by Gabriel Gasque, PhD

 We are biased. When evaluating someone’s competence in a professional setting, we tend to consider other factors beyond qualifications and achievements. One particularly troubling example is gender bias. We still see women as less competent for certain professional tasks, and there is evidence to back up this claim. Smart research done by Corinne Moss-Racusin and colleagues at Yale University elegantly supports the existence of gender bias in science.

The experiment was elegant due to its simplicity. The authors objectively tested faculty from research-intensive institutions for their gender bias. The scientists conducting the study made up a resume and asked 127 faculty scientists to rate the qualifications of the fictional student (the volunteer scientists did not know the student was fictional) who was applying for a research assistant position. A female name (Jennifer) was assigned to half of the faculty participants, and to the other half, a male name (John) was assigned. Otherwise, the CVs were identical. Probably not surprisingly, the male student was rated higher for competence and hireability, even by female faculty. In addition, faculty would be willing to give the male student more mentoring and a higher starting salary.

Know thy bias F1

I wrote “probably not surprisingly” because the gender gap in Science and Engineering (S&E) fields is still alarmingly large, suggesting discrimination against women. By 2010, only 21% of full-time faculty positions were held by female scientists. But correlations do not imply causality, and the gap between male and female scientists in faculty positions has been attributed to factors other than discrimination, including a lack of women who desire advanced scientific positions or who are as equally skilled as male applicants. Harvard’s ex-President Lawrence Summers endorsed both arguments in his remarks at an S&E conference several years ago.

The “lack of skills” argument has been built, in part, upon the fact that over the past 20 years, more male than female high school students have performed at the top 5% in math tests.  The reasons for this difference in performance has been extremely difficult to pin down. One tested hypothesis is that males and females respond differently to testing, and thus math tests results may not be predictors of actual math skills.

So why was ‘Jennifer’ rated less highly than ‘John’ by the Yale faculty who participated in this study? Most scientists consciously strive to be objective. We are trained to consider hard facts, and to move beyond emotions when making judgments and statements.  But research has shown that people who highly value objectivity and fairness are particularly prone to express biases. This probably happens because “objective” people overestimate their invulnerability to hidden bias and fail to monitor its influence on behavior [1, 2].

In this context, I felt a responsibility to test myself in an Implicit Association Test (IAT) for gender-science association. IAT is a computer test that measures the strength of association between classes of concepts (like male-female vs. career-family) by observing response times in categorization tasks. The idea is that it is easier to make a decision when concepts that we think belong together share the same response key. For example, faster responses for the {male career/female family} task compared to the {male family/female career} task indicate a stronger association of male and career over female and career.

Know thy bias F2

IAT is a robust test; it displays internal consistency, independence towards familiarity of the subjects with the method, and resistance to faking (reviewed in ref. 3). More importantly, IAT is predictive of behavior. In socially sensitive topics, IAT is a better predictor than self-reported attitudes [3].

So, I tested myself, and I think that everybody interested in being egalitarian should do it too. I tried two different tests, several times: (1) male, female vs. career, family and (2) male, female vs. science, liberal arts.

I found some variability in my results, which is not unexpected. The average of my tests could be summarized as a very slight association of male with career and female with family, and male with science and female with liberal arts.

Would I have rated Jennifer lower than John in Moss-Racusin’s experiment? Probably yes. As shocking as this might sound, I am not discouraged. Being aware of our hidden biases is the starting point to gaining control over them and to minimize their influence on our behavior. If you obtain an implicit preference you don’t like, well, you might choose to remember that IAT is not perfectly accurate, and move on. Alternatively, you could accept the fact that you might actually show bias and then try to undo it. For example, avoid situations that reinforce the stereotypes.

As a final remark, I would love to read similar research about underrepresented minorities in S&E. African Americans and Hispanics, who represent 30% of the American people, account for only 5% of the work force in S&E. In addition, African Americans are 10 percent less likely to be awarded NIH RO1 funds compared to whites. Could these statistics reflect another implicit bias that we need to acknowledge and fix? I think they do. How about you?

References

  1. Monin, B. & Miller, D.T. Moral credentials and the expression of prejudice. J. Pers. Soc. Psychol. 81, 33-43 (2001).
  2. Ulmann, E.L. & Coehn, G.L. ‘‘I think it, therefore it’s true’’: Effects of self-perceived objectivity on hiring discrimination. Organ. Behav. Hum. Decis. Process. 104, 207–223 (2001).
  3. Greenwald, A.G., Poehlman, T.A., Uhlmann, E.L. & Banaji, M.R. Understanding and using the Implicit Association Test: III. Meta-analysis of predictive validity. J Pers Soc Psychol 97, 17-41 (2009).

About the Author

IMG_2377[2]

Dr. Gabriel Gasque was born and raised in Mexico City. He has been a little of a globetrotter and has lived and studied in Mexico, Denmark, and Chile, until he found home in New York City. He obtained his Ph.D. from the National Autonomous University of Mexico, and has postdoctoral trainings from Columbia University and The Rockefeller University. When bench science stopped being enough to satiate his desires to understand the natural world, he turned into publishing and science communication. He loves cooking and eating, and is a devotee of yoga and introspection.

 

Join the conversation!