The convocation was organized in four main sessions: biological components of success in science and engineering, social components of success, institutional structures that affect recruitment and retention of women scientists and engineers, and a final session on current institutional transformation efforts. Several major themes emerged from these sessions.
SEX DIFFERENCES IN COGNITIVE ABILITIES
The members of the first panel, “Cognitive and Biological Contributions,” presented evidence of differences between males and females in the trajectories of brain development and in average performance on verbal, mathematical and spatial cognitive tasks.
Janet Hyde explained that the largest differences were seen in the extremes of performance distributions, with more males found in the top and bottom tails; even so, within-gender differences were much larger than between-gender differences. She presented meta-analyses that suggested that the current stereotypes in which boys and men are believed superior in mathematics skills and girls in verbal skills should be replaced with a “gender similarities” hypothesis, in which women and men are more psychologically similar than they are different. Hyde also provided data that showed girls in Taiwan and China outperformed US boys in mathematics; at issue, she emphasized, is not whether US boys do better than US girls but that US children in general are underperforming relative to nations the United States is competing with economically.
Cognitive differences showed strong dependence on age and experience, explained Jay Giedd. He provided evidence from longitudinal MRI studies of
adolescents. Bruce McEwen provided evidence of differential responses to stress and suggested that there may be sex differences in learning strategies. On the basis of data showing that the influence of experience on brain development is strong and lifelong, Giedd and McEwen independently suggested that any difference between males and females be viewed more as an opportunity for education and research than as an intrinsic constraint on cognitive capability.
There was some discussion among the panelists on whether sex differences in cognition were large enough to account for the size and nature of the discrepancies between male and female representation among academic scientists; disagreement centered on the degree to which small differences could accumulate over time and have a substantial effect on careers. In that context, Diane Halpern proposed a biopsychosocial model of development, in which experience alters the biological underpinnings of behavior, which in turn influences the types of experiences to which we are exposed.
GENDER STEREOTYPES AND ACADEMIC PERFORMANCE
Science and engineering are widely stereotyped as male domains in American culture. That context influences the performance of those not expected to succeed, explained Toni Schmader. A person’s belief that he or she belongs to a group stereotyped as inferior in a given ability may, when combined with certain contextual cues, trigger a phenomenon termed stereotype threat by Claude Steele. When this happens, the person’s cognitive performance, particularly on tests of mathematics ability among women and tests of general intellectual ability among members of racial and ethnic minorities, is negatively affected. Schmader explained that contextual factors, such as predominant stereotypes, can discourage people, especially women and minority-group members, from aspiring to and pursuing science and engineering education and careers and from taking leadership roles. They also reduce their chances of being accepted into educational programs whose admission requirements emphasize test scores.
Pervasive unexamined bias against women in science and engineering influences evaluations of women scientists’ motivation, determination, promise, seriousness, and productivity and can undermine the perception of the quality of their work throughout their careers, explained Mahzarin Rustum Banaji. Small differences in advantage can accumulate over the span of a career into large differences in status and prestige. That results in male scientists often receiving greater rewards for their accomplishments than female or minority-group scientists, said Donna Ginther.
Modern gender bias, in addition to being pervasive, is automatic, ambiguous, and ambivalent, said Susan Fiske, who presented data showing the female gender
role underlying this bias is both prescriptive and descriptive, demanding from women such traits as subservience and caring, thereby limiting their ability to be perceived as effective in traditionally male roles.
CAREER TRAJECTORIES IN SCIENCE AND ENGINEERING
Several panelists—including Ginther, Yu Xie, Robert Drago, Joan Williams, and Angelica Stacy—examined the factors that affect career trajectories in science and engineering.
Since the 1970s, there has been tremendous growth in the overall number of bachelor’s and doctorate degrees awarded to women, but Ginther showed that women’s representation is dependent upon field. In the physical sciences and engineering, women earn no more than 20% of the doctorate degrees, while in social sciences and biology women earn no less than half of doctorates.
Independently, panelists found that the factor most detrimental to career progression was family status. Their data indicated that married women scientists are disadvantaged, particularly if they have children: they are less likely to pursue careers in science and engineering even with an advanced degree, they are less likely to be in the labor force, they are less likely to be promoted, and they are less likely to be geographically mobile.
Married men with young children are 50% more likely to enter tenure-track jobs than comparable women, said Stacy. Ginther presented data showing that regardless of field, young children significantly decrease the likelihood of women—but not men—in obtaining a tenure-track job.
BIAS AGAINST CAREGIVERS
Fiske presented data showing that American culture in general strongly stereotypes caregiving—whether of children, the elderly, or sick or disabled family members—as work appropriate to females. As described above, several panelists independently identified motherhood as the factor most likely to keep a woman with science training from pursuing or advancing in a scientific career.
Scientists are generally well aware of the bias against caregiving, and those seeking fast-track academic careers use a number of strategies to avoid damage to their careers by caregiving responsibilities, said Robert Drago. Bias avoidance disproportionately affects those who shoulder primary caregiving responsibilities. What Drago termed productive bias avoidance involves minimizing family commitments that interfere with career progress. The most obvious methods are to avoid marriage or delay having children. What Drago termed unproductive bias avoidance involves efforts to deflect attention from the family responsibilities that a person in fact carries. For example, faculty members may decline opportunities to reduce their workload or to take parental leave in order to appear dedicated to their careers.
Joan Williams described family responsibilities discrimination. Also known as the “maternal wall”, this model aims to describe in concrete terms the unrecognized patterns of stereotyping that negatively affect women in academe, to train people to recognize this bias for what it is, and to highlight an important new trend in federal employment lawsuits of which employers must be mindful. Williams discussed the federal employment laws under which employees can sue—and employers can be sued—for family responsibilities discrimination, including Title VII of the Civil Rights Act of 1964, the Pregnancy Discrimination Act, and the Family and Medical Leave Act. In sum, Williams argued for the need to create a new model for spurring institutional change that specifically names and identifies unexamined bias and considers the risk of family responsibilities discrimination lawsuits.
The traditional tactics of increasing female representation on faculties, what Joanne Martin called add women and stir, do not overcome systemic issues that limit women’s opportunities. Policies and practices that appear to apply equally to everyone often have very different effects on men and women because of the differences in their overall social situations.
Martin explained that the widely used 7- to 10-year tenure clock and the requirement that candidates for tenure show early promise, although ostensibly gender-neutral, often create a severe conflict with the biological clock that limits women’s reproductive years, forcing women to choose between taking time out for pregnancy, childbirth, and child care and pursuing fast-track careers. Other requirements of career success such as travel, relocation, and long workdays are much more difficult for people who have major caregiving responsibilities—overwhelmingly women—than for people who do not—usually men.
ETHNIC AND RACIAL MINORITIES
Women’s interest in obtaining science education and pursuing scientific careers varies among ethnic groups, explained Joan Reede. While minority-group women are more likely than white women to major in and earn a PhD in science and engineering, they are rare in academic science; and once their careers have begun, they often face dual negative stereotyping. Because of their extremely small numbers on science faculties, they suffer in an exaggerated way from the problems of isolation, high visibility, unreliable feedback, inauthenticity, lack of role models, and difficulty in obtaining mentoring and camaraderie that afflict many female academic scientists generally. In addition, as members of groups underrepresented on campus, women academic scientists are under great pressure to serve on large numbers of committees.
Like their minority-group women colleagues, pioneering female faculty members—those who are among the first to be hired into a department or who are the only women in their departments—face social isolation and extreme visibility, explained Joanne Martin. They are often viewed as tokens rather than genuine colleagues. Their singular status often results in unreliable feedback and difficulty in being accepted as team members and leaders. The result is that women quit, Martin said, and minority-group women quit more often.
SUMMARY AND SUGGESTIONS
From the panelist presentations and ensuing discussion, several themes emerged, primary among them that male and female careers in science and engineering generally follow different trajectories. Panelists presented data showing that sex differences in cognitive and intellectual abilities do not account for the numerical discrepancies between women and men in faculty positions. Women and minorities, because of their small numbers in faculty and leadership positions, lack the requirements of career success including mentors, camaraderie, networking possibilities, and social support. In addition, pervasive explicit and implicit gender bias—practiced by both men and women, white and minority group members—has played a major role in limiting women’s opportunities and careers. Panelists provided demonstrations and data to show that bias is a complex phenomenon that requires multiple remedies, the first among them an explicit examination of the effects of bias on evaluation. Ostensibly gender-neutral institutional policies often disadvantage women scientists, particularly those targeted at women to accommodate family caregiving responsibilities, because women who take advantage of such programs are seen as less serious than their male colleagues. Women scientists who belong to ethnic and racial minorities face additional issues of stereotyping, isolation, and tokenism.
Panelists proposed a wide range of steps that institutions can take to reduce bias and inequity against women and improve opportunities for them to succeed in academic science careers. On the whole, presentations focused in on how restructuring institutional policies could alleviate problems caused by gender bias, isolation, and caregiving responsibilities. Among the steps proposed were:
Using new metaphors and descriptions to discuss bias, in particular calling bias or stereotyping unexamined places the responsibility on the person who holds or acts on the bias or stereotype.
Training people to see and identify unexamined bias in their own and others’ actions.
Establishing flexible-time policies such as family leave, flex time, part-time tenure, and temporary stoppage of the tenure-clock; and, just as
importantly, an atmosphere that allows faculty members to take advantage of these policies without fearing damage to their careers.
Restructuring hiring and promotion procedures to reduce bias and encourage diversity, particularly the training of search committees, deans, and department chairs to recognize and reduce bias in hiring, evaluation and promotion.
Establishing programs to provide mentoring and support to women and other underrepresented groups.
Changing the context of test-taking to eliminate stereotype threat.
Continued or enhanced funding of research into social and institutional structures and field testing of methods to reduce bias and stereotype threat.
A complete summary of the presentations, including figures and references, is presented in the next section. That is followed by the papers of several of the convocation speakers and the abstracts of the research posters presented at the meeting.
In addition to this workshop report, based on the information presented at the Convocation and other research that the study committee gathers, the committee will issue a consensus report presenting conclusions and recommendations (see http://www.nap.edu/catalog/11741.html). Several convocation participants emphasized that greater workforce diversity will strengthen the American scientific enterprise and that universities and other institutions can do much to improve the opportunities for female and minority scientists to succeed in academic science.