-WO. O... r -.V-WHVM.W w—OPENOOOF. - 1-1.--.m.._-o.m-«W 1 EAL-$300: .3 Com .5 s
...N0.0- E0 0000.0 0000.0- 000 E0 0N0-00. E0000 :00000E «0.000200200000000 3 s 0 0 3 0
-.-0-0..0-.---.-.E.0:--...00.N.0.._0.0.-....N-0.0.0..0.--.T....0_0. 00.0- -. E00.-- 0.0.000 WNW-0.0m -.0._0E.000E- .0205

00 00 000 005 00. 00. 000 080 00.000005 0000000003

0003:0000— 000080.005 20.5-— 00.0:5000 0000:0000mm_0>0- E 0:20:03: 0.0800252 0:000:335 I

 

 

9.00 00:00 .93. 5-00 5.00 .N .0 000050903 000000—090. 2E 08080000000 E003-N .00>0A-m .08 000300.030 00.08805 .00 030-0.

152

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-N.0,.0.-.-0-.: 0000.0 0000.0- 0N: - .-
000 . ....
-000--- 0 ----0.0.00.0- ..N000 0 0.0.0-.-- w 0.- 0000.0 000.00....,--.....E._.00E... 0-00.0.0-

-.....0.E..0..-- .0....- 0.0.0.00- 000.000.. 000 N00. - 0 00.000 -:0.0N.0..0 0.0E§E- .00....”000000 020.000.0000. .0 000.0 .
...-000.0110- -.. -00.0.0.0 0-00.0.0..- 0.00.. 0 .NN0.0_..0- .0000...0...-.......0E000E. 00.0.0.0 0000.00.00 a . N. 0 003 00
000.N- 0 0000.0- 0000 0.3.0.00 000 .0-- 0E.000 N.N0.0-0.- 0.0000000- 00000:”.2508
.-....0..0...--.00 .. 10.8.0.0 EN00.0. 000 00 .0800 0000.0 ..0E..00E .0000 000.000

.EE.0.N-: 00--- 0000.0 0.0.00.0...- ..........N.0 0E ----.00-. 00E00 10.0.0.0..0 0.0000000 3000003000.“..00000

....E0..N.....: 00 - . E0000 0.00.0.0- 0..-00. -00 . EN000 ;00000-......:0E\.00E- 4.000.000.00-

- N.0...0..-. -00- - -00000 . 000.00-- 000 .000. .00 0000.0 00000 ..0.00000..r ...-000.003.0000 0.00000 .0 0 0

.000...- 00 - 0.0.00.0- 080.0..- .. 00 .0.... 00. NN00.0- 00000-- ...00E0.00.E 0.00 0.00.00 3 s 0 0 000000 00
0000- 00 0000.0 E0000. 000.000- 00 N0000- .0000N0 .,.0.0E..000E 000...“...0ﬁummwa

.mc..—.--0-i -860 c.086- -30.. 0 Good Soc-o- 0.000000% . 00

.20. -...0: 0N000 0000.0 ....00M00- 0. 00000 0.0.00.0 0.00000E. 3000.0.0000020000

00.0., .0-. NE.0.00- 0000.0 000 -. 0. 00000 .380. 000.00.... -. ...000...00...0....0.0.
.. ..0.N.0.. . 0 - EE000 .0.0.N0.0- 000.. E0-..- 0 - 0000.0 0N00..,0 0.000000 30000003000000 020.000.0000 .0 000

-.hNd-ni .w- -. NON-ob wNoouo- 000.. o- w .Nmood. 300.0- .000EE00000 00004.0- 0._-0.:-0-0...0- 3 . N0 00 00000-0 00
00,0- 0 0000.0- 0000.0. 0.0.0.000-- 0-. -..0N00.0 -0000-.0 00000;... 30000000000500
0- .0N00-.---0N -- 1.0.0.000- 00000. 00.0 0N 0000.0 N000.0 0.00.0000 00000020200

-20.- 0N 00.0.00- 0000.0. .0000--- .0N- N0000- - 00.0-0.0; -00....00 30.000.000.000.

-0.E....0w-;00: .0.-0.00.0- 00.00.00- -000-- 00: 0000.0. 00.000 000.000. --.-0.0.0.0000.

-00..- 00-; 000.00-- 000.0... ......0..-_0.--, 00, 00000 N0.0...N.--.-000000- 00.0.0020 0.0.0... .20

-.00...0..-- -00 .. 0000.0 .0.-0.0.00- -,0N.0. 00- 00000 .0000.0-.-........0E.....00». 0..-00.0.0.0.0-00-00-00. 0.. s 0 0 0000-0 0E
000.0 0. 0M0 NNM0000 .N..0.00.0- H0000- 00-. 00000 000.0 0.0000; 3,000H000ﬁummm

.0000 .0.—- :0 Mm 300
00:00:30 .00 080.. 0033.00
n.— » «mm—9'04 33::me hoaoﬁagamNB>uu~ 0.» «cu—509“” 9.3.0.3632 «mwﬂMMm—Luab— :

 

 

I“ a“ a I“ a I a I a a .
N0 0 w a n n 00 n 00 N 0 00000080000: 00030000000 2A3 00000000005 E0034” 4054-0 00-0 00000000000 @0885 30000000053 00 030-0-

153

.38. 8:85:me :onmuouoE 232 05 ca Eon E 833/ .mova n a
.35.:8 :38 2me 803 SESE> =< .855 N 98 3380b. a 368 ﬁ 82% 833 n a

 

we as. w...-2:: : 28.0 .3» :
-.--mwpn-z.2-.-..-.»www- :.-:. $8.0. .- -_ 88.0- _ . N2, saéészséo
2:- :3... .. .. 13.3.0--. .- :35me 55...... .3222:

23:~§_w§o_ -.N..@.w-m .. , mé..- .. --
N; €222.52.....a.-~.8..o-......-..2862.-. 2.3»

._ - Go.- ., 82 . - 38.0 28.0- s»: . sage-3:5?
fww.On...-a -...-. -.NOWWLE: - 2NW.~. ~20.- .. NomﬁOO-L: . ..- .6»- .wommon 25.4

3 33.: 25:. we coanoi 3 365% ...-.68; ac

.22..~._m.m._.s8?--.-- -meégigw. - ...é. : -
Egg . £80-. . :23- -. . -é. -. . .

22%.... .é..-..-MW 3%-... . N32- g ......__..._.....§
30 3E:- -.mwoo.o omwod- 5» -. .3382“:

......owmmm 32 $3.62..-- @399. . ., 8*

 

2.. 8mm woo: wooed 3»

Ill.t.22-2‘Q2-t..u.ll\1§ 2... ..Il...4.. 41q|.. . 21.-‘2|. ..c:.\ ....l. .‘ ... i.. u 2 2 .l

8. 7 82 m H 8.0 2 cos. § 3 as: gees a m
“ac-m...- -. . momm- -. -- . v3.0.0 . .mwmwd... .. . .. . 55>. . . @350 3332 29:28 :oEo oi 3395 .253 on

www.mmo 32 58.0 83m ‘ .. -8».

154

 

#5. moon- . . -280; 8.0.0.0 - a»
. r . . .. . .. - . . . 33533353850
.....3.~.... .- -..-maom .. .. ”8.0.0-.- . Ema? - E. - .225
.. 2 ...-2-- 22..-.- 21.2 : .. 2: 2.2.2..- .- -. -..-.-.2 .- .. - 2- 2 .2.2--.... -..... .- 2.2.85.2-..- 222..- ..-..2. - .----..:- 2 2-22.- 3 ESE 25:28 coEomoi «ioncow .533 m

5".)2L’If 12,-...l2...:|.2 . 112-02 2.: ...Ivtl.!.nuy.-'2l.i. .. 21.12:..2 :2lo.12'.u-u.tn.‘5 l..- in! l 2.».ll I v. .2! . 2

a.,: a: . 2.80.... - $39.-

3 mmcsﬂ Eoan_o>ov

-.2,m..m- .. nag-...... _. «NEE-o...» w... .- E-_...9 . .. - -- i- 2.- .225

tawnwmo-yé 82 38.0 .33.}. -. ._ a»...

 

 

 

 

L- :U Hm ~80 > 83323 2.2.595 2898—52 .8335»

 

33:53— $85.23 N .954 33:25::— :

 

 

 

 

 

 

 

WM can .NH .ow .05 .00 com .M m0m®£woa~hm ”m0m%~d§ 21mm COMHQNHOHCH kmguN JO>01~2N ..Om 3:;QO m0 NCQEEDW .OH 23ng

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

...-E a: . _‘ GE..- :80 é

-. MM-Haa-I ..2 . WMMMi-i: ..2-.M.©©w..o- -.- ©~0mﬂou- no»... 3 mmcta «magnum-co 8 Sn
L... .223..- .2:- -:..-. . 2:22. .- .-- wao O OVVM O a». .9550 3 6 “owns 3 mmsah U_~Cowa.w_om N—
333. can $3.0.--- §-w...m - 8»
2%..- -.2 ...ec--- -- . ._ - ES -- ~22 g

-Nb 7.. 2 .- meg - - NQOO o NNOO Q-n S..- mmcsﬂ mmocoaﬁoto

...-3 m. - - £2 -2- - 82 o - 0:3- . E 3 . .2333E
...-ommﬁ- .32.-.. , ...... 88¢, - . $an -- . E

-.HE: - 0mm“. .. - £8 om. . 38.0 a». -

0v mu 2 - ©mm~2 - - mmoo O. - Magoo-..o- - -. 3% mwczﬂ ozzowa
-.uwﬂ-M 2M..I2.:.-l. -.. . .IW-m-mzH-f- 2!..1; - Vd me O w—FN.O.I . _Ox. 3 .mhwon 07“th Smqudh o_a~:ho COCLOQObnm n4 EDUCUm uoﬂmod ow
2%“.Qm . om: ..m.w~-o..o§- 22mm...” _ 8». .
m: 22 ”mygdm-m :03 .. é.

.*mm.m....;i .. .Ommm .2.w.NO.O.O. - ”...-mood” - no» . mmcsaﬁacsEEoo

.....a,.n-. GB- ; 2%. - 2%.? g. .2 223.2.-

232; 32 £5.02- . 23w - 8»
2:. E. .. ES O82 é
mm MM” mewwmmmm -. .. --me ...... sasﬁaaazuéo
.-.-...-..2...-- - .1213..- . 2.- - 2 2 .333 oEEom
...»ﬁwﬁ..---... 5.2... .. .nwmod an? - .8». -. -
.2852..- - -- 5.2--- - 2 E8 0 -0890 ........§. . . - --
23b.- --5-22 $8-0; - goo... -. -.-.-ﬁ-.-- $52353
.-...W-miawus-J - -%m.«.q:-E-2: 2 - N25v2~ O bwbo Qu- - ...—ox. -. 3.3mon owns—om 3382 22.-E cofoaoi 3.091% .8va on
...-ﬁg - E. 3.8. o; .- ;.--awn...“ -- -. ‘ -.-é .. .-
- mg... 2--- Saw-2::-- snow-ow €8.0- -. -3» - - --
225..- - EV: - ago-o - -... mano- -- we»--- gaging-.0
03 $322-.-.- gwmao E: o- 5» .masﬁzea

13w: E. 28.0- .mwwow--- 8» -

L. .6 Hm ..oou » no .— 832:5
3:3: m ..o 2:53 33.—5w 32:5 0 Ea «..o c a .
u H w m N .954 n g u E Banana—on:— E

 

 

E can .9 .ow .on do .om .m 8850qu ”83:23 225 song-.85 >a3-m .~o>o-H-N .8 333-"Co E885 .

835803 3 29¢

155

.156

 

ML... EN; 2:; N2; £2.
0m .02 meN no: .0 095.c- 8» 3 mmcsﬂ mmocotﬁobo

“mg-......--......-.H.. ..ma.m-..M.---.... .-wzw.-- . so: -.. . .- _£.--- ...-..5

...-.m.wwm.l..... EN... -- $3....9mN.....”Ewn...._-..  £.

...-.9. . ..M..._.,.-...-m.£N.-N,.-w...w.--...-omém .....H -.-. mw-mé.m...-.._ .4 .£..-: - -

_ :.0...- - an ,2. new: -. . ...-Eon... -‘ -.--£- . ...-£222.. 2.... .
-.m-S...N-.-- ...... -82?! 85d.-.- ...-22.?- é. .225 s_ _ Y: 3 u “.3 2
...-20.93.25... .83-- ---.,-.-_- «.23 - .. #3.me- . ..o:-

-..-....m.~...- .-.”...NN2LE... .53--.. 2.80%. .- .£..,. .- - -

.-....m_.w.m..m..--..- ......sﬂ............-._.m -....._._.&£o.-M-- , “9...“.6..R...--._... ---...£..-...-.,. 3 as... ...-seas

2:32. -..-:3 83... +. .£ - .28
:.0-2“ $2 3:5 923 8»

 

 

 

 

 

L. .=. mm ..oou > SEE-:5 33:83: 339832 .83325

8:2:ch 38.52am N .33 2.03335 m

 

 

 

 

 

 

 

 

E 25 :2 .om .3. do .8 .m momoﬁomxm ”39:23 33E cos-2285 Nag-N .3522” 5m gamma («0 $883 $035303 2 2an

.380 858%; 5.5889: N- :32 05 2a Son 5 3203 .3. VQI 2
.3888 52: 98% 803 8335» =< A295 N 98 6380b 0 368 m. beacon $9032 2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.0.002- .0.-----20000-02- -0.0,00..0.2---00.0w-- 0 :00 0 0.00.00“.-. 2.00.00--- ...-.0... .02...
--..00...0- .. .. 0 .. ‘00-0.0.02-.--0.0.00_0 0000.000.-- .0 ..--0000-02 00.0.0.0..-- 0£202.. ...-.222...- .......§3 .0.
tam-.0.---- .m-2-; ..o-So...o -.--mano- £5 22:00.2... .285 Soc-o :30» 0 300.002.05.58 3 . s
3.0. 0 886 9.8.0.- 23 SN 0 .. 38o oww-Nﬂmiiasé. =20.
2.0-0.0..m- .00. 0.0.000-- -.-0.m-o.0“o...-- 20.0-00-2 0.02 - 000-000- 0300..€.0a. 3.0005050.
2.000... .00.. 20000.0: 00000-- . ...-00,000.00 00.000. .0000-0-a..0_00£ - - 30.00.. ...-.3 ...-2.2.2 .2
0 3.0. 3. Soc-o :8 o- :.0 a. 08cc 88¢ £200 3.00.02.25.58 : :
Mom-m“; ~00..-- 20M820-02 ammo-o.“ -.--....S 52-3. 2.32.3 20.0.90.---208000; :3
Who: FQQN bMﬁOAV ”GHQ-O *@w~ .N FQQN OMH 0.0 WWNOO ”CC—ox. 3 awn—mun." mmUCo>wwooto 3 mwcmwau
22.-......2222 -- .---2.21- -,-222-2-2-2-.---.- 222-2. 0 ..--..-.- . .. -. .2- 2 22 .. .2.-2 2208-8»..- ﬁoﬁo 33.28» 833 2.2522600 v
.03 a: N93 380 22 0 E 2.0 33 3.25 203 . ...-.25
o
F :0 am w 5 L- :0 mm Eu 0 8:35» 283252 020.0003 m
888th £825.80— N _o>3 8:2:th 53250.3.” N .964 :.0—Eugen 32050—205

 

 

@2280 98 v 38533; ”mow-bag SSE 0508880: xmB-N Agog-N 8.0 3.33% mo meEzm .ON 033-

157

s

.38: 8532c? :ouﬂuvoa .838 .296— 2: 2m Eon 5 82.5 .mng n a
622:8 538 33w 82$ 833:3 =< .8235 m 98 A295: u cocoa & 395m 3234 u ..

 

158

 

 

 

 

 

 

a; omw~2 mung-.. 2. .0.-2.0.6.? - ..5 ,.
-..... - .... _-.m.w.€.p..-m---Ewe... --.. ...; -
-. -.....- a... 2.2,. -.....o..o.-....-. ,_ . ,, ..., .

...-SN--. ...-3% - 28o - ...-soda---» -- ....» -

~V.OI - . -OwwN 22.132-22.- Vcw00221 .VMMOOM22 an». . .. . Emma—Cd.— “lye—own
...... _-__.,- ._,.._....M....,..._-..m--.......,...._.-_,._-...-..........-....,-..._-,---.....-...._... ...-...... .
. . mwd - Liam-MN.-- 2-. 22.- ..Cd-ob- 2. mambo .. ...Q. .moEEEonnm v5. .mommom . 3 umbuﬂam ...mmcaﬁﬁczE—coo
.32.: 3% :23 28d E .2223... ..5 .338

NS- ...mmei-: 2.2.0. E.~..o:é -, ¥ ...;

...-SN-.. mam--2--.- .Nm2..-o. .-mmwwd- w 0;.

N2 .. .- ..aww~.2., aﬂodm; .. - 5.0.0:; . ._ .5... -

28.2w : $3-. .. $8.0. .. $9.9m: . §

06 o GEN
L. nouns—“mm SBEEEHNoﬁq .— U > 335:5 E353: 9.89.282 «mhuomubm I

 

 

 

 

 

 

 

a £8523: “£32214 225 55825 .334” ._o>o2_-m .8 S_smom «0 $883 gm 2an

.6885 we 338252 05 08 vocaoE—Hma Mo 0002 2: 20 Eon E 3205 .mon n N.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

80.0 . 0000 00009 30.0 00008 200000 8
-..2- man nmmm 3.35 080m 0588th 088 ﬁagzmpaﬁom
.000 00 0 0 a 0 .0050 .0003 . 2
000.0 - . 0000 00.0000. Hwy “0002300888 00
2.3.5.. .2 WWW wmwm “wwch 22 2.-... ..wmwmow 0030M- EoEonm 082 BuivSnEQﬁm
80.0 -- 0000 ”0.00002 .222: .. 2m0-N-02H2
00.0 0000 2.0000 0.0mm. 00002 302088000 :
.0: 8.0 0 :.0. .0080 .0003
000.0 - - : : 0000 0:000: - ;..-mm?
00.0 500 00.002 Sam
:.0 2:000 N 00; 382 20.: mo Nx. ... howsomMauww-JL mmcuﬁ oucowa bow
0.0..N 0: N - 202 . 2 22. 2320009002.».
00 .0 00.0 N 00.0 .0050 508-0 000
000.0 2 2..-.. 2000-0.. ...-0.0.0.0002. -. . .. - -. 230m:
2 -0.“...0 - -..-00%--.. awn-00.0... ..mmmm.
mc.N 22.0 N $0 $88 208 mo 0\0 ... 02050 $084 0958 3:38:80 bow
-..mmN-2 0.002. N . .03....»- - s 2.....0N.0._.0...0....0-
200.00 00.0 N 00.0 . . N.00....000000020
NM 9% m2 MU mm ooh—cm gang—ax? “paw—5&0; m

 

 

 

 

 

 

2 can .2 .2: momoﬁogm 8.0 338% 3:033 mo 30.2050. .NN 050.0.

159

Figure l. Denison Leadership Development Model (Denison & Neale, 2001)

External Focus

 

Flexible Stable

 

160

Figure 2. Example 3-Level, 2-Way Interaction HLM Model for Hypotheses l, 2, Sa-b, 6a-b,
7a-b, and 8a-b (Taken from Hypothesis 1)

 

 

 

 

 

 

 

 

 

 

 

 

Level 3 Industry
Level 2 Leader Gender
Others’
Level 1 Development
Ratings

 

 

 

Level-1 Model
Others’ Development Ratings = no + e

Level-2 Model
1:0 = [300 + BO! (Leader Gender) + r0

Level-3 Model
Boo = 7000 + 7002 (% Women in Industry) + uoo
B02 = 7020 + You (% Women in Industry) + um

Where:

Others’ Development Ratings = example of level-l outcome variable
Leader Gender 2 example of level-2 predictor variable

% Women in Industry = example of level-3 predictor variable

no = mean ratings for Leader Gender within Industry

[300 = level-2 intercept

301: level-2 slope

7000 = grand mean

7002 = level-3 slope

7020 = level-3 intercept

'Yon = level-3 slope (cross level 2-way interaction: test of moderation hypotheses)
e = within group variance

r0 = between Leader Gender variance

uoo , um = between Industry variance

161

Figure 3. Example 2-Level, 2-Way Interaction HLM Model for Hypotheses 3, 5c, 6c, 7C,
BC, 12, and 14 (Taken from Hypothesis 3)

 

 

Proportion of

 

 

 

 

 

 

 

 

 

Level 2 Leader Gender Male Raters
Others’
Level 1 Development
Ratings

 

 

 

Level-l Model
Others’ Development Ratings = Bo + r

Level-2 Model

[30 = 700 + 70. (Leader Gender) + 702 (Proportion of Male Raters) + 703 (Leader
Gender*Pr0portion of Male Raters) + uo

Where:

Others’ Development Ratings = example of level-l outcome variable

Leader Gender = example of level-2 predictor variable

Proportion of Male Raters = example of level-2 predictor variable

Leader Gender*Proportion of Male Raters = example of level-2 interaction term
[30 = mean ratings for Leader Gender

700 2 grand mean

701, 702 = level-2 slope (cross level main effects of level-2 predictors on the level-1
outcome variable)

703 = level-2 slope (single level 2-way interaction on level-l outcome variable: test of
moderation hypotheses)

r = variance within Leader Gender

uo = variance between Leader Gender

Figure 4. Example 2-Level, 2—Way Interaction HLM Model for Hypotheses 4 and lOa—b
(Taken from Hypothesis 4)

 

 

 

 

 

 

 

 

 

Level 2 Leader Gender
Others’ Others’
Level I DCVCIOPment Effectiveness
Ratings Ratings

 

 

 

 

 

 

Level-1 Model
Others’ Effectiveness Ratings = Bo + [31 (Others’ Development Ratings) + r

Level-2 Model
BO = 'Yoo + YO] (Leader Gender) + u0
Bl = 710 + "Y“ (Leader Gender) + u]

Where:

Others’ Effectiveness Ratings = example of level-1 outcome variable

Others’ Development Ratings = example of level-1 predictor

Leader Gender (in ﬁg equation) = example of level-2 predictor of level-l intercept
Leader Gender (in B1 equation) = example of level-2 predictor of level-1 slopes
[30 = mean ratings for Leader Gender

[31: level-1 intercept (main effect of level-1 predictor on outcome variable)

700 =grand mean

701: level-2 slope

710 = level-2 intercept

711 = level-2 slope (cross level 2-way interaction: test of moderation hypotheses)
r = variance within Leader Gender

uo, u] = variance between Leader Gender

163

Figure 5. 2-Level, 3-Way Interaction HLM Model for Hypotheses 9

 

Leader Gender

 

 

Level 2

 

 

 

 

 

Level 1 Rater TYPe
(Boss or

Subordinate)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Communal _ Effectiveness V
Ratings Ratings

 

 

Agentic Ratings

 

 

 

 

 

 

 

 

 

Level-l Model

Effectiveness Ratings = [30 + B, (Rater Type) + B; (Communal Ratings) + [33 (Agentic Ratings) +
[34 (Rater Type*Communal Ratings) + [35 (Rater Type*Agentic Ratings) + r

Level-2 Model

Bo = You + Y0! (Leader Gender) + no
[31 = 710 + 7“ (Leader Gender) + u,
[32 = 'Yzo + 72, (Leader Gender) + ii;
[33 = 730 + 731 (Leader Gender) + u3
B4 = 740 + Y4! (Leader Gender) + u4
[35 = 'Yso + 75. (Leader Gender) + u5

Where:

Effectiveness Ratings = level-l outcome variable

Rater Type, Communal Ratings, Agentic Ratings = level-1 predictors

Rater Type*Communal Ratings, Rater Type*Agentic Ratings = level-1 interactions
Leader Gender = level-2 predictor variable

[30 = mean ratings for Leader Gender

[3,, B2, B3 = level-1 intercepts (main effects of level-l predictors on outcome variable)
[34, [35 = level-1 intercepts (single level 2-way interactions)

700 = grand mean

710. 'Yzo. 730. Y4o. 750 = level-2 intercepts

70., y] 1, 721, Y3! = level-2 slopes (cross level 2-way interactions)

741, 751 = level-2 slopes (cross level 3—way interactions: test of moderation hypotheses)
r = variance within Leader Gender

uo, ul, uz, u3, u4, us = variance between Leader Gender

164

 

Figure 6. Example 2 and 3-Level HLM Null Models

2-Level HLM Null Model

 

Level-l Model
Others’ Communal Ratings = Bo + r

Level-2 Model
Bo = 700 + 110

 

Where:

Others’ Communal Ratings = example of level-1 outcome variable
Bo = mean ratings for Leader Gender

“Yoo = grand mean

r: variance within Leader Gender

uo = variance between Leader Gender

 

3 -Level HLM Null Model

 

 

Level-1 Model
Others’ Communal Ratings = no + e

Level-2 Model
710 = Boo + f0

 

Level-3 Model
[300 = 'Yooo + 1100

Where:

Others’ Communal Ratings = example of level-1 outcome variable
no = mean ratings for Leader Gender within Industry

Boo = mean ratings for Industry

Yooo = grand mean

e = variance within groups

ro = variance between Leader Gender within Industry

uoo = variance between Industry

 

165

 

 

Appendix A. Denison Leadership Development Survey Demographics Questions

(Denison & Neale, 1996)

S_egg

1. Female

2. Male

3. Prefer not to respond

lf-S

Under 20

20-29

30-39

40-49

50-59

60 or over

Prefer not to respond

SOFAAPJN?‘

Education (mark highest level)
High school

Some college
Associate’s/Technical degree
Bachelor’s degree

Some graduate work
Master’s degree

Doctoral degree

Other:

Prefer not to respond

>090>19~MJ>93N~

Years With Organization
1 Less than 6 months
2 6 months to 1 year
3. l to 2 years
4. 2 to 4 years
5 4 to 6 years

6 6 to 10 years

7 10 to 15 years

8 More than 15 years
9 Prefer not to respond

166

Salary (AnntLal)

FOWSQMPP’NC‘

$25,000 or less
25,001 to 35,000
35,001 to 50,000
50,001 to 75,000
75,001 to 100,000
100,001 to 150,000
150,001 to 200,000
200,001 plus

Prefer not to respond

Ethnic Baciground

9‘1”?!“3"?

Asian

African American
Hispanic
White/Caucasian
Other

Prefer not to respond

Appendix B. Procedure for Administering and Reporting the DLDS (taken from Denison
& Neale, 2001)

Who should participate? The Denison Leadership Development Survey is designed for
leaders and managers within all types of organizations. The participant does not have to
be at any particular organizational level or hold a speciﬁc title. The important criterion is
that the participant is acting in some type of leadership role within the organization,
whether that be as CEO, manager, supervisor, team leader, or special task force leader.
The survey is designed to look at the participant's leadership skills and highlight areas of
strength, as well as areas that would beneﬁt from focused developmental efforts.

What criteria should a participant utilize when choosing respondents? The
participant should distribute the surveys to people who work closely with him/her, and
whose opinions and judgment s/he respects. The basic package for the Denison
Leadership Development Survey includes 25 Description by Others Surveys. The
participant or survey facilitator should contact us if more than 25 surveys are desired. If
the participant desires results of peers and direct reports/subordinates to be displayed
separately, s/ he must be sure that a minimum of three surveys for each group is returned
for scoring to ensure the conﬁdentiality of individual respondents.

What will the participant see in the Summary Report? Depending on how many
survey questionnaires the participant distributed and which co-workers s/he chose to
respond to the survey, the Summary Report will contain as few as two proﬁles (Self and
Combined Other), or as many as ﬁve proﬁles (Self, Combined Others, Direct Reports,
Peers and Boss.) Three surveys for each category are necessary to protect the anonymity
and conﬁdentiality of each individual respondent.

Self and Combined Other Proﬁle Sheet: The ﬁrst proﬁles in the Summary Report are
the Self and Combined Other proﬁles. At minimum the participant will always receive
these two proﬁles. They appear together on the same page to enable comparison. This
provides the participant with a visual image of how s/he perceives ability to lead, how co-
workers perceive those abilities, and where the differences and similarities lie. Self - This
survey contains participant's own responses about his/her leadership skills. The data in
the resulting proﬁle and line-item report come from a single survey. Combined Other -
This is a composite of all the responses of “Others” completing the survey. At least three
completed surveys are required to generate this proﬁle and line-item data.

Peer and Direct Report Profiles: A participant’s Summary Report may include one or
both of the following proﬁles depending on the breakdown of the surveys received. Peers
- This is a composite of co-workers who are at approximately the same level within the
organization as the participant, or other colleagues. This category can also include
customers, suppliers, etc. A minimum of three surveys is needed to generate this proﬁle
and line-item detail. Direct Reports - This is a composite of co-workers who report
directly to the participant. A minimum of three surveys is needed to generate this proﬁle
and line-item detail. As with previous proﬁles, these proﬁles allow the participant to
quickly compare skill area assessments from several vantage points within the

167

organization. These different perspectives provide an excellent foundation for individual
development. Research on 360° feedback suggests that direct reports will generally rate a
leader higher than peers.

Boss Proﬁle: The ﬁnal proﬁle included in the Summary Report is based on the survey
responses of the participant’s boss(es). This proﬁle is generally comprised of the
responses from one individual, but in a matrix with two bosses or when the participant
requests a survey from his/her boss' boss, the proﬁle may represent data from more than
one individual. This proﬁle is provided only if the boss has given permission for his/her
responses to be displayed separately. In some instances, the boss may ﬁnd it undesirable
to have data reported in a separate proﬁle. In that case, the participant will not receive a
separate boss proﬁle and the boss data will be included only in the Combined Other
proﬁle.

Line-Item Report: In addition to the graphic proﬁles, the Summary Report includes a
Line-Item Report that lists each item in the survey, groups them by index and provides
percentile scores for each item broken down by respondent (self, combined other, boss,
peers and/or direct reports). A participant's Summary Report will contain a minimum of
two breakdowns on the Line-Item Report - Self and Combined Other. The other
breakdowns are provided only when there are three or more surveys for that category.
The Line-Item Reports are color-coded to match the Denison Leadership Model (the
Involvement pages are printed in green, the Consistency pages in yellow, Adaptability
pages in blue and Mission pages in red). You will ﬁnd a "+" and "-" sign next to the
highest and lowest 10 line-item scores. The Line-Item Reports are most helpful for
developing individual action plans because they highlight scores for speciﬁc leadership
skills and practices. It is an important part of the Summary Report because it helps to
identify the speciﬁc areas of strength and weakness that are not detailed in the graphic
proﬁle.

Overview of how the system works: Each participant in a study is e-mailed an
invitation letter with their own unique 3ClickTM link. By clicking on the link, they can
deﬁne their raters and relationships (e. g, Boss, Direct Report, Peers), take the "self"
survey, track survey counts and download their own report (optional). Raters receive an
e—mail from the participant (sent out by the 3ClickTM system), which explains the study
and provides a link to the on-line survey. Reminder letters are automatically sent to only
those raters who have not completed the survey. Participants can see their survey counts
by relationship category (but they can not identify which participants have chosen to
respond or not respond). The survey data is saved in a database system. If participants or
raters disconnect from the survey (to continue later or because of internet connection
issues), their partial survey data is saved and they can ﬁnish the survey at a later time.
Electronic reports can be requested by the participants and/or, optionally, the project
administrator can have the reports produced in batch mode for all the participants.

168

Appendix C. Description of Denison Leadership Development Survey Items (taken from
Denison & Neale, 2001)

Involvement: Building human capability, ownership, and responsibility. Individual
managers who know how to create "high-involvement" strongly encourage others to be
involved and create an environment of experimentation and exploration, as well as a
sense of ownership and responsibility. Highly-involved individual managers depend on
informal, voluntary, and implicit leadership skills to move their work group or
organization forward rather than formal, explicit, bureaucratic directives. Out of this
sense of ownership grows a greater commitment to the organization, an increasing
capacity for leadership, and a sense of autonomy. Receptivity to the ideas of others
increases leadership quality and improves implementation of new ideas. The skills of the
Involvement Trait are:

0 Empowers People

0 Builds Team Orientation

o Develops Organizational Capability

Empowers People - Creates an environment where individuals have the authority,
initiative and ability to develop and manage their own work. Employees have a sense of
ownership and responsibility toward the organization. Survey items that comprise the
Empowerment skills are:

Sees that decisions are made at the lowest possible level.

Shares information so that everyone gets the information s/he needs.

Creates an environment where everyone feels that his/her effort can make a difference.
Involves everyone in shaping the plans and decisions that affect him/her.

Conveys conﬁdence in people's ability to do their job.

Ensures that the necessary resources are available to do the job.

Encourages others to take responsibility

Delegates authority so that others can do their work more effectively

Builds Team Orientation - Places value towards working cooperatively toward
common goals and knows how to use team eﬂort to get work done. Establishes a sense of
mutual accountability for the accomplishment of goals. The survey items that comprise
the Builds Team Orientation skills are:

Builds effective teams that get the job done.

Encourages effective teamwork by others.

Knows how to use a team approach to solve problems.

Knows when to use a team approach to solve problems.

Fosters teamwork within the work unit.

Knows how to design work so that it can be done by a team.

Values the contributions of the people s/he works with.

Acknowledges and celebrates team accomplishments.

169

Develops Organizational Capability -— Continually focuses on the development of
employees’ skills and knowledge to meet ongoing organizational needs. Eﬂectively uses
the diverse capabilities of the workforce. The survey items that comprise the Develops
Organizational Capability skills are:

- Builds the capabilities of employees into an important source of competitive
advantage. 0 Knows how to utilize the diversity of the work force.

0 Coaches others in the development of their skills.

0 Is sensitive and responsive to diversity issues when dealing with others.

Helps direct reports create realistic development plans.

Uses rewards and recognition to motivate good performance.

Develops direct reports/subordinates so that they are ready for promotion.

Builds employee skills so that the organization always has good "bench strength."

Trait Two: Consistency

Deﬁning the values and systems that are the basis of strong leadership. Consistency
provides a central source of integration, coordination, and control. Consistent individual
managers develop a mindset and a set of operations that create an internal system of
governance based

on consensus. They have highly committed employees, key central values, a distinct
method of doing business, a tendency to promote from within, and a clear set of “do's and
don'ts.”

Consistency produces leadership based on a shared system of beliefs, values, and
symbols that are widely understood by members of a work group or organization.
Implicit control systems based on internalized values can be a more effective means of
achieving

coordination and integration than extemal-control systems that rely on explicit rules and
regulations. The power of leadership consistency is particularly apparent when
organizational members encounter unfamiliar situations, when it enables leadership to
react in a predictable way to an unpredictable environment by emphasizing a few general,
value-based principles on which actions can be grounded. The skills of the Consistency
Trait are:

- Deﬁnes Core Values

0 Works to Reach Agreement

- Manages Coordination and Integration

Deﬁnes Core Values — Communicates and lives by a set of nonnegotiable

core values. Helps to deﬁne the work group or organization 's culture, values and ethics;
and helps employees learn to apply the organization ’s values when dealing with
customers,

stakeholders and other employees. The survey items that comprise the Deﬁnes Core
Values skills are:

0 Does the right thing even when it is not popular.

0 Practices what s/he preaches.

- Has an ethical code that guides his/her behavior.

0 Helps deﬁne the organization's culture, values and ethical standards.

170

Helps employees learn to apply the organization's values when dealing with customers,
stakeholders and other employees.

0 Lives up to promises and commitments.

- Has earned the conﬁdence and trust of others.

Clearly articulates a set of fundamental beliefs that are not negotiable.

Works to Reach Agreement — Helps to reconcile diﬂerences when

they occur by actively promoting constructive discussion of conﬂicting ideas,
incorporating diverse points of view into decisions, and working toward win-win
solutions. The survey items that comprise the Works to Reach Agreement skills are:
Helps people to reach consensus, even on difﬁcult issues

Works to ﬁnd alternatives that will beneﬁt all when confronted with a disagreement.
Helps people in his/her organization be effective at reaching agreement on key issues.
Promotes constructive discussion among people with conﬂicting ideas.

Is willing to compromise when necessary in order to reach agreement.

Works toward win/win solutions when disagreements occur.

Reconciles differences by seeking to clarify and understand other's points of view.

Manages Coordination and Integration — Ensures that different functions or units of
the work group or organization are able to work together well to achieve common goals.
Establishes necessary contacts and coordinates resources in other groups to prevent
organizational boundaries from interfering with getting work done. The survey items that
comprise the Coordination and Integration skills are:

0 Works hard to foster the alignment of goals across all functional areas.

Builds coordination across departmental boundaries.

Uses informal networks to get things done.

Builds relationships with key people in other ﬁrnctions and levels.

Helps create an environment that facilitates coordination of projects across ﬁrnctional
units.

Makes certain that things do not fall between the cracks.

o Builds support for ideas through contact with other departments.

- Establishes mechanisms that facilitate effective cross-functional communication.

Trait Three: Adaptability - Translating the demands of the organizational environment
into action. Successful individual managers hold a system of norms and beliefs that
support his/her capacity to receive and interpret signals from the environment, and to
translate them into internal behavioral changes that increase the work group or
organization's chances for survival, growth, and development. Three aspects of
adaptability influence an individual manager's effectiveness. First is the ability to
perceive and respond to the external environment. Successﬁil individual managers are
very focused on their customers and their competitors. Second is the ability to respond to
internal customers, regardless of level, department, or ﬁinction. Third is the capacity to
restructure and reinstitutionalize a set of behaviors and processes that allow the
organization and its employees to adapt. Without this ability to implement adaptive
response, an organization cannot be effective. The skills of the Adaptability Trait are:

171

- Creates Change
0 Promotes Organizational Learning
0 Emphasizes Customer Focus

Creates Change — Knows the organizational environment and quickly reacts to current
trends, and anticipates future changes. Continually creates adaptive and innovative
ways to meet changing needs. The survey items that comprise the Creates Change skills
are:

Continuously looks for new and better ways to do work.

Encourages creative thinking.

Challenges the way that things have always been done and looks for a better way.
Champions change that goes beyond the scope of his/her job.

Challenges organizational practices that are nonproductive.

Foresees problems before they arise.

Serves as a model that creates change in other parts of the organization.

Generates innovative ideas and solutions to problems.

Emphasizes Customer Focus — Driven to clearly understand the present and future
needs of the customer, seeks ongoing input from the customer, continuously strives to
improve customer service, and ensures that all employees are driven by a concern to
satisfy the customer. The survey items that comprise the Emphasizes Customer Focus
skills are:

Encourages direct contact with customers.

Responds quickly and effectively to customer feedback.

Ensures that employees have a deep understanding of customer wants and needs.
Uses customer comments and recommendations to change organizational practices.
Actively seeks feedback from customers.

Continuously tries to improve service to customers.

Incorporates customer input into the planning process.

Recognizes the need to respond quickly to customer concerns.

Promotes Organizational Learning - Encourages innovation, risk taking and
continuous improvement. Sees mistakes as opportunities for gaining knowledge and
developing capabilities. The survey items that comprise the Promotes Organizational
Learning skills are:

Deals constructively with failures and mistakes.

Views failure as an opportunity for learning and improvement.

Creates a working environment in which learning is an important objective.
Openly accepts criticism without being defensive.

Works well under conditions of ambiguity and uncertainty.

Knows the strengths and weaknesses of the competition.

Encourages others to learn about the best practices in the industry.

Helps others to understand the big picture

Trait Four: Mission

Deﬁning a meaningful long-term direction for the organization The individual manager's
mission provides purpose and meaning by deﬁning a social role and external goals for
his/her functional area or unit. It provides a clear direction and goals that serve to deﬁne
an appropriate course of action for the individual manager and his/her employees. The
individual manager is able to align the mission and goals for his/her functional area or
unit to the mission and goals of the organization. A sense of mission allows an individual
manager to inspire, to direct activities, and to formulate strategy by envisioning a desired
future state. Being able to translate his/her mission into action contributes to both short
and long-term commitment to the organization. Success is more likely when individual
managers and organizations are goal directed. The skills of the Mission Trait are:

0 Deﬁnes Strategic Direction and Intent

0 Deﬁnes Goals and Objectives

0 Creates Shared Vision

Deﬁnes Strategic Direction — Communicates the organization ’3 overall strategies so
that everyone can see the relationship between their work and the accomplishment of the
work group or organization ’s goals. Eﬂectively implements short and long—term
strategies to meet organizational needs. The survey items that comprise the Deﬁnes
Strategic Direction and Intent skills are:

Provides employees with a clear mission that gives direction to their work.
Implements strategies by developing clear goals, objectives and tactics.

Focuses on long-term strategies, rather than quick ﬁx "band- aid" solutions.
Effectively allocates resources in line with strategic priorities. ’

Helps deﬁne strategies and tactics that keep his/her organization competitive.

Has a clear strategy for the future of his/her own part of the organization.

Is able to meet short-term demands without losing sight of the long-term strategy.
Communicates a clear and compelling rationale for the business strategy.

Deﬁnes Goals and Objectives —— Encourages high employee accountability in setting and
accomplishing organizational goals. Communicates a clear set of goals and objectives
that can be linked to the mission, vision and strategy of the work group or organization.
The survey items that comprise the Deﬁnes Goals and Objectives skills are:

Sets clear goals that are ambitious, but realistic.

Holds individuals and teams accountable for achieving goals and objectives.
Provides clear directions and priorities for employees.

Establishes high standards of performance.

Involves employees in the goal-setting process.

Tracks progress against stated goals.

Effectively communicates the goals and objectives of the organization.

0 Aligns goals and objectives with the strategy and vision.

Creates Shared Vision — Helps create a shared view of a desired ﬁxture state for his/her
organizational unit. Inspires others with this vision, translates it into everyday activities
and engages others to ensure buy-in and commitment. The survey items that compromise
the Creates Shared Vision skills are: .
o Helps create a shared vision of what the organization will be like in the ﬁiture.
Communicates the organizational vision to his/her employees.
0 Uses the vision to create excitement and motivation for employees.
Realizes short-term goals without compromising long-term vision.
0 Organizes work so that everyone sees the connection between the vision and his/her
day-to-day activities.
Translates the vision into reality in a way that helps guide individual action.
0 Inspires others with his/her vision of the future.
- Engages others in ways that ensure buy-in and commitment.

174

Appendix D. Communal and Agentic Development Items

Communal Items (32 total)

I.
2.
3.

4.
6.
7.

10.
11.
12.
I3.
14.
15.
16.
19.
20.
23.
27.
29.
33.
34.

35.

36.
37.
38.
39.
40.
42.
48.
59.
61.
63.
85.

Secs that decisions are made at the lowest possible level.

Shares information so that everyone gets the information s/he needs.

Creates an environment where everyone feels that his/her effort can make a
difference.

Involves everyone in shaping the plans and decisions that affect them.

Conveys conﬁdence in people’s competence to do their job.

Encourages others to take responsibility.

Encourages effective teamwork by others.

Knows how to use a team approach to solve problems.

Knows when to use a team approach to solve problems.

Fosters teamwork within the work unit.

Knows how to design work so that it can be done by a team.

Values the contributions of the people s/he works with.

Acknowledges and celebrates team accomplishments.

Coaches others in the development of their skills.

Is sensitive and responsive to diversity issues when dealing with others.
Develops his/her own people so that they are ready for promotion.

Has an ethical code that guides his/her behavior. '

Helps employees learn to apply the organization’s values when dealing with others.
Helps people to reach consensus, even on difﬁcult issues.

Works to ﬁnd alternatives that will beneﬁt all when confronted with a
disagreement.

Helps people in his/her organization be effective at reaching agreement on key
issues. '

Incorporates diverse points of view when making decisions.

Promotes constructive discussion among people with conﬂicting ideas.

Is willing to compromise when necessary in order to reach agreement.

Works toward win/win solutions when disagreements occur.

Reconciles differences by seeking to clarify and understand other’s points of view.
Builds coordination across departmental boundaries.

Establishes mechanisms that facilitate effective cross-functional communication.
Ensures that employees have a deep understanding of customer wants and needs.
Actively seeks feedback from customers.

Incorporates customer input into the planning process.

Involves employees in the goal-setting process so goals and objectives are
understood and shared.

175

Appendix D (continued). Communal and Agentic Development Items

Agentic Items (22 total)

32.
49.
51.
52.
53.
55.
70.
73.

74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
86.
88.
94.

Clearly articulates a set of ﬁindamental beliefs that are not negotiable.
Continuously looks for new and better ways to do work.

Challenges the way that things have always been done and looks for a better way.
Champions change that goes beyond the sc0pe of his/her job.
Challengesorganizational practices that are nonproductive.

Serves as a model that creates change in other parts of the organization.

Knows the strengths and weaknesses of the competition.

Provides employees with a clear mission that gives meaning and direction to their
work.

Implements strategies by developing clear goals, objectives, and tactics.

Focuses on long-term strategies, rather than quick ﬁx “band-aid” solutions.
Effectively allocates resources in line with strategic priorities.

Helps deﬁne strategies and tactics that keep his/her organization competitive.
Has a clear strategy for the ﬁrture of his/her own part of the organization.

Is able to meet short-term demands without losing sight of the long-term strategy.
Communicates a clear and compelling rationale for the business strategy.

Sets clear goals that are ambitious, but realistic.

Holds individuals and teams accountable for achieving goals and objectives.
Provides clear directions and priorities for employees.

Establishes high standards of performance.

Tracks progress against stated goals.

Aligns goals and objectives with the strategy and vision.

Translates the vision into reality in a way that helps guide individual action.

176

Appendix E. Denison Leadership Development Survey Effectiveness Items (Denison &
Neale, 1996)

1.

2.

Overall, this individual is a highly effective leader.

This individual’s leadership style serves as a role model for others in the organization.
This individual has great potential as a ﬁiture leader in our organization.

Overall, this individual is one of the most capable leaders in our organization.

This individual develops high quality relationships with internal and external
customers.

This individual and his/her organization are consistently high performers.

This individual is capable of leading the organization through ﬁiture changes and
transitions.

. 177

 

 

Appendix F. Denison Leadership Development Survey Organizational Level Options
(Denison & Neale, 1996)

Organization Level

Non-management

Line management (supervising non-management personnel)
Middle management (managing managers)

Senior Management

Executive/ Senior Vice President

CEO/President

Owner

Prefer not to respond

”S9MPWNE

I78

 

Appendix G. Denison Leadership Development Survey Functional Area Options and
Corresponding DOC Codes (Denison & Neale, 1996)

1.

10.

11.

Finance and Accounting
(DOC Code: Financial Managers; 52.1% women)

Engineering
(DOC Code: Engineers, 10.4% women)

. Manufacturing and Production

(DOC Code: Production Occupations; 24.6% women)

Research and Development -
(DOC Code: Scientists; 34.3% women)

Sales and Marketing
(DOC Code: Sales Occupations; 49.4% women)

Purchasing
(DOC Code: Purchasing Managers; 42.6% women)

Human Resources
(DOC Code: Personnel and Labor Relations Managers; 68.2% women)

Administration
(DOC Code: Executive, Administrative, and Managerial; 46% women)

Support Staff
(DOC Code: Administrative Support Occupations, Including Clerical; 78.7% women)

Professional Staff
(DOC Code: Professional Speciality; 53.7% women)

Prefer not to respond

179