Supplementary material to “The Science of Diversity”

Jennifer D. Laird and Robin E. Bell, Lamont-Doherty Earth Observatory, Columbia University, Palisades, New York; Geraldine Downey, Columbia University, New York; Stephanie Pfirman, Barnard College, New York.

Citation:
Laird, J. D., R. E. Bell, G. Downey, and S. Pfirman (2007), The science of diversity, Eos Trans. AGU, 88(20), 220. [Full Article (pdf)]

Academia has clearly come a long way since the 1960s when women PhD students were considered “bad investments” because they interrupted their careers for marriage and children (NAS, 2006). As of 2004, women earned over one-third of the PhDs granted in the Earth sciences, a significant increase from 1974 when they earned less than 10% (NSF, 2004). While the number of women and minorities in science is increasing, the rate of increase in the upper ranks of academia is much slower than the rate of increase in high school and undergraduate science programs. Institutional leaders across the country are faced with a diversity dilemma: at every academic milestone, the proportion of women in science declines, even in fields where there has long been gender parity in the pipeline (NAS, 2006). Both universities and the federal government are responding to these bottlenecks by investing in diversity initiatives. The National Science Foundation’s ADVANCE program seeks to increase the participation of women in science and engineering by using behavioral and social science research to provide fundamental insights about individual and organizational behavior. By employing a research-driven approach, ADVANCE is systematically addressing the diversity dilemma in science and engineering.

On November 17, 2006, the ADVANCE program at Columbia University’s Earth Institute convened a symposium on “The Science of Diversity” in partnership with the Provost’s Task Force on Diversity in Science and Engineering. The objectives of the symposium were to identify factors that lead to bias and exclusion and to provide strategies for universities to create more inclusive environments. The symposium brought together scholars in psychology, business, and law to address four topics:

  1. Cognitive bias, stereotype threat, judgment, and decision-making,
  2. The effects of subtle environmental changes on individuals,
  3. Techniques for structuring inclusive environments, and
  4. Strategies for institutions to translate research into action.

Stereotypes can have a strong effect on scholarly acceptance and recognition. Madeline Heilman, professor of psychology at New York University, described how gender stereotypes affect hiring practices. Stereotypes, such as “men are agentic; they get things done” and “women are communal; they care for others” are not inherently bad perceptions. The potential for gender bias arises when the perceived criteria for the most prestigious positions are closely associated with the male stereotype. When the male stereotype is pervasive, there tends to be an implicit expectation that women will perform poorly in the position (Lyness and Heilman, 2006). Women may also engage in self-limiting behavior when they try to ensure a fit between their behavior and others’ expectations (e.g., in a negotiation situation). Heilman recommended specific strategies for counteracting biased behavior, such as removing ambiguity in evaluation processes and increasing accountability in decision-making.

Group hires have also proven to be an effective method for counteracting unconscious bias in hiring decisions. Using experimental laboratory studies, Valerie Purdie-Vaughns, professor of social psychology at Yale, has found that people will choose more diverse job candidates when hiring a group rather than hiring a series of individuals. When the selection process is focused on individuals, the prototypical candidates are framed as scarce resources. When groups are hired, the demographics of the hired cohort are more likely to reflect the applicant pool. In addition, Purdie-Vaughns has found that hiring in groups can reduce the anxiety that people often feel when they are working under a mandate to increase diversity.

Stereotypes are rarely overt; they are usually expressed through subtle cues in the environment. Tufts psychologist Nalini Ambady and others have shown that in environments where women are asked to identify their gender, women perform worse than men on math tests (see, e.g. Steele and Ambady, 2006; Dar-Nimrod and Heine, 2006). This effect, known as a stereotype threat, is only evident in environments where women associate their performance with their gender and not their individual ability. It is possible to counteract stereotype threat by changing the environment. At the workshop Stanford psychologist Carol Dweck described how perceptions of performance ability can impact susceptibility to stereotypes threat. Those who believe their ability to be fixed are much more susceptible to stereotype threat than those who perceive their ability to be malleable (Aronson et al., 2002). Through a series of neurophysiological experiments, Jennifer Mangels, from Columbia University, has recently shown how subtle cues in performance-oriented environments can reduce the impact of stereotype threat. When teachers present material emphasizing the malleability of intelligence, the performance of all students increases, with the largest increase observed in the minority group.

The applications of this research to diversity efforts in academia are not straightforward. Purdie-Vaughns’ research indicates a department may be more likely to hire diverse candidates for faculty positions if it is given the opportunity to hire in groups, but group hires may be unfeasible for some departments and universities. Mentoring programs could be developed based on Dweck and Mangels’ research, but the curriculum changes used by teachers in their experiments would have to be modified to be relevant in a professional setting. Faculty mentors may not feel comfortable addressing an issue as abstract as the malleability of intelligence with their junior colleagues. By the time scientists become junior faculty members, it may be too late to change their perceptions about their innate abilities.

To build a diverse academic environment where women and minorities will thrive, leadership from the top is crucial. Susan Sturm, a professor at Columbia’s Law School, outlined the requirements for achieving institutional change at universities. Change is difficult to implement when institutional authority is fragmented and when any single individual has limited power. Sturm’s research has focused on the role of organizational catalysts, individuals who serve as information entrepreneurs at pivot points within an institution (Sturm, 2006). To successfully implement interventions like cluster hiring or research-driven mentoring, institutional roles have to be created that place people with knowledge and credibility in positions that can influence processes where biases operate.

The symposium featured a number of organizational catalysts. Joan Girgus from Princeton and Jean Howard from Columbia described their efforts to influence change at their institutions. As the Special Assistant to the Dean and a member of the Princeton psychology faculty since 1977, Girgus has the knowledge and the credibility required to introduce institutional interventions ranging from university-supported emergency care for children to a policy of automatically stopping the tenure clock when any junior faculty member, man or woman, has a child. Howard, an experienced administrator and well-respected scholar, is also in a position to influence change as the Vice Provost for Diversity Initiatives at Columbia. In her first year as Vice Provost, she developed a $15 million hiring fund to support the recruitment and retention of women and minority faculty. The ADVANCE program at Columbia has worked with Howard to bring the behavioral and social science research on decision-making and stereotype threat to search committees and department chairs.

University departments can be highly autonomous and resistant to change. Institutional leaders need to have support for diversity objectives a priori to ensure long-lasting success. Concrete institutional interventions that emerged from the symposium include:

  1. Removing ambiguity from evaluation processes,
  2. Increasing accountability in decision-making,
  3. Hiring in groups whenever possible, and
  4. Developing mentoring programs that convey to young women and minorities that their ability is not fixed but rather a product of engagement and interest.

By understanding and applying the science behind diversity, academic communities can begin to undertake the cultural shift that will be required to make science a more inclusive institution.

Acknowledgements

We thank the National Science Foundation for their support of this meeting through the ADVANCE award to the Earth Institute at Columbia.

References