|Home | About | Journals | Submit | Contact Us | Français|
Concurrent with the evolving role of the department chair in academic medicine is the entry of women physicians into chair positions. Because implicit biases that stereotypically masculine behaviors are required for effective leadership remain strong, examining faculty members' perceptions of their chair's leadership in medical school departments with women chairs can provide insight into the views of women leaders in academic medicine and the complex ways in which gender may impact these chairs' leadership style and actions.
We conducted semistructured interviews with 13 male and 15 female faculty members representing all faculty tracks in three clinical departments chaired by women. Inductive, qualitative analysis of the subsequent text allowed themes to emerge across interviews.
Four themes emerged regarding departmental leadership. One dealt with the leadership of the previous chair. The other three described the current chair's characteristics (tough, direct, and transparent), her use of communal actions to help support and mentor her faculty, and her ability to build power through consensus. Because all three chairs were early in their tenure, a wait and see attitude was frequently expressed. Faculty generally viewed having a woman chair as an indication of positive change, with potential individual and institutional advantages.
This exploratory study suggests that the culture of academic medicine has moved beyond questioning women physicians' competence to lead once they are in top organizational leadership positions. The findings are also consonant with experimental research indicating that women leaders are most successful when they pair stereotypic male (agentic) behaviors with stereotypic female (communal) behaviors. All three chairs exhibited features of a transformational leadership style and characteristics deemed essential for effective leadership in academic medicine.
The skills required of department chairs in academic medicine have evolved tremendously.1–5 In the past, the chair was an honorary position given to a talented, revered physician who was strong in the three missions of clinical practice, education, and research. Today, the chair must be a department head with complex leadership skills, including protean communication abilities, an understanding of funds flow from various sources, the ability to inspire multiple stakeholders toward a collective vision, and the flexibility to adapt to the changing and frequently unpredictable environment of academic medicine.1–5 Concurrent with the evolving role of the department chair is the entry for the first time of women physicians in substantial numbers into chair positions.6,7 According to the Association of American Medical Colleges (AAMC), in 2007–2008, women accounted for 34% of all medical faculty and 17% of full professors. Although women have comprised >30% of medical students for nearly three decades 8 and much has been written on the disproportionate lack of advancement of women physicians into senior and leadership positions,9–12 the AAMC data indicate that women are achieving leadership positions in the top echelons of academic medicine in unprecedented numbers. Of all departments in academic medical centers (AMCs) 12% are now led by women (vs. 6% in 1998) and only 6 of 126 AMCs remain with no women chairs. 8
Implicit biases that stereotypically masculine behaviors are required for effective leadership remain strong.13–15 Examining faculty members' perceptions of their chair's leadership in medical school departments chaired by women can provide insight into the evolving views of women leaders in academic medicine and the complex ways in which gender may impact these chairs' leadership style and actions. Qualitative methods are ideally suited to probe this issue because they allow exploration of real life behavior. 16 These methods enable research participants to speak for themselves, with the underlying belief that perceived reality is constructed by the social, historical, and individual contexts in which it is lived. 17 This stands in contrast to quantitative research, which generally works from the assumption that there is an absolute true reality that can be discovered. We chose to use the qualitative method of grounded theory, which is widely used in the study of social phenomena to explain processes or generate theories that are derived from (i.e., grounded in) systematic analysis of the rich data collected by interviewing participants.18–20 Unlike quantitative research where the goal is to select participants randomly, participants in a study using grounded theory methods are chosen because the investigators believe they will represent the range of beliefs and experiences relevant to the issue being studied. This is referred to as “theoretical” or “purposive” sampling.21,22 The unit of analysis in grounded theory is the incident or condition and not the person or patient as in quantitative research. In this study, we interviewed faculty members in three clinical departments with women chairs. We explored issues of leadership and interpreted the findings in the context of socialized gender roles, research on leadership, and the prevailing views of requisite skills for chairs in the current academic medicine environment.
Participants were faculty members who had served under at least one previous chair at the University of Wisconsin School of Medicine and Public Health and who were currently in one of the three clinical departments (out of 16) with a woman chair. The study was approved by the Institutional Review Board, and all participants gave written informed consent. Using a theoretical sampling strategy, the senior author (M.C.) identified cases expected to have the desired range of experiences from lists of all faculty members in targeted departments. She sent email invitations for an interview to those selected, including physicians at all ranks (assistant, associate, and full professor) and in three faculty tracks: tenure (clinician-researcher), clinical health sciences (clinician-educator), and clinical (clinician-practitioner). Invitations stated that the study was investigating leadership in clinical departments, without mention of gender or which departments were being studied. Of the 65 faculty members contacted, 30 did not respond after two follow-up email invitations. Of the 35 who responded, 1 declined, 1 was ineligible (service under only current chair), 33 were scheduled for interview, 29 were interviewed, and 1 withdrew after interview. Three departments were represented in the sample with the following responses: department A, 11 (5 male, 6 female); department B, 10 (5 male, 5 female); and department C, 7 (3 male, 4 female) faculty participants. Table 1 lists the interview questions. The first author (L.G.) performed all interviews at sites convenient to participants during the summer of 2008; the interview lengths ranged from 15 to 35 minutes. L.G. digitally audiorecorded, transcribed verbatim, and removed all identifying information from the interviews. Pseudonyms were assigned to the interviews for tracking. Participants reviewed and verified transcripts (sometimes referred to as “member checking”). 23
We followed Strauss and Corbin's steps 20 for sequentially coding data and grouping subsequent codes. First, each interview transcript was examined line by line or sentence identifying initial concepts (codes) in what is referred to as “open coding.” Next, codes that were viewed as conceptually linked were grouped together (axial coding). Finally, these concepts were integrated and synthesized into categories (selective or theoretical coding) to identify the core categories to which other categories relate. 20 We adhered to the central principle of data analysis in grounded theory research, which is an iterative cycle of data collection, analysis, and comparison with previously collected data looking for coherence until unifying and recurrent themes emerge from the data itself and no new codes occur.19,20
NVivo, a computer software program for managing qualitative analysis, facilitated data organization, coding, and retrieval. 24 L.G. defined the set of initial codes in the open coding process (Fig. 1). The second author (C.I.) reviewed and verified these initial codes in the first five transcripts. Examples from this open coding were “action,” “being decisive,” “change,” “closed door,” and “difference.” Axial codes were developed from the initial open codes in weekly meetings of the research team, linking like concepts around the axis of a category. Examples of the axial codes were “damage,” “direct,” “support through recognition of value,” ‘actions for faculty,” and “power in consensus.” From selective coding of the axial codes, core categories were identified and synthesized into final themes and a theoretical model grounded in the data (i.e., a grounded theory) (Fig. 2). C.I. verified the final selective coding by independently categorizing the data derived from L.G.'s initial codes. The intercoder agreement ranged between 82% and 100%, with an average of 89%. As commonly occurs in qualitative analysis of text, double coding could occur (e.g., a statement could contribute conceptually to two categories of codes).
All codes were continually compared within their conceptual categories as new data were collected. Discrepancies in coding or categorization were resolved by discussion until consensus was reached. Data saturation 25 in which no new data codes, categories, or themes emerged occurred after 19 interviews. Validity of the results in grounded theory is judged by the degree to which they present a consistent and coherent story line. 19 In discussing qualitative research data, Koro-Ljungberg 26 prefers the term “validation” to “validity.” In this study, validation of the emerging themes was enhanced by the presence on the research team of an experienced qualitative researcher with a background in educational leadership (C.I.) and an investigator (M.C.) with extensive knowledge of the extant research on implicit gender stereotypes. Because coding is not preliminary to analysis but an integral part of the interpretation of the social condition under investigation, the requisite assumptions for the application of statistical tests (confirmatory analyses) in grounded theory (exploratory analyses) are not met.
Open coding identified 34 initial codes, which were integrated into 8 axial codes, each containing 2–5 codes. These were ultimately synthesized into four major themes that comprehensively defined all codes regarding departmental leadership (Tables 2A, ,2B,2B, ,2C,2C, and and2D).2D). One dealt with the leadership of the previous chair. The other three described the current chair's characteristics (tough, direct, and transparent), her use of communal actions to help support and mentor her faculty, and her ability to build power through consensus.
Because each chair began within the previous 36 months (6–36 months), the prior environment was frequently mentioned in the interviews (Table 2A). In addition to noting frequent absences, one faculty member explained, “His office was a corner office far on the edge of one of the hallways which almost no one ever went down…it was literally as far away from the center of activity as possible.” Some embraced this absence as “complete autonomy—as complete as you can [get].” Other faculty members found that “different silos just popped up where you would have somebody who was powerful, influential, or just had vision…[who] would then build a program within the department that was strong…but never with any overarching goal of the department.” Faculty described departmental processes as “paddling with our oars not quite all the way in” and lacking follow-through. Another explained, “Number one is to…pick up from all of the neglect of 10 years prior to her arrival.” Although some experienced success with autonomy, others described “neglect” and “damage” in conjunction with “lost opportunity” and “lack of recognition.”
Participants reluctantly criticized the “weighed down” and “tired” previous chairs but reported, “We all liked him—he just wasn't ever here.” They also described the former chair as not being “open to other perspectives.” One woman reported, “I remember raising my hand and saying…can't we just keep [program] anyway? And he just looked at me and said, ‘No. Next question.” Faculty members thought that the former chair “really wanted some individuals to succeed” but with others “distanced himself” when there was conflict.
Some interviewees reflected gratitude to the current chair for taking on the position: “God bless her…. I wouldn't want to do that, but she does.” “I know that she loves being a leader.” “She wanted to be a chair in the true sense of the word.” Yet another wondered, “I've always been amazed that there aren't more people vying for this type of position…. There really weren't that many highly qualified candidates applying.”
Each of the three chairs was seen as exceptionally qualified. The faculty members spoke admiringly of her national prominence, her ability to understand both research and clinical practice, and her personal expertise. Although each current chair was described with unique personality attributes ranging from “soft spoken” to “energetic,” three qualities—toughness, directness, and transparency—emerged as common to all (Table 2B). Interviewees contrasted the “leadership void” of the previous chairs with the transparent, direct, and authoritative leadership of the current chairs. One faculty member in describing the chair explained, “The leader really has to make decisions…. Communication couched within a strong style is important…. There's not going to be consensus on many issues, but I think just keeping everyone informed about decisions that are happening and why is probably the most important thing.” One participant praised the chair's transparency as an important factor in allowing her to make tough decisions palatable, stressing that if major changes are made without transparency, faculty respond by “hunkering down.” Although the transparency of the chairs' style was uniformly praised, the tough and direct behavior of their chair was not always described positively. One senior female faculty member stated, “She doesn't have all the interpersonal skills to make you feel warm and fuzzy…. Our department chair is very abrupt and sometimes isn't schmoozie enough.”
Interviewees readily identified relationship-building activities by the new chair and noted her consistent focus on development of faculty and staff (Tables 2C and and2D),2D), exemplified by the quote in the theme title. A clear sense emerged that the chairs wanted “each individual, whether it be faculty or staff or resident…to succeed and then, the logical consequence of that is that the department succeeds as a whole…. I see her really caring about individuals…but I see no ego involved.” A male faculty member described this as “facilitation of others' excellence.” Key factors for creating a communal atmosphere identified by male and female faculty included providing connections, preventing faculty from feeling “lost,” and listening to faculty. Examples of statements in this category were: “Her style is one of careful listening and weighing in at appropriate times,” and she has found the balance “to be flexible enough that [she's] not telling everybody else what they need to be doing.”
It was frequently noted that the chairs took time to know faculty personally—mentoring, offering connections, and soliciting advice on certain issues (Table 2C). One faculty member described a subtle style in facilitating faculty development: “She has a mentoring style that is just inserted into conversations that she'll have, saying things like, ‘that would be really cool to write about.’” One male faculty member thought the chair's kind of mentoring provided a “role model for young faculty,” and a junior female faculty member noted that her chair “…is somebody who could be a real mentor, an academic mentor for me, which has been lacking in our department” and acknowledged that “for me, being a female, it's a wonderful role model to have.”
The chair's communal actions were also revealed through her advocacy within the institution for her department and its faculty: “She's willing to get out there and shake the trees and garner support outside the department.” Another woman faculty member echoed praise for such departmental advocacy, noting that her chair is one “…who's not afraid to go to the medical school and say, we generate a lot of income, we do a huge service to the community, we generate…referrals to this medical school, and you better start listening to us and not treating us like the second cousin….” Faculty members in all three departments saw their current chairs as visionary, although many also expressed a wait and see attitude about the chair's long-term leadership capabilities, as exemplified by the statement of this male faculty member: “…she is very energetic and well meaning, but it's too early to say how she will do. Even bad marriages start with honeymoons, okay?” Another male tenured professor said, “I think time will still tell.”
In line by line coding for overtly negative and positive statements, most related to the previous chair. Of 123 pages of transcription, we identified 18 statements about the current chairs that were coded as overtly negative and 35 coded as clearly positive. Male interviewees accounted for 81% of the text coded negative, and female interviewees accounted for 86% of the text coded positive. Those with negative statements also had positive statements, except for two senior male faculty members whose text accounted for 13 of the 18 negative statements.
Although the description of the chair's power (Table 2D) bifurcated from “no power” to “absolute power,” interviewees most often described the new chairs as building power through consensus and presence: “She has power because…she has consensus…. She's built up respect and consensus and people listen to her. And so if you go up against her, you're probably going to lose…. She used the power to pull together the department.” Faculty members indicated that they listen to their chair because she listened to them as individuals, and they approve of her because “she understood the department, she was willing to be flexible and understood the different needs of different sectors.” One female faculty member described how, “She's been doing a lot of information…[gathering] and developing rather than making significant changes so far.” Another summarized that, “She's been meeting with all of us individually, which I think has really helped see…what our needs are, what our thoughts are…. It's much more subtle—I can't say there's been any big announcements.”
One tenured male faculty member implied that the chair's position is easy and suggested, “She doesn't really have to exert much authority at this point to get cooperation in most venues—everybody is trying to make things work for her….” Another male faculty member joked that she has power simply because “the Dean made her chair.” One interviewee noted, however, that his chair does not intend to collect the most “popular” opinions, and that “power is power and she will have to exert some.” Another described his chair's power as getting “people to work together…without stepping on toes” and “getting consensus so that she doesn't have to make decisions on her own….”
Not all faculty thought that communication had improved nor felt included in decision making in the department. One senior male faculty member noted feeling outside of the sphere of influence: “I don't feel a part of [the] mission because I haven't been included in [forming] it and nobody's come and asked for my input.” Another senior male in a different department said, “There are only a couple of people from whom she seeks much advice.” These are both examples of statements coded as negative.
We examined statements in which gender and being a chair were mentioned. None of these statements linked being a woman with any doubt in competency to fulfill the chair's duties. One male interviewee did observe that, “The big time guys that came through here took one look at the chair position [and] had little interest.” Having a woman as the chair was generally seen as an indicator of positive change in the department (Table 3). Multiple statements from both male and female faculty members suggested that rather than a hindrance, having a woman chair is seen as a source of pride: making colleagues at other institutions “jealous,” a confirmation that “we're not stuck in the Dark Ages,” and a boon to recruitment. One female faculty member reported, “There's not a more natural department to fill with a woman chair than this department.” Women seemed particularly energized to have a female role model, as exemplified by statements in Table 3.
We discuss our findings in the context of socialized gender roles, research on leadership, and the qualities deemed desirable for future leaders in academic medicine, concluding that the three women chairs in our study appear to evince all these qualities.
A large body of research confirms a deep-seated bias in the assumption that men will be more competent leaders than women, especially in positions historically occupied by men.14,27–33 Rooted in the different social roles traditionally occupied by men and women is the implicit assumption that women will exhibit more behaviors and traits that fall under the rubric of “communal” and reflect primarily concern for others' welfare (e.g., kind, sympathetic, nurturing, and gentle), whereas men will exhibit more action-oriented “agentic” behaviors and traits (assertive, ambitious, forceful, and independent)29,34–36 As most leadership positions are also viewed as requiring agentic behaviors, this can lead to what Schein has termed the “think-manager-think-male” phenomenon30,37,38 and what Eagly and Karau 14 have termed “role congruity” for men in leadership. These implicit assumptions result in more positive ratings of men being considered for or functioning in leadership positions27,39,40 and simultaneously lower ratings of women (because of their role incongruity), even when accomplishments and credentials are identical. 41 These phenomena have been documented with attitudinal questionnaires30,42 and in controlled experiments.39,43 A meta-analysis 44 of 94 studies in which male and female leaders were evaluated found a tendency to favor men over women for competency, especially in male-dominated fields. In the context of this large body of research, it is heartening that competence of the three women chairs in our study never surfaced as a concern in any of the interviews.
In addition to the attribution of less leadership competence to women from the lack of alignment of implicit gender and implicit leadership norms, Heilman et al.39,43 have repeatedly documented experimentally that when women demonstrate agentic competence in traditionally male leadership positions they may trigger additional negative responses by violating socialized gender norms. Eagly et al. 44 similarly found that women who adopted an authoritarian (i.e., more stereotypically masculine) style were less favorably rated than those who used a more democratic or interpersonal (i.e., more stereotypically feminine) style. In our study, one senior woman faculty member criticized her chair as not being “schmoozie enough” and not “making you feel warm and fuzzy.” Such criticisms of agentic women in male-dominated positions are fully consonant with the extant research, which would also predict that similar criticisms would not be leveled at men in the same leadership positions. Perhaps more revealing in terms of the changing views of women's leadership, however, is that the tough, direct, agentic actions of our women chairs were generally not criticized but seen as welcome and effective, even when they were specifically acknowledged to be counterstereotypic (e.g., first statement for Department B in Table 2B).
Several studies provide evidence that clearly competent women leaders in historically male fields can avoid or mitigate negative responses elicited by gender role violation if they also exhibit communal actions, particularly in the professional setting.28,45 Our findings are fully congruent with this research in that the perceived effectiveness of our three women chairs was attributed to their ability to be both agentic (e.g., “tough,” “direct,” and “willing to get out there”) and communal (e.g., “really caring about individuals,” and”supporting all the talented people”). Our three chairs were further able to establish through their words and actions their communal motivation for becoming leaders, thus avoiding the perception of being self-promoting. Research from controlled experiments concludes that it is particularly important for women leaders not to appear to be self-promoting. Ridgeway, 46 for example, found that proof of group-oriented rather than self-oriented motivation was a prerequisite for effective leadership by women (but not men) in small groups. In different experimental paradigms, Rudman47,48 has shown that women (but not men) who are perceived as self-promoting suffer social penalties. This research would predict that the effectiveness of our chairs is enhanced by (or even predicated on) the view by their faculty that they are motivated to help others and benefit the department and that they are not self-promoting.
Although negative responses triggered by women leaders can be subtle,49,50 the tone used and statements made by the faculty regarding their chair seemed overall to be genuinely positive. This perception is confirmed by finding twice as many statements that could be coded as overtly positive vs. overtly negative. Women offered a larger percent of statements coded positive, and the only faculty members who were consistently negative were two senior men whose statements involved feeling left out of decision-making channels or referred to some specific personal negative interaction with the chair. We cannot explicitly invoke gender in their negative perceptions. However, the higher status accorded to men in socialized gender roles51,52 would predict that senior, experienced male faculty members would feel more negative toward a female chair than a male chair. This is consistent, again, with Eagly's work, 44 in which women in leadership were evaluated more harshly by male than female evaluators. Perhaps these faculty members had a closer relationship with the previous chair or had unfulfilled expectations of having greater influence with the new chair because of their long tenure in the department. For whatever reason, their statements run counter to those of most of the participants interviewed, including other senior male faculty.
In spite of the persistent mental model of leaders as male, research on gender and leadership indicates only small effect sizes in differences in effectiveness of male and female leaders.53–55 Differences in effectiveness are also found to be influenced by the degree to which the role is gender typed (i.e., male leaders are viewed as more effective in male-dominated fields). Similarly, differences in leadership style between men and women are small. In one meta-analysis of 162 wide-ranging studies on leadership style, the strongest evidence for gender difference in style was a tendency for women to adopt a more democratic or participative style and for men to adopt a more autocratic or directive style (although this difference was less in male-dominated fields). 56 The recurring description of each of our chairs as “collaborative” and “democratic” is consistent with this previous research.
Considerable research on leadership describes three general styles of leaders: transformational leaders who inspire and mentor their subordinates, transactional leaders who largely reinforce the rules of the organization, and laissez faire leaders who are essentially absent when decisions need to be made.57,58 Transformational leaders are found to be the most effective across multiple fields, 58 including academic department chairs 59 and physicians. 60 Eagly et al. 54 performed a meta-analysis of 45 studies that measured gender differences in these three categories of leadership. The effect sizes were small but consistent in finding that women in leadership were more likely than men to exhibit more features of a transformational style. In an academic setting, Rosser 61 examined the leadership effectiveness of 22 deans and also found that women were more likely than men to be viewed as effective and to exhibit qualities consistent with transformational leadership. The descriptions of the behaviors of our three women chairs are consistent with transformational leadership: listening to all departmental stakeholders, developing and exciting constituents toward a common vision, instilling a sense of pride, and nurturing and mentoring subordinates. Congruent with Eagly et al.'s findings, 54 our women chairs were also comfortable with the transactional activities of management (e.g., a “ship-shape” leadership style, willing to make unpopular decisions). In contrast to the numerous statements consistent with a laissez faire leadership style of the previous chairs, no statements suggested this view of the current chairs.
Eagly and Carli 62 note that historically appointing women to leadership positions often signifies a change in values, direction, and goals. Our interviewees frequently express this sense of their new chairs. Ryan and Haslam 63 referred to a “glass cliff” phenomenon whereby women may be appointed to high status positions historically occupied by men when circumstances make the position precarious or less attractive. Some element of this is reflected in the recurrent description of departmental “damage,” “neglect,” and the chair's need to “clean house” and also in the observation by one faculty member that prominent male candidates for the position showed little interest. Overall, however, there was no sense that the department had lowered its standards in recruiting a woman chair.
Much has been written on the need for a new style of leader in academic medicine,1–3,64–67 calling for the future-oriented department chair to have strong communication skills, to be a good listener, to possess teamwork skills, to effectively develop others through mentorship and coaching, and to be able to articulate a shared vision. Souba2,65 stresses that leaders in the current and future climate of academic medicine must be skilled at building relationships, enhancing networking, and fostering human connectivity. Many of these are transformational leadership qualities, and all are consistent with descriptions of our three department chairs by members of their respective faculties. Souba further states that academic medicine needs leaders who have “softer and more qualitative than quantitative” characteristics than past leaders. Given this emphasis on the importance of stereotypically female skills and traits, the invisibility of women in the multiple articles on leadership in academic medicine is striking, as is the absence of any reference to the extensive experimental, field, and theoretical research examining gender and leadership.1–3,66 Nevertheless, our three women chairs appear to have met all Souba's criteria for desirable physician leaders in academic medicine.2,65,68–72
The findings of our study are limited in generalizability by the nature of qualitative research, which is highly dependent on context (e.g., time in history, nature of the AMC, unique features of the departments studied, and short tenure of the current chairs vs former chairs). The generalizability of the themes that emerged from our analysis will be tested in how well they resonate with the lived experiences of others in similar situations and by future research. However, this work stands as an exploratory investigation into the relatively new phenomena of clinical departments in AMCs chaired by women. Qualitative research paradigms believe that the researcher is an important part of the research and that analysis is invariably and directly influenced by the researchers' perspectives. This is openly acknowledged as “researcher bias.” 23 In this tradition, we acknowledge that we, the researchers, are all white women raised in the United States who are at different career stages in academic medicine (medical student, postdoctoral fellow, and tenured professor). None of us are members of the departments studied, but we are all affiliated with the institution where the research took place.
This exploratory study suggests that the culture of academic medicine has moved beyond questioning women physicians' competence to lead once they are in top organizational leadership positions. The findings are consonant with experimental research indicating that women leaders are most successful when they pair stereotypic male (agentic) behaviors with stereotypic female (communal) behaviors. All three women chairs in this study are exhibiting behaviors and traits characteristic of transformational leaders that are heralded as desirable for current and future leaders in academic medicine. Our findings underscore the ability of women physicians to succeed in negotiating the complexities of gender to become and function as effective department chairs. This work also suggests that continued silence on gender issues in projecting the future leadership of academic medicine could cause the unaffordable loss of many potentially transformational leaders.
L.G. was supported by the School of Medicine and Public Health's 2008 Shapiro Summer Research Program. Dr. Isaac was supported by grant No. T-32 AG 00265 from the National Institute on Aging. Dr. Carnes is employed part time by the William S. Middleton Memorial Veterans Hospital.
The authors have no conflicts of interest to report.