In this issue, Nomura and colleagues1 report that in a large study of primary care resident physicians in Japan, women on average rated themselves less confident than their male counterparts in most but not all of their clinical skills. Male physicians on average expressed greater interest in leadership pathways, and female physicians on average expressed greater family orientation—both of which are congruent with socialized gender norms. Perspectives on work and life, while different on average for men and women, were not significantly related to confidence levels. The authors fit their findings into some of the research on gender differences in physicians from the United States and Europe.
Additional insights into the complex ways gender may be influencing Nomura and colleagues' findings come from social psychology. Research in this field suggests that the social expectation of women to be modest, the phenomenon of stereotype threat, and the influence of gender stereotype priming may significantly influence women's assessments of their own competence. Given the differences in social roles for men and women and the prescriptive behavior-related gender norms, more noteworthy than the observed differences in this large sample may be how small these differences are for most items and the degree of similarity between male and female physicians' responses. The following discussion focuses on relevant research on gender issues and cautions against relying on mean differences between groups of people to form conclusions about individual members of those groups.
Prescriptive Gender Norms for Women: “Don't Brag”
In all societies, gender is a powerful social category that carries with it lifelong socially constructed messages about how males and females should (and should not) behave. Across all societies, the prescriptive gender norms for females involve communal behaviors, such as being nurturing and supportive, while those for males are generally more agentic, such as being assertive and independent.2 Both males and females suffer social penalties if they exhibit behaviors outside these gender norms. The fact that our language contains many pejorative terms for men (e.g., “wimpy”) and women (e.g., “bossy”) whose behavior is perceived to violate gender norms is an indication of the social costs of doing so. Congruent with gendered behavioral norms, girls early in life generally receive negative messages from parents and others if they boast about their skills or accomplishments—rather, they are imbued with messages that reinforce modesty as the socially desirable behavior for girls. Such messages are less likely to be part of the behavioral script for boys for whom boldness may be encouraged. Indeed, studies document the tendency for men, including male physicians,3 to overestimate their actual task competence. To succeed professionally, women must learn the negative consequences of violating the prescriptive norms for female modesty in employment settings. In carefully designed, controlled experiments, Rudman has repeatedly demonstrated that women—but not equivalently qualified men—who self-promote (i.e., brag about their expertise) have less favorable employment and economic outcomes than nonbraggers and that the presence or absence of such behavior is given more weight than competence in hiring decisions for women but not men.4
More salient than physician gender differences in self-expressed confidence are measures of actual clinical competence. On this matter, several studies (e.g., Lind and colleagues) indicate that female physicians are rated as equivalently or more competent than their male counterparts.3 Whether female physicians' lower self-assessment of their skill contributes to their lower likelihood of career advancement relative to men or whether this modesty favorably facilitates their career advancement remains unknown. In support of the latter is the finding of Bartels and colleagues5 that residents of both genders described the need for female residents to adopt more stereotypically female behaviors, including speaking with a nicer “tone,” to enhance their effectiveness. Modesty may be another learned behavior that promotes rather than hinders women's success in traditionally male fields by providing explicit evidence of communality.6 The inability to predict the impact of such behavior on either actual competence or career outcomes of female physicians underscores the multifaceted and complex effects of gender at the individual and organizational level. As indicated by Nomura and colleagues, the correlation between the results of these questionnaires and the career outcomes of the respondents will require longitudinal follow-up.
Because of the socialized differences in role expectations, it is not surprising that Nomura and colleagues found that more women than men surveyed indicated family as the most important thing in life or that more men than women expressed an interest in activities that lead to leadership (e.g., research, administration, and an advanced degree). The penalties for gender role violation would predict that even if an individual woman did not endorse the female stereotype of prioritizing family or did endorse the male stereotype of aspiring to leadership, she would be less likely than a man to acknowledge this. Against this backdrop of prescriptive gender norms, it is perhaps more remarkable that many female residents surveyed did admit to prioritizing professional commitment (18%) and to having leadership aspirations (13–28%) and that many men did express an orientation toward family (54%) and work-life balance (54%).
Gender Stereotype Priming and the Threat of Incompetence
Exposure to information that reinforces gender stereotypes reliably promotes gender bias in subsequent decision making. Physicians are bombarded with such gender stereotype priming in their environment: women are predominant in subordinate caregiving roles as nurses and support staff, men are heavily overrepresented in high-status fields (e.g., surgery) and in hospital and department leadership roles, and portraits of male physicians and scientists frequently line the walls of hallways and conference rooms in hospitals and medical schools. Steele's group first described and has conducted extensive research on the phenomenon of stereotype threat.7 This construct holds that if a negative stereotype about the competence of a group exists, members of that group, fearing that they will be judged in accordance with the stereotype, may perform at a level below their actual ability. In a U.S. study, Davies et al.8 examined stereotype threat for women and leadership and found that female gender stereotype priming with media images made women (but not men) less likely to select a leadership role in a subsequent group task. In Nomura and colleagues' study, stereotype threat could have been triggered for female residents taking the questionnaire if they were asked to identify their gender at the beginning rather than the end of the survey or simply by the stereotype priming ever-present in their environment. Self-perceived competence (self-efficacy) has been closely linked to successful task performance and career persistence. However, the possible influence of stereotype threat activation, in conjunction with the lack of association between female physicians' self-rated competence and actual performance in other research, cautions against relying on Nomura and colleagues' study to draw conclusions about comparisons between female and male residents' actual competence. As more women enter medicine, a useful intervention to mitigate the effect of stereotype threat and promote women's self-efficacy would be to permeate the training and practice environment with clear messages that “research indicates no gender differences in the ability to perform any of the clinical tasks required of physicians.” Davies et al. found that such statements of affirmation eliminated the negative impact of stereotype threat on women's self-selection of a leadership role (and had no impact on men).
It is interesting that Nomura and colleagues found that female residents rated themselves the same or slightly better than male residents on perceived competence in blood drawing and insertion of urinary catheters. It is possible that of all the clinical tasks assessed, female residents on average actually have selectively greater competence in these tasks than male residents with the same amount of experience. More compelling, however, is to consider whether the discrepancy between gender differences for these tasks and that for other clinical tasks assessed (e.g., lumbar puncture or chest compression) is conflated with gender differences in who typically performs these tasks and the resultant assumption of technical difficulty and status associated with them. Tasks usually performed by men are often implicitly assumed to be technically more difficult and have higher status whether this is true or not. In the United States, blood drawing and urinary catheter insertion are most often performed by nurses or nonphysician clinical staff, most of whom are women. If this is also the case in Japan, residents may not perceive these tasks to belong as definitively in the “masculine-task” domain as the other clinical tasks. Therefore, questions about competence on these tasks may be less likely to trigger stereotype threat for female residents and also less likely to invoke the need for modesty in self-reported confidence in performing these lower status tasks.
Promoting Development of All Talent
Because biological differences do not constrain men's or women's performance as physicians, the challenge is to prevent socially constructed gender-based assumptions from inhibiting the full participation of men and women in all the complex aspects of improving the health of patients and populations. An important first step is to acknowledge that gender norms for behavior exist, are powerful, and may differentially influence female physicians' early self-assessed confidence in some stereotypically male agentic clinical tasks, and perhaps may affect male physicians' self-assessed confidence in some stereotypically female communal clinical competencies, such as empathy.9 As suggested by Nomura and colleagues, this information could be incorporated into the medical curriculum at multiple levels. Studies on mitigating gender bias also support including and often repeating in the curriculum the explicit message that research has found no gender difference in the ability to function as physicians, leaders, or scientists. With all due respect to the past patriarchs of medicine, the visual display of their portraits in modern academic medical centers may trigger stereotype threat for women. Such male- stereotype gender priming may also lead to more positive evaluations of male than female physicians or scientists with comparable expertise working in this environment. Removing these portraits can send an important message to all stakeholders about the value of inclusiveness in academic medicine. Alternatively, given the evidence that counterstereotype imaging (i.e., intentionally imagining a “strong” woman) can reduce unconscious gender bias,10 interspersing portraits of women who are successful physicians and scientists with those of men may send an equally effective message.
Male and Female Physicians: More Alike Than Different
The title of Nomura and colleagues' article emphasizes the gender difference in self-rated clinical confidence. Such an emphasis may be misleading and a disservice to both male and female physicians. More noteworthy is that in spite of the powerful prescriptions of gender and the male-dominated workplace of Japanese medicine noted by the authors, these differences were very small. Most female and male residents scored themselves in the “feel somewhat confident” range. While the differences between male and female residents' levels of confidence were smaller in the relational aspects of patient care, their responses, falling closer to the “not very confident” range, suggest the need for improvement for both genders. The authors provide the confidence intervals around the mean difference for male and female residents, but they provide no data on the dispersion of the actual responses. The small difference in average responses for male and female residents suggests that there was considerable overlap in the distribution of responses—that there were many women who expressed greater confidence than men and many men who expressed less confidence than women. Similarly, it would be statistical discrimination to assume that all female residents favor family over professional commitment when 18% indicated that professional commitment was more important. More men than women aspired toward leadership, but many women also expressed these career goals. Both men and women expressed an equivalent desire for work-life balance. Because Nomura and colleagues conducted a large study with a high response rate, the gender differences they report allow for considerable reflection regarding gender and medicine, and correlations with future career outcomes will be interesting. One must exercise caution, however, in relying on the average differences between a large group of residents to draw conclusions about all male or all female physicians. Using the findings of Nomura and colleagues to reinforce gender stereotypes will continue to impede each individual female physician's career advancement and each individual male physician's struggle for work-life balance.
Dr. Carnes thanks Eve Fine, PhD, and Carol Isaac, PhD, for their thoughtful reading of and contributions to earlier drafts of this commentary.
Dr. Carnes' research on gender and the advancement of women in academic science and engineering is funded by NSF SBE- 0619979 and NIH R01 GM088477-01.
1 Nomura K, Yano E, Fukui T. Gender differences in clinical confidence: A nationwide survey of resident physicians in Japan. Acad Med. 2010;85:647–653.
2 Heilman M. Description and prescription: How gender stereotypes prevent women's ascent up the organizational ladder. Journal of Social Issues. 2001;57:657–674.
3 Lind DS, Rekkas S, Bui V, Lam T, Beierle E, Copeland EM, 3rd. Competency-based student self-assessment on a surgery rotation. J Surg Res. Jun 1 2002;105:31–34.
4 Phelan JE, Moss-Racusin CA, Rudman LA. Competent yet out in the cold: Shifting criteria for hiring reflect backlash toward agentic women. Psychology of Women Quarterly. 2008;32:406–413.
5 Bartels C, Goetz S, Ward E, Carnes M. Internal medicine residents' perceived ability to direct patient care: impact of gender and experience. J Womens Health. 2008;17:1615–1621.
6 Heilman ME, Okimoto TG. Why are women penalized for success at male tasks?: The implied communality deficit. J Appl Psychol. 2007;92:81–92.
7 Steele CM. A threat in the air. How stereotypes shape intellectual identity and performance. Am Psychol. 1997;52:613–629.
8 Davies PG, Spencer SJ, Steele CM. Clearing the air: Identity safety moderates the effects of stereotype threat on women's leadership aspirations. J Pers Soc Psychol. 2005;88:276–287.
9 Hojat M, Gonnella JS, Mangione S, et al. Empathy in medical students as related to academic performance, clinical competence and gender. Med Educ. 2002;36:522–527.
10 Blair IV, Ma JE, Lenton AP. Imagining stereotypes away: The moderation of implicit stereotypes through mental imagery. J Pers Soc Psychol. 2001;81(5):828–841.