Does a Female Manager Make a Difference?
M. Lee Williams, Ph.D.
Professor in the Department of Speech Communication
Southwest Texas State University, San Marcos, TX 78666
Office Telephone: 512-245-3127
Victoria Nevin Locke
Research Associate at The Psychological Corporation, San Antonio, TX
Paper Presented at the Institute for Behavioral and Applied Management (IBAM) conference,
Annapolis, MD, November 4-6, 1999. Best Paper in Division I, Human Resources Management.
Does a Female Manager Make a Difference?
Research investigating the management styles of male and female supervisors is mixed. In
an effort to clarify this research, the present study examined the interaction effect of gender of
supervisor and gender of subordinate on perceived mentoring. Results revealed the least amount of
mentoring occurred between female supervisors and female subordinates, but the greatest mentoring
took place between female supervisors and male subordinates. Results were explained in terms of
managerial experience and the double-bind of female managers. These findings challenge the
traditional advice that female subordinates should seek female supervisors.
Does a Female Manager Make a Difference?
Mentoring, whether it be formal or informal, is a common practice in organizations today. In
this dyadic relationship an older, more experienced member of the organization helps the younger
employee learn to navigate in the workplace (Kram, 1983). Even though considerable research has
investigated mentoring, very few studies have focused on the amount and type of mentoring
provided by supervisors (Burke, McKenna & McKeen, 1991). In addition, virtually no research has
explored the nature of supervisor mentoring in same- and cross-gender supervisor-subordinate
relationships. Therefore, the purpose of this paper is to explore research investigating supervisor
mentoring as well as to test hypotheses addressing the kinds of mentoring provided by male and
female supervisors to male and female subordinates.
This research is of value for three primary reasons. First, as indicated above, research has
not considered the relationship between supervisor mentoring and gender combinations. This is an
unexplored area of inquiry. Second, the results of this investigation should expand the already rich
body of literature on mentoring, supervisor-subordinate relationships, and gender. While
understanding has been gained in these separate areas, efforts to integrate the varying theoretical
frameworks has been limited. New insights can be gained from synthesizing this literature and
providing empirical testing. Finally, this study is of value for a very practical reason. Since women
now comprise over 46% of the workforce (Bureau of Labor Statistics, 1998), it is important that we
better understand the dynamics of gender, managerial styles, and employee perceptions. While some
men find it difficult to work under the supervision of a female (Fairhurst, 1993), some women are
uncomfortable supervising men. It is important to discover if these cross-gender situations impact the
type of mentoring provided. As the work place becomes more diverse and complex, organizations
need to anticipate problem areas and develop strategies for managing these complexities.
Review of Literature
The unique characteristics of mentoring clearly establish the relationship as interpersonal
(Kalbfleisch & Davies, 1993). The mentor guides, counsels, and encourages the younger person,
known as a protégé. Within the mentoring relationship, there are stages of coming together and
coming apart, and feelings of intimacy may occur (Kram, 1983). Mentoring combines aspects of a
work relationship as well as a personal relationship.
The mentor provides both career development and psychosocial support (Kram, 1983). The
career development function includes sponsorship, visibility to others, protection, and offers of
challenging assignments. The psychosocial function enhances a protégé’s feeling of competence and
sense of identity in the organization. From these two types of support, the protégé learns increased
self confidence, ways of dealing with people, self insight, ways to approach problems, and an
increased understanding of the organization (Burke, 1984). Furthermore, research indicates protégés
enjoy the additional benefits of having greater job satisfaction (Mobley, Jaret, Marsh & Lim, 1994;
Riley & Wrench, 1985), being promoted more often (Whitely, Dougherty, & Dreher, 1991; Dreher
& Ash, 1990; Henderson, 1985), attaining an executive position at a younger age (Henderson,
1985), and receiving greater total compensation, including salary, bonuses, and benefits (Whitely,
Dougherty, & Dreher, 1991; Dreher & Ash, 1990; Henderson, 1985).
Mentors can also benefit from the relationship. They find internal satisfaction from knowing
they have made a positive impact on a younger person. Mentors also receive recognition within the
organization from colleagues for successfully developing younger talent (Kram & Isabella, 1985).
On the downside, however, mentors may feel rivalrous and threatened by a protégé’s growth and
development (Kram, 1983). Lack of recognition can also be a problem. In a survey of public
managers, 47% of those that thought of themselves as having been a mentor were not identified as
such by their protégés (Henderson, 1985).
Research indicates that mentoring tends to occur in naturally hierarchical relationships
(Burke, 1984). Burke, McKeen, & McKenna (1993) found that approximately one-half of
mentoring relationships occur within the protégé’s hierarchical line of responsibility. Tepper (1995)
showed that while 30% of employees in his study had a mentoring relationship with their supervisor,
18.5% of those were informal, and only 11.5% were formal. Findings also indicate that managers
offer both protégés and nonprotégé subordinates career support (i.e., provide feedback, career
planning, teaching, sponsorship); however, managers offer protégés more psychosocial support
(Burke, McKenna & McKeen, 1991). Protégés are perceived as being more promotable and more
similar to the mentor. Managers are also more likely to confide in a protégé than a typical
The quality of the supervisor-subordinate relationship is of vital importance to the employee
as well as the organization. It is the primary relationship articulated by the organization (Shockley-
Zalabak, 1988), and structurally it is the most important communication link in the organization
(Downs, Clampitt, & Pfeiffer, 1988). Group meetings and top executives are desirable sources of
information, but employees identify the immediate supervisor as the most preferred source of
information (Foehrenback & Goldfarb, 1990). In addition, employees indicate the immediate
supervisor is the primary source for actually receiving information (Foehrenback & Goldfarb, 1990).
Major reviews of supervisor-subordinate communication have helped synthesize the research
findings for this vital dimension of organizational life. Jablin (1979) classified the literature into nine
categories. He explored areas such as interaction patterns, openness in communication, upward
distortion of information, the gap in understanding between supervisors and subordinates,
supervisor feedback, and the communication qualities of effective versus ineffective supervisors.
Updates to this research were provided by Jablin (1985) and Dansereau and Markham (1987). The
most recent update by Jablin and Krone (1994) added the component of social support.
Social support is the communication between people that helps, comforts, cares for, and aids
one or both persons (Albrecht & Adelman, 1987). This unique form of human contact reduces
uncertainty, provides a sense of personal control, and creates a stronger bond between the
participants. Social support can serve as a buffer to shield the negative consequences of stress
brought on by organizational factors such as role ambiguity, work overload, and job uncertainty
(Cohen & Wills, 1985). The most consistent finding in the social support research investigating
organizations is that the immediate supervisor is the person most likely to provide this support and
thus reduce employee stress (Anderson, 1991; Fenlason & Beehr, 1994; Fisher, 1985; Jones-
Johnson & Johnson, 1992; Ray, 1987; Ray & Miller, 1991).
Gender and Management
Since the 1960s and 1970s when women began to enter the workforce en masse, researchers
have sought to determine whether male and female managers have different styles or exhibit
alternative patterns of communication. The results of that research have produced a variety of
Powell (1990) reviewed the literature on gender differences in managerial behavior and
concluded there were no differences in task-oriented behavior, people-oriented behavior, and
subordinates’ responses to actual managers. Eagly and Johnson (1990) reported that males and
females were rated as equally effective by their subordinates. These findings would lead one to
conclude that either gender has a trivial, almost inconsequential influence on behavior (Daniels,
Spiker, & Papa, 1997), or female managers over time reject the feminine stereotype and adapt to the
male-dominated corporate culture (Berryman-Fink, 1997).
Other research, however, identifies gender differences. Witherspoon (1997) reviewed the
leadership literature and noted several key differences between male and female managers. Men
assume more task roles, give more opinions, are argumentative, and do not disclose personal
information. Men tend to take over decision-making discussions, and criticize the opinions and ideas
of other people. Conversely, women assume nurturing roles, interrupt for clarification, are more
disclosive about information, and more supportive of other speakers. Women also try to avoid
conflict by seeking compromises and talking through problems (Witherspoon, 1997).
Some contend that these differences between male and female managers create a female
advantage in today’s organizations (Helgesen, 1990). Because young girls are socialized to be
cooperative, understanding, supportive, interpersonally sensitive, and flexible, they are more
inclined to develop different managerial styles when they grow up and assume leadership positions
in organizations (Helgesen, 1990). Furthermore, these traditional “feminine qualities” are more in
line with contemporary organizations which value sharing information, collective decision-making,
developing relationships, empowering others, and resolving conflict in nonconfrontational ways.
Some research investigating gender differences in organizational conflict has found that
women were less competitive, more accommodating, more willing to share power, and more willing
to discuss divergent viewpoints than men (Burrell, Buzzanell, and McMillan (1992). Korabik, Baril,
and Watson (1993) noted gender differences in conflict management styles were present only in less
experienced female managers. These women rated themselves as more obliging and compromising
than did men. There were, however, no differences between male and female managers with
experience. These findings seem to indicate that less experienced female managers exhibit more
traditional feminine behavior acquired from socialization earlier in life, while more experienced
female managers temper these feminine attributes and adopt more masculine qualities.
Another line of research investigating gender differences in management takes the position
that men and women can learn from each other’s unique leadership qualities, and the most desirable
type of management is one which is androgynous (Maier, 1992). The most effective management
style is the one that captures the best of both. According to this approach, men can learn from
women, and women can learn from men.
Some studies have sought to integrate various elements of research investigating mentoring,
supervisor-subordinate communication, and gender differences in managers. While research from
the 1970s and early 1980s indicated men were more likely to be mentored than women (Cook, 1979;
Hall & Sandler, 1983), more recent research has shown women are just as likely to be mentored as
men (Mobley, Jaret, Marsh, & Lim, 1994; Ragins & Cotton, 1991). However, women perceive
more barriers in acquiring a mentor (Ragins & Cotton, 1991). These barriers include restricted
access to potential mentors (i.e., senior executives), the perception that mentors were unwilling to
enter into a relationship with them, and the feeling that other people would disapprove of the
relationship, or misconstrue it as being sexual rather than professional (Ragins & Cotton, 1991).
Some research shows that mentors provide more psychosocial functions to women than men (Burke,
McKeen & McKenna, 1993), women managers mentor more women than men (Ragins & Scandura,
1994), and women prefer female mentors (Kram, 1985). In addition, research indicates women
often feel more isolated in the organization, and feel like they receive less mentoring from their
immediate supervisor (Goh, 1991).
One limitation of much of the supervisor-subordinate research is that it has focused on gender
primarily as a main effect. Studies have looked at the impact of the gender of the manager or the
gender of the subordinate, but rarely have they investigated the same- or cross-gender combinations
(i.e., an interaction effect). In particular, the nature of the relationship between a male subordinate
with a female supervisor is not well understood. The main reason for limited research in this area is
because fewer women are in upper-levels of management, especially in male-dominated industries
Therefore, in an effort to better understand the role of gender in supervisor-subordinate
mentoring relationships, the present study investigated gender as an interacting variable. The
research question guiding the investigation was:
RQ: Do female subordinates with female supervisors receive more mentoring than other gender
combinations of supervisors and subordinates?
Since prior research indicates women tend to be more relationship-oriented (Fairhurst, 1993),
and since mentors provide more psychosocial functions to women than men (Burke, McKeen &
McKenna, 1993) it is expected that female managers will have greater empathy with female
subordinates and thus provide more mentoring and social support to female subordinates than male
subordinates. The following hypotheses are presented for testing:
H1: Female subordinates with female supervisors will receive significantly more mentoring
than female subordinates with male supervisors.
H2: Female subordinates with female supervisors will receive significantly more mentoring
than male subordinates with female supervisors.
Supervisor mentoring was measured using a variation of the Mentoring and Communication
Support Scale developed by Hill, Bahniuk, Dobos, and Rouner (1989). This instrument measures
four factors: career mentoring, task support, coaching, and social support. The measure has a total
of 15 items with Likert response options ranging from strongly agree (5) to strongly disagree (1).
Career mentoring contains four items that identify the presence of a personal relationship (e.g., My
immediate supervisor has shown a parental-like interest in me and my career). Task support contains
four items that represent a reciprocal, collaborative relationship focused on sharing and exchanging
work assignments and ideas (e.g., I work jointly on major projects or cases with my immediate
supervisor). Coaching contains three items that focus on teaching the rules, goals, and politics of the
organization (e.g., My immediate supervisor has coached me about office politics). Social support
contains four items that measure a reciprocal friendship focused on sharing and exchanging personal
problems and confidences (e.g., My immediate supervisor and I are friends as well as coworkers).
Reliability of the scale ranges from .75 to .89 for managers and from .76 to .85 for the
general population (Hill, Bahniuk, & Dobos, 1989). Reliability of the sub-factors was confirmed in
a separate study. Among a sample of managers, the four factors of task support, coaching, career
support and social support had alphas of .89, .85, .75, and .75 respectively (Dobos,
Bahniuk, & Hill, 1991). Criterion validity is evident, as the four factors correlate positively with
satisfaction, promotions, and fast-track mobility (Hill, Bahniuk, & Dobos, 1989).
The measure used in this analysis was slightly modified from the original version. Items
which originally focused on coworker collegial support were rephrased to indicate supervisory
collegial support. Other items which asked if someone of a higher rank had provided support were
also rephrased to indicate supervisory support. With these modifications, all 15 items evaluated
subordinates’ perceptions of supervisor mentoring, not mentoring in general.
The questionnaire employed in the present study also asked respondents to complete an
open-ended item. Subjects recalled a communication experience when support from his/her
immediate supervisor was especially important. Relating communication experiences allowed
subjects to reveal information that otherwise would be missed by a Likert scale. This rich, qualitative
data also is invaluable when interpreting data from questionnaires (Downs, 1988).
Questionnaires were administered to employees of two metropolitan daily newspapers in the
southwestern United States. The two newspapers are in the same state, approximately 80 miles
apart, with a similar readership. These organizations were selected because of the nature of their
business and employees. They are information-intensive, deadline-driven organizations, with many
upwardly mobile employees. Many entry-level jobs require a college degree. Both organizations
have relatively equal numbers of male and female employees, as well as male and female
Prior to data collection, a liaison in the human resources departments of the two organizations
informed employees they would be asked to participate in a study. Anonymity and voluntary
participation were emphasized. Employees completed the supervisor mentoring questionnaire and
answered other demographic questions. Upon completion, questionnaires were placed in a drop box
located in the employees’ work area.
The revised version of the Mentoring and Communication Support Scale used in the present
study was submitted to a confirmatory factor analysis with varimax rotation. Four guidelines for
selection of factors were used. Each factor needed to have an eigenvalue of 1.0 or greater, each item
needed to load at least .60 on one factor (but no higher than .40 on any other factor), and at least
three items needed to load on a factor. Finally, each factor needed to have a reliability of .70 or
greater. These criteria are typically used in factor analysis research (Smith, 1988).
Three factors with an eigenvalue greater than 1.0 were produced. The coaching factor
contained three items and explained 27% of the variance (Cronbach alpha = .85). Collegial social
contained four items and explained 20.2% of the variance (Cronbach alpha = .82). Both factors were
identical to those on the original scale by Hill, Bahniuk and Dobos (1989). A third factor, titled task
mentoring, contained six task mentoring and career mentoring items from the original scale by Hill,
Bahniuk and Dobos (1989). This factor explained 19.1% of the variance (Cronbach alpha = .88).
Two items which were included in the career mentoring factor of the original scale did not load at .60
or greater on any factor; therefore, these items were eliminated from the analysis.
There are several explanations why the confirmatory factor analysis produced three factors
instead of the four in the original scale. The primary explanation is that the scale was revised. In the
original scale, the career mentoring items were worded to state that "someone higher up" had