Dannels, Sharon A. PhD; Yamagata, Hisashi PhD; McDade, Sharon A. EdD; Chuang, Yu-Chuan MS; Gleason, Katharine A. MPH; McLaughlin, Jean M. MA; Richman, Rosalyn C. MA; Morahan, Page S. PhD
Many medical schools have committed substantial resources to the development of leaders through internal and external programs.1–4 However, the evaluation of leadership programs is notoriously difficult.5–7 The impact of any such program is not likely to be immediate, and evaluation should ideally include long-term follow-up.8 Assessment is further complicated because it must occur in the midst of the day-to-day life and careers of busy professionals rather than in a controlled laboratory situation. Most often, evaluation of leadership development programs has relied on pre–post surveys from participants, with the posttest administered immediately after completion of the program.4,8 The pre–post survey format has occasionally been complemented by measures such as the tracking of participants' completed projects, peer-reviewed publications, and appointments or promotions.3,4,9–14 To our knowledge, longitudinal studies of leadership program outcomes using comparison groups have been exceedingly rare,8,9,15 especially within academic medicine.16
In addition to enhancing the general talents and skills necessary for leadership positions in academic health centers,1,2 several programs have addressed the persistent underrepresentation of women in leadership positions at academic medical centers.17,18 According to the Association of American Medical Colleges (AAMC) Faculty Roster, as of May 2006, only 29% of full-time female faculty members had obtained the rank of full professor or associate professor compared with 52% of their male colleagues, and only 11% of deanships at U.S. medical schools were filled by women.19 This failure represents a loss not only of intellectual capital20,21 but also of a leadership style that often offers a broad perspective and innovative approaches.21–25 Now in its 13th year, the Hedwig van Ameringen Executive Leadership in Academic Medicine (ELAM) Program for Women, within the Institute for Women's Health and Leadership at Drexel University College of Medicine, provides executive leadership training to women faculty at the associate or full professor level at schools of medicine, dentistry, and public health (www.drexelmed.edu/ELAM).18,26
This study addresses the following question: Does participation in the ELAM program enhance the leadership and career development of women faculty and their readiness to take on the challenges of leadership, compared with women from two comparison groups?
We conducted this study at the Drexel University College of Medicine and The George Washington University. We chose a pre-/posttest methodology and a longitudinal structure to compare the impact of ELAM participation on respondents' aspirations to leadership, mastery of leadership competencies, and attainment of leadership roles, using three groups considered to be similar in terms of training and professional standing. Baseline data (Time One) were collected in 2002, with follow-up data (Time Two) collected in 2006. This study was approved by the IRB processes at Drexel University College of Medicine and The George Washington University.
The ELAM group was made up of 78 eligible women from the ELAM classes of 2001–2002 (Cohort Seven; n = 44) and 2002–2003 (Cohort Eight; n = 45). These cohorts were the seventh and eighth ELAM classes since the program began in 1995, and we selected them for inclusion because they fit the time frame of the funded study. Eleven women were excluded from the original cohorts (total of 89): nine dental school faculty and two medical school faculty whose demographic and/or medical school appointment information was not identified.
AAMC comparison group
We used the AAMC Faculty Roster database to select a group of women faculty to correspond to the ELAM faculty.27 Each of the 78 fellows in the ELAM group was matched with six women medical school faculty members (n = 468). Before matching, all current and previous ELAM fellows were removed from the faculty roster list, as were those who had applied and were not accepted. In addition, faculty members from the AAMC database whose information was used in our previous studies were also removed. Next, each ELAM participant was matched with six individuals from the AAMC database. The following matching criteria were used in this priority: academic rank, department chair status, race/ethnicity, discipline, degree type, basic sciences versus clinical department, age, medical school ownership, and the fiscal year 2000 National Institutes of Health extramural awards ranking. These criteria were relaxed as necessary on the lower-priority criteria to achieve six matches. No AAMC member was matched with more than one ELAM participant. Comparison between the groups revealed that the academic rank and chair status of both groups were very close (Table 1).
The final group (NON) was composed of 63 women who applied for Cohort Seven or Eight of the ELAM program and were not accepted in those groups or in subsequent years. The NON group represented yet another comparison group which, although not formally matched on the measurable criteria used to identify the AAMC group, was known to be similar to the ELAM group in the important attribute of aspiration to leadership, as well as in stages in career progression in academic medicine.
The AAMC group most closely resembled a traditional control for the ELAM group and served as the norm for midcareer women faculty at academic medical centers against which the ELAM group could be compared. However, a desire to occupy a leadership position is believed to be a substantial predictor of leadership attainment, and the AAMC data provide no information on this characteristic. For this reason, information obtained from the NON group was particularly valuable.
The questionnaire we used was a modification of a questionnaire used with previous ELAM cohorts.18 It consisted of 34 questions with a seven-point scaled response, plus eight discrete choice questions that requested a listing of academic and administrative positions, participation in leadership programs, and national leadership positions. Topics included in the scaled-response questions were self-assessment of administrative knowledge and skills; attitudes toward administrative responsibilities; the nature and extent of networking; aspirations for administrative leadership positions; participation in leadership training programs; and academic and administrative rank, as well as other leadership positions held. We created eight composite scores considering leadership competencies from the 34 scaled-response items: knowledge of leadership theory (four items), environmental scanning (three items), financial management (three items), communication skills (seven items), conflict management (two items), diversity competence (three items), reconciliation to the demands of leadership (six items), and leadership positioning (seven items). All Cronbach alphas approached or exceeded 0.70. We field-tested the revised instrument with nine women faculty from The George Washington University School of Medicine. The final instrument we administered in 2002 was evaluated by a panel of experts, who determined it to have sufficient content validity. After experience with the instrument in 2002, the response format for questions about networking and leadership positions was modified to shorten the questionnaire before it was readministered in 2006. The 2006 questionnaire also requested verification of 2001 administrative and academic rank (recalled 2001 information) and other leadership positions held within and outside each participant's school. We used the data from the questionnaire to analyze 16 leadership indicators: administrative leadership attainment (four items), full professor academic rank (one item), leadership competencies and readiness (the eight composite scores detailed above), and leadership aspirations and education (three items).
Time One data collection
In late fall 2002, we sent the ELAM Cohort Seven, NON, and AAMC groups a cover letter, an informed consent form, and the Time One questionnaire. A total of four follow-up messages and questionnaires were sent to those who did not respond initially, extending into summer 2003. For ELAM Cohort Eight, we distributed consent forms and a longer questionnaire that included the Time One questionnaire during the first day of the ELAM program in September 2002. Additionally, ELAM Cohort Seven received the earlier, shorter version of the questionnaire during its first day of the ELAM program in September 2001.18
Time Two data collection
In spring 2006, we sent the ELAM, NON, and AAMC participants who had responded at Time One the Time Two questionnaire (available from the corresponding author on request). There were four rounds of follow-ups, using methods we had learned from previous experience to increase response rate.28
Because of the delay involved in obtaining IRB approval from the collaborating institutions,28 we were unable to administer the Time One survey to members of ELAM Cohort Seven before their participation in the program. In an effort to maintain a longitudinal focus, we made the decision to base much of the data analysis on the 2006 survey (Time Two), with information recalled about 2001 status serving as a proxy for the Time One survey. When we used information from the Time One survey (i.e., leadership competencies), we relied mainly on data from ELAM Cohort Eight, with supplemental information on Cohort Seven taken from the earlier version of the questionnaire completed at enrollment in 200118 where the data were deemed sufficiently reliable.
Data verification, adjustments, and analysis
Two reviewers manually coded all questionnaires (Y.C. and J.M.) and, when it was clear that responses had been incorrectly marked, edited them. Data entry was fully verified, and questionable data were identified using computerized edits and checks against the original data. We coded missing data as missing and did not include them in the analyses. We also coded logically inconsistent responses as missing.
We assessed the internal consistency reliability for the eight leadership readiness/competency composites for both the Time One and Time Two data. Cronbach alphas were above 0.70, with the exception in the Time One data of conflict management (α = 0.658) and leadership positioning (α = 0.688). We calculated eight ANCOVAs using the Time Two composite scores as the dependent variable, academic rank and the associated Time One composite score as the covariates, and the group as the “independent” variable. Assumptions of homogeneity of slopes (MANOVA) and homogeneity of variances (Levene), tested with an alpha of 0.1, indicated that they were met for all eight tests. We pursued significant ANCOVAs with pairwise comparisons of the adjusted group means, using the Holm sequential Bonferroni procedure29 to control for Type 1 error (α = 0.05). Strength of association was determined using partial η2. Eight composites were formed for ELAM Cohort Seven, using items available from the shorter version of the survey administered at the beginning of the ELAM program. We pursued three ANCOVAs (environmental scanning, financial management, and diversity competence); the other five tests failed to meet either the assumption of homogeneity of slope and/or variances.
We calculated the χ2 test of homogeneity of proportions to evaluate differences in the distribution of leadership indicators, behaviors, and aspirations across the groups. All assumptions of the χ2 test were met, although in some cases it was necessary to collapse categories and/or to exclude the NON group from the comparison to ensure that at least 80% of the cells had an expected frequency of five or greater (Ejk >5). We interpreted φ-coefficients using the conventional values of 0.10, 0.30, and 0.50 to suggest small, moderate, and large effect sizes, respectively. Analyses were conducted using SPSS statistical software, version 15.0 (SPSS Inc., Chicago, Illinois).
Response rate and evaluation of matching procedure
Response rates varied across the three groups. A high percentage (57 of 80, or 71%) of the ELAM group participated. There was a lower response rate in the groups that had no experience with the ELAM program; 28 of 63 (44%) (n = 28/63) of the NON, and 178 of 468 (38%) of the AAMC individuals, responded to both the Time One and Time Two questionnaires. Both mail and online questionnaires were returned from all three groups. The mean values of the data collected were similar for the two methods of data collection.
The matching process indicated very close concordance between AAMC and ELAM groups for academic rank and department chair administrative position, and the Time Two recall data indicated no significant differences on these two measures (Table 1); yet, information obtained from self-report did not always mirror that of the matching process. For example, the proportion of respondents reporting a position of chair was considerably higher than indicated by matching. Pearson χ2 tests did not show any significant differences among the various groups for academic rank. For academic rank matching in 2001 with the AAMC and ELAM groups, χ21 = 0.005, P = .944. For the 2001 information on academic rank recalled in 2006 (AAMC, NON, ELAM), χ22 = 2.395, P = .302, φ = 0.097. Pearson χ2 tests also showed no difference on the matching for chair position in 2001, but they did show differences in the administrative positions in the 2001 information recalled in 2006. For the 2001 department chair matching data for AAMC and ELAM groups, with continuity correction, χ21 = 0.002, P = .966. For the 2001 information on administrative positions recalled in 2006 (using AAMC and ELAM groups only, so that assumption of sufficient sample size could be met), χ24 = 36.525, P < .001, φ = 0.409.
The most likely explanation for the difference between matching and self-report is the delay in schools' voluntary reporting and updating of faculty information to the AAMC Faculty Roster. Also, although the percentage of respondents reporting a position of chair was very close across groups (12.4% AAMC, 14.8% NON, 14.0% ELAM), there was a significant overall difference between the AAMC and ELAM groups for administrative positions (Table 1). Additional potential explanations for these differences include a differential response rate, both in terms of the overall proportion of the eligible pool responding and by academic position (i.e., ELAM associate professors may have responded more than ELAM full professors, and AAMC full professors may have responded more than AAMC associates), and recall bias contributing to the difference in rank in 2001 as recalled at Time Two (2006). To control for the differences between the matching and self-report data, we thus decided to use 2001 academic rank as the covariate in analyses when appropriate.
Leadership roles and attainment
There was a significant difference in the highest administrative positions reported among the three groups for the time period 2001 to 2006 (Tables 2 and 3). Pearson χ2 tests showed a significant difference for administrative rank, and no difference for academic rank. Comparison for highest administrative rank on the five position categories shown in Tables 1 and 2, and using AAMC and ELAM groups only so that assumption of sufficient sample size could be met, yielded values of χ24 = 41.670, P < .001, φ = 0.452. Comparing AAMC, NON, and ELAM groups on highest academic rank held between 2001 and 2006 as recalled at Time Two (2006) generated values of χ22 = 3.046, P = .22, φ = 0.112.
Almost two-thirds of the ELAM group (33, or 63.5%) reported positions of department chair or greater (chair or center director; senior administrative staff in the dean's office, including vice, executive associate, senior associate, associate and assistant dean; or positions of dean and above, including vice president, associate vice president, provost, president, chancellor, and chief executive officer). This was considerably greater compared with the 36 (22.5%) reported by the AAMC group and 10 (37.0%) by the NON group (χ22 = 29.96, P < .001, φ = 0.354) (Table 3). When the data were analyzed by types and levels of positions, administrative roles below chair (e.g., course, clerkship, residency, fellowship, or program director; vice or associate program, division, or section director, chief, or head; division/section director, chief, or head), these positions were reported by a larger proportion of the AAMC group (89, or 55.6%) and NON group (14, or 51.9%) compared with the ELAM group (18, or 34.6%). Moreover, 27 (16.9%) of the AAMC and three (11.1%) of the NON groups reported having no administrative roles, in contrast with only one (1.9%) of the ELAM group (Table 2).
Consistent with the reports on administrative positions, the proportion of the ELAM respondents (11, or 44.8%) reporting that 50% or more of their time was devoted to administrative duties was significantly greater (χ22 = 10.821, P = .004, φ = 0.214) than for either the AAMC (32, or 17.8%) or NON (6, or 21.4%) groups, as was the percentage of respondents chairing one or more college or university committee (χ22 = 6.86, P = .03, φ = 0.168) (Table 3). No significant differences were found in the percentage holding one or more national position, although the ELAM group reported the highest proportions (Table 3).
The data for academic promotion to full professor were mixed (Tables 1 and 2). ELAM fellows reported an increase in attaining the rank of full professor, from 26 (44.8%) in 2001 to 37 (69.8%) between 2001 and 2006. The AAMC group reported a higher proportion of full professors in 2001 (96, or 55.4%), which increased to 118 (68.6%), similar to the ELAM group. The NON group reported a percentage of full professors in 2001 similar to that of the ELAM group (12 of 26, or 46.1%), and 12 of 25 (48.0%) in 2006.
Knowledge and confidence in leadership areas
In all eight leadership composites, the ELAM group mean (based on a seven-point scale) was greater than the means in both the AAMC and NON groups (Table 4). The ANCOVAs revealed that the difference in five composites—knowledge of theory; environmental scanning; financial management; communication skills; and tolerance for the demands of leadership—was significant at P < .001. Two composites, conflict management and leadership positioning, were significant at P < .05. Only one composite, diversity competence, did not yield a statistically significant difference between groups (means for all groups were high, with a spread of 5.98 to 6.20). ANCOVAs for the three composites that were analyzable for ELAM Cohort Seven (environmental scanning, financial management, and diversity competence) showed results similar to those for ELAM Cohort Eight.
Pairwise comparisons of the significant ANCOVAs revealed that the ELAM group was significantly higher than both the AAMC and NON comparison groups for six of the seven significant composites (Table 5). In the leadership positioning composite, the ELAM group was significantly different from the NON group, but not the AAMC group. In all eight composites, there were no significant differences between the AAMC and the NON comparison groups.
Leadership aspirations and advanced leadership education
Our inquiry at Time Two about aspirations to leadership positions both inside and outside academic health centers revealed that people in the ELAM group were significantly more interested than those in the other groups in obtaining a higher leadership position inside an academic health center (χ22 = 12.903, P = .002, φ = 0.223) (Table 6). In terms of change over time for this sample, the proportion of AAMC respondents reporting such aspirations remained unchanged (89, or 50.6%, in 2002; and 88, or 49.4%, in 2006), decreased slightly for the ELAM group (23, or 85.2%, in 2002 [only Cohort Eight data available]; and 42, or 76.4%, [both cohorts], in 2006), and decreased considerably for the NON group (24, or 88.9% in 2002; and 17, or 63%, in 2006).
The percentage of participants aspiring to leadership positions outside academic health centers was similar and essentially unchanged over time. The ELAM group reported 10 (35%) in 2002 (only Cohort Eight data available) and 19 (33.9%) in 2006 (both cohorts); AAMC reported 35 (20%) in 2002 and 51 (28.3%) in 2006; and NON reported 11 (44%) in 2002 and 9 (37.5%) in 2006 (Table 6).
As a final indicator of leadership aspiration, a significantly greater proportion of the ELAM group (41, or 78.8%; χ22 = 37.73, P < .001; φ = 0.392) reported participation between 2001 and 2006 in one or more leadership or management education certificate or degree-granting programs than did either the NON (9, or 34.6%) or the AAMC (52, or 31.1%) groups (Table 6).
The overarching question for this study was: Does the ELAM Program for Women have a positive impact on its participants four to five years after completion of the program? This question was examined by comparing women who participated in ELAM with two related groups: women who applied but were not accepted into the ELAM program (NON group), and women faculty matched from the AAMC Faculty Roster (AAMC group). The AAMC group represented the norm for midcareer women faculty in academic medicine. The NON group provided a comparison group similar to ELAM participants in general backgrounds and leadership aspirations. The results of this study support the hypothesis that the ELAM program has a beneficial impact on ELAM fellows in terms of leadership knowledge, behaviors, and career progression.
We considered 16 leadership indicators for this study: administrative leadership attainment (four items), full professor academic rank (one item), leadership competencies and readiness (eight composite scores), and leadership aspirations and education (three items). The means and frequencies reported by the ELAM group were greater than those of the NON and AAMC comparison groups for 15 indicators. For 12 of these indicators, the difference was statistically significant, indicating higher levels of leadership development among the ELAM fellows. About 70% of the AAMC group reported having full professor status, 23% held leadership positions of chair or higher, 18% spent more than half their time in administrative responsibilities, 28% chaired one or more committees, and 49% aspired to leadership positions within academic health centers. The NON group mirrored the AAMC group. In contrast, for the same time period, the ELAM group matched or surpassed the AAMC and NON groups, reporting 70% at full professor level, 64% in leadership positions of chair or higher, 45% spending more than half their time in administrative responsibilities, 47% chairing one or more committees, and 76% aspiring to leadership within academic health centers.
These differences were consistent with the differences found in the ANCOVA analyses of the leadership knowledge composites for the ELAM group compared with the AAMC or NON groups. The ELAM group had significantly greater means in seven of the eight composites. Only the diversity composite scores were similar; however, the reported scores were very high in all groups, suggesting that most of the senior women faculty believe they can work effectively with different gender, race/ethnicity, and sexual orientation groups.
Our findings regarding the impact of an aspiration to leadership on leadership competencies are not as clear-cut. As discussed above, pairwise comparisons of the eight leadership composites showed no statistically significant differences between the NON and AAMC groups. However, in five of the other eight leadership measures, the responses reported by the NON group were between those of the AAMC and ELAM groups, suggesting a level of attainment midway between the two groups. These findings suggest that, at least for this group of academic women, aspirations to leadership alone are not enough to develop most of the leadership competencies that we examined. Given that a higher percentage of ELAM fellows have achieved leadership positions, these competencies may best be acquired through a combination of leadership training such as ELAM and on-the-job experience.
The finding that about one third of all three groups of senior women faculty aspired to a higher leadership position outside academic health centers is troubling. These results may indicate that academic health centers are at risk of losing the intellectual capital of this group of accomplished senior women faculty.20,21,30 Similar studies comparing the leadership aspirations of men and women faculty over time are warranted.31,32
Our study also emphasizes the many challenges in attempting to evaluate prospectively the impact of a leadership program.5,33,34 First, the difficulties encountered in the implementation of the study design included the delayed distribution of Time One surveys to ELAM Cohort Seven, as well as issues of recall and response bias. The difference observed between respondents' self-report data and the data from the AAMC database highlight the difficulty in obtaining completely matched comparison groups in real-time, real-world settings. This difficulty somewhat limits the confidence with which we can interpret our findings, because there is some indication that the ELAM group was ahead of the AAMC at study entry in the proportion of women in administrative positions, though not academic rank.
A second, related issue is that the three comparison groups cannot be considered a case–control situation of the type used in clinical studies. Although numerous efforts were made to control for threats to internal validity, we cannot conclude that participation in ELAM is the only factor responsible for the observed differences between groups. However, we can reasonably conclude that such participation does have some impact. The ELAM program is designed both to increase leadership knowledge and skills of the fellows and to increase their visibility and perceived value to AHC leaders including deans.
These inherent limitations of leadership evaluation studies emphasize the importance of using multiple sources of data to establish outcomes. In addition to the present study, we have completed a within-group pre- and postsurvey of leadership development18 as well as a qualitative study of how the ELAM program fosters development of self-efficacy among the women participants.35 Our other ongoing studies include longitudinal, qualitative, and quantitative repeated-measures research and an ongoing survey of medical and dental school deans regarding the impact of ELAM participants on their home institutions. Through the compilation of such studies, we wish to establish solid findings on the impact of women's leadership development programs on the advancement of women in academic medicine, and to develop theory about leadership development for women.
The goals of the Hedwig van Ameringen ELAM Program for Women are to increase the number of women in leadership positions in academic medicine, dentistry, and public health; to ensure that women are successful in leadership positions; and to effect change in the culture of academic health centers so that the unique contributions of women are valued. This study is one of the very first to attempt to assess the midterm impact of a leadership program in academic medicine on the basis of comparison groups. Taken together, our findings support the thesis that the ELAM program provides tangible benefits to the women participants in terms of attainment of leadership positions, mastery of leadership competencies, and aspirations to and education in leadership.
This research was supported in part by grants from the Robert Wood Johnson Foundation and the Jessie Ball duPont Fund, as well as support from the Mayo Medical School, University of Michigan Medical School, Vanderbilt University School of Medicine, and the Wright State University School of Medicine. None of the sponsors had a role in the design or conduct of the study; collection, management, analysis, or interpretation of the data; or preparation, review, or approval of the manuscript.
The authors thank Gregg Jackson, EdD, for his critical role in the initial conception and design of the study, and for acquisition and initial analysis of the data from the survey; Victoria C. Odhner, for her contributions to locating participants, administration of the questionnaires, and other research administration; and several graduate students, Ray Francis, Elizabeth Verostek, and Brian Sponsler, in the Center for Educational Leadership and Transformation at The George Washington University, for their assistance in data entry and research administration.
1 Grigsby RK, Hefner DS, Souba WW, Kirch D. The future-oriented department chair. Acad Med. 2004;76:571–577.
2 Morahan PS, Kasperbauer D, McDade SA, et al. Training future leaders of academic medicine: Internal programs at three academic health centers. Acad Med. 1998;73:1159–1169.
3 Wilkerson L, Uijtdehaage S, Relan A. Increasing the pool of educational leaders for UCLA. Acad Med. 2006;81:954–958.
4 Simpson D, Marcdante K, Morzinski J, et al. Fifteen years of aligning faculty development with primary care clinician–educator roles and academic advancement at the Medical College of Wisconsin. Acad Med. 2006;81:945–953.
5 Hannum K, Marineau JW, Reinelt C, eds. The Handbook of Leadership Development Evaluation. San Franscisco, Calif: Jossey-Bass; 2006.
6 Tucker PD, Henig CB, Salmonowicz MJ. Learning outcomes of an educational leadership cohort program. Educ Considerations. Spring. 2005;32:27–35.
7 McDade S. Evaluating Leadership Development Programs. Developing Administrative Excellence: Creating a Culture of Leadership—New Directions for Higher Education. Vol 87. San Francisco, Calif: Jossey-Bass; 1997.
8 Reinelt C, Russon C. Building Leadership Bridges. College Park, Md: International Leadership Association; 2003.
9 Hirst G, Mann L, Bain P, Pirola-Merlo A, Richver A. Learning to lead: The development and testing of a model of leadership learning. Leadersh Q. 2004;15:311–327.
10 Millett R, Reinelt C, Weber W. Developing leadership in an international context: A summary of the final evaluation report of the Kellogg International Leadership Program II. Battle Creek, Mich: W. K. Kellogg Foundation; 2000.
11 Gruppen LD, Frohna AZ, Anderson RM, Lowe KD. Faculty development for educational leadership and scholarship. Acad Med. 2003;78:137–141.
12 Hewson MG, Copeland HL. Exploring the outcomes of faculty development programs. Acad Med. 1999;74(10 suppl):S68–S71.
13 Markus G. Building Leadership: Findings From a Longitudinal Evaluation of the Kellogg National Fellowship Program. Battle Creek, Mich: W. K. Kellogg Foundation; 2001.
14 Kalet AL, Fletcher KE, Ferdman DJ, Bickell NA. Defining, navigating, and negotiating success: The experiences of mid-career Robert Wood Johnson Clinical Scholar Women. J Gen Intern Med. 2006;21:920–925.
15 U.S. Fulbright Scholar Program Outcome Assessment. Washington, DC: Bureau of Educational and Cultural Affairs, Office of Policy and Evaluation; 2002.
17 Bickel J, Wara DW, Atkinson BF, et al. Increasing women's leadership in academic medicine: Report of the AAMC Project Implementation Committee. Acad Med. 2002;77:1043–1061.
18 McDade S, Richman RC, Jackson GB, Morahan PS. Effects of participation in the Executive Leadership in Academic Medicine (ELAM) program on women faculty's perceived leadership capabilities. Acad Med. 2004;79:302–309.
20 Morahan PS, Bickel J. Capitalizing on women's intellectual capital in the professions. Acad Med. 2002;77:110–112.
21 Kramer VW, Konrad AM, Erkut S. Critical Mass on Corporate Boards: Why Three or More Women Enhance Governance. Wellesley, Mass: Wellesley Centers for Women; 2006. Report no. WCW 11.
22 Coughlin L, Wingard E, Hollihan K, eds. Enlightened Power: How Women Are Transforming the Practice of Leadership. San Francisco, Calif: Jossey-Bass; 2005.
23 Eagly AH, Johannsen-Schmidt MC. Transformational, transactional and laissez-faire leadership styles: A meta-analysis comparing women and men. Psychol Bull. 2003;129:569–591.
24 European Commission. Women in Science and Technology—The Business Perspective. Brussels, Belgium: European Commission; 2006.
25 Carnes M, Bland C. Viewpoint: A challenge to academic health centers and the National Institutes of Health to prevent unintended gender bias in the selection of clinical and translational science award leaders. Acad Med. 2007;82:202–206.
28 Morahan PS, Yamagata H, McDade S, Richman RC, Francis R, Odhner V. New challenges facing inter-institutional social science and educational program evaluation research at academic health centers. Acad Med. 2006;81:527–535.
29 Green SB, Slakind NJ. Using SPSS for Windows and Macintosh, 4th ed. Upper Saddle River, NJ: Pearson Prentice Hall; 2005.
30 Waldman JD, Kelly F, Arora S, Smith HL. The shocking costs of turnover in health care. Health Care Manage Rev. 2004;29:2–7.
31 Buckley LM, Sanders K, Shih M, Hampton CL. Attitudes of clinical faculty about career progress, career success and recognition, and commitment to academic medicine: Results of a survey. Arch Intern Med. 2000;160:2625–2629.
32 Buckley LM, Sanders K, Shih M, Kallar S, Hampton C. Obstacles to promotion? Values of women faculty about career success and recognition. Acad Med. 2000;75:283–288.
33 Schall E. Learning to love the swamp: Reshaping education for public service. J Policy Anal Manage. 1995;14:202–220.
34 Norcini J, Burdick W, Morahan PS. The FAIMER Institute: Creating international networks of medical educators. Med Teach. 2005;27:214–281.
35 Sloma-Williams L, McDade SA, Richman RC, Morahan PS. The role of self-efficacy in developing women leaders: A case of women leaders in academic medicine and dentistry. In: Dean DR, Bracken SJ, Allen JK, eds. Women in Academic Leadership: Professional Strategies, Personal Choices. Sterling, Va: Stylus Publishing; 2007.