Study reveals implicit gender bias in recommendation letters

AddThis

Photo Credit: Will Byargeon | Daily Texan Staff

When UT students apply for jobs, graduate programs or professional schools, recommendation letters can play an integral role. What professors and students may not realize, however, is that these letters might be unconsciously biased by gender, according to a recent study from Columbia University. 

The comprehensive study showed that women applying for geosciences fellowships are less likely to receive outstanding letters of recommendation in comparison to men, regardless of what region the letters came from or the recommender’s gender.

Kuheli Dutt, lead author of the paper and assistant director for academic affairs & diversity at Columbia University’s Lamont-Doherty Earth Observatory, said she wanted to explore why only 10 percent of geoscience professorships are held by women when they hold 40 percent of doctoral degrees.  

The five-year study analyzed the tone and length of over 1,000 recommendation letters written for geosciences postdoctoral fellowships. The study had a large data set of letters from 54 countries and controlled for regional differences. 

The letters were stripped of any identifying factors, such as gender or race, then classified into three categories: excellent, good and doubtful. 

Letters classified as “excellent” clearly described the applicant as outstanding, using descriptions such as “trailblazer” or “brilliant scientist and role model,” whereas “good” letters consisted of weaker descriptors such as having a “thorough understanding of the subject.” “Doubtful” letters questioned the applicant’s scientific and leadership abilities. 

Results of the studies revealed that female applicants were only half as likely to receive “excellent” letters compared to male applicants, a consequence Dutt attributes to implicit, or unconscious, gender bias.

“Our study uncovers what appears to be a very real problem that is consistent with implicit bias,” Dutt said. “Given the way society views men, men are more likely to be described as ‘confident’ and ‘dynamic’ whereas women are more likely to be described as ‘mature’ and ‘caring.’”

According to Dutt, while these labels might be applied with good intentions, adjectives that tout leadership and innovation are more favorably viewed by businesses. Companies promote people to lead and spearhead projects, whereas people described as “mature” and “team-builders” are not selected as readily for these opportunities, Dutt said. 

Arthur Markman, UT professor in the Department of Psychology, said that even though this study focused on geosciences fellowships, implicit bias contributes to the leaky pipeline of other STEM fields, the legal professions and industry jobs, to name a few. 

“The leaky pipeline phenomenon shows that although women have degrees in STEM fields, as we move up the higher education ladder, promotions and fellowships and tenure faculty prominence of women drop off significantly,” Markman said. “This is also prevalent in the workforce.”

According to Markman, the stereotypes associated with gender can also factor into unconscious bias. To combat this issue, decision-makers must acknowledge that a problem exists and be more upfront about criteria when choosing applicants, Markman said.

“Despite 40 years of attempts to create gender equity [in terms of hiring practices and equal pay] on college campuses and in workplaces, there are still gaps,” Markman said. “When you combine the real world data of persistent differences as a result of race, gender and ethnicity, with data from well-controlled studies, it becomes harder and harder to argue that there’s something else going on other than these biases that are creeping into the system.”

Future steps include studying implicit bias for non-binary genders, as well as advancing studies focused on the intersectionality of minorities and gender,” Dutt said. 

According to Yael Niv, associate professor at Princeton University’s Neuroscience Institute, compounding effects of race and gender can create obstacles for minority groups. For example, Asian women in science are expected to be demure, Niv said. 

“So when an Asian woman speaks up at a [research] conference, people are like, ‘Wow, she’s so aggressive’ or ‘Wow, she’s so strong-willed and unrelenting,’ whereas a guy could say the same thing and people would just say, ‘Oh, what a thoughtful comment,’” Niv said.

According to Niv, getting these problems out in the open is crucial for change. 

“It really isn’t about judging people, it’s about how to fix the situation,” she said. “Implicit biases are unintentional, but we must intentionally fix them.”