Skip to main content

How Implicit Bias and Lack of Diversity Undermine Science

The first step toward fixing the culture of STEM is recognizing that it’s broken

This article was published in Scientific American’s former blog network and reflects the views of the author, not necessarily those of Scientific American


When neuroscientist Ben Barres delivered his first seminar, an audience member praised him, commenting that Ben’s work was much better than that of his sister, Barbara Barres. The irony? Ben Barres (now deceased), a transgender scientist was Barbara Barres before he transitioned to male. When New York Times columnist Brent Staples was a graduate student in Chicago’s Hyde Park, he found that white people on the street perceived him, an African American, as a threat to their safety. They were visibly tense around him, clutched their purses and sometimes even crossed the street to avoid him. But when he started whistling tunes from classical music, people suddenly weren’t afraid of him anymore—they relaxed and some even smiled at him.

Implicit bias runs far deeper than we realize. A riddle used at implicit bias trainings goes like this: A father and his son are in a terrible car crash. The father dies at the scene. His son, in critical condition, is rushed to the hospital; he’s in the operating room, about to go under the knife. The surgeon says, “I can’t operate on this boy—he’s my son!”

The audience is then asked how that’s possible. Responses include several scenarios: two gay fathers; one biological and one adopted father; one father and one priest (religious father); all of which are possible. However, an obvious answer that most people miss: the surgeon is the boy’s mother. Whether we like it or not, we are conditioned to associating surgeons with being male.


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


Similarly, we often associate scientists with being male, especially since STEM fields are usually male-dominated. The National Academies of Sciences, Engineering, and Medicine (NASEM) report on sexual harassment highlighted the staggering rates of sexual harassment in academic STEM fields. Academia has the dubious distinction of having the second highest rate of sexual harassment in the workplace (the military having the highest). Several reasons come to mind: one, the pervasive nature of implicit bias; two, male domination and a lack of diversity; three, the nature of scientific inquiry; and four, the overall culture that enables harassment.

Numerous studies have documented implicit gender bias: in recommendation letters, CVs, invited speakers, academic reviewers, teaching evaluations ... the list is endless. The overarching message is clear: women’s contributions in STEM are not valued as highly as men’s. Additionally, such biases extend to racial/ethnic minority and LGBTQ+ individuals. Since their contributions are not valued as highly, they are less likely to be respected, and given the power dynamics and structural hierarchies in the STEM workplace, it is no surprise that they face marginalization and harassment.

In general, the less diverse or inclusive a field, the heavier the reliance on stereotypes and implicit biases towards underrepresented groups (be it women, racial minorities or LGBTQ+), the greater the harassment they face (sexual or otherwise) and the more likely they are to leave STEM fields. It then takes a movement like #MeToo to expose, to the horror of many, just how widespread and pervasive the problem is. Take, for example, the sexual harassment survey that showed that a whopping 67 percent of women had been harassed during fieldwork.

The subject matter of the physical sciences is disconnected from human behavior, so the human being is situated outside of the subject. Contrast this with the social sciences, where the subject matter pertains to human behavior, and the human being is situated within the subject. While this is appropriate for science, it leads to not placing much importance on human behavior in STEM fields. Further, there is a perceived sacredness of quantitative data, with a tendency to dismiss human experiences and narratives as “merely anecdotal.” This promotes a culture where good science (and everything that comes with it—funding, publications, seniority, tenure) is much more prized than good behavior. When we combine all these factors, we have patterns of bias and harassment deeply entrenched within STEM culture. While harassment is not unique to STEM, the culture in these fields enables harassment. 

A couple of years ago, I did a study where my colleagues and I explored for gender differences in recommendation letters. Our results were exciting and worth publishing. However, I knew that my being a woman of color, an administrator (as opposed to faculty) and without a STEM background (despite my PhD in the social sciences) would potentially subject me to bias in the review process. The irony did not escape me: my worrying about bias when submitting a paper pertaining to bias, especially when male STEM faculty have been found to be skeptical of gender bias research. Not just that: while other fields had been the subject of similar studies this was the first the geosciences, so there would likely be extra scrutiny. I was relieved to find that Nature Geoscience offered a double-blind option in the review process. It became my shield of protection.

We all have biases, regardless of our gender, race, age, sexual orientation, academic background, etc. Affinity bias, or the tendency to prefer people like ourselves, is extremely powerful and exists within all of us. So if a certain group dominates a discipline, the new hires, future leaders, and role models in that discipline will also likely belong to that group and will continue to perpetuate the existing culture.

In popular culture, the reality TV show Survivor groups contestants into different categories: “brains vs. brawn vs. beauty” or “heroes vs. healers” or “David vs. Goliath.” People are grouped with others like themselves. Even though these groups exist only until the “merge,” i.e., when it becomes one big group, the affinity bias is so strong that people continue to identify with those labels, which guides their behavior throughout the game. And since they operate under stress, their biases are further intensified. Until recently, most contestants (and therefore most winners) were white, and minorities were usually voted out early in the game—not due to conscious racial bias but rather to the need to fit in with similar people to move ahead in the game.

How do we reduce our individual biases? Here are some simple tips to start with. When you write a recommendation letter, first identify the traits that you think are necessary for the job (e.g., creativity, analytical skills, leadership skills), and then look at the candidate’s CV/portfolio for any evidence of those skills. This will help reduce the usage of gendered or racial language. Whoever the candidate, ask yourself: what is the best thing you’ve written about that person, and does it adequately reflect their skills? Be mindful that men are more likely to be hired for their potential (i.e., credit for something they haven’t yet done), and women for their actual accomplishments. Avoid qualifiers—there is no need to say, “For an Asian, his English is good”; you can just say, “His English is good.” Women’s personalities are often described in communal terms (e.g., caring, nurturing); you can counteract that with terms like confident, dynamic, highly motivated, etc.

Most people aren’t fully aware of their biases, and if you view yourself as completely unbiased, then you are a part of the problem. STEM culture is in desperate need of repair, and even though institutions are trying to hire underrepresented groups, at the end of the day if those groups are harassed, they will continue to leave STEM fields. And we will continue to read alarming accounts of their experiences in STEM. Acknowledging implicit bias and promoting diversity and inclusion doesn’t only enable better science; it is a crucial first step for fixing a broken STEM culture.

Kuheli Dutt, PhD, is the assistant director for academic affairs and diversity at Lamont-Doherty Earth Observatory, Columbia University, and serves as Lamont's diversity officer. Kuheli leads diversity and inclusion efforts at Lamont, and her research interests lie in implicit bias, gender and STEM diversity. Her publications include Women in the Geosciences, published by the American Geophysical Union, and "Gender Differences in Recommendation Letters for Postdoctoral Fellowships in Geoscience," published in Nature Geoscience. For more information visit https://diversity.ldeo.columbia.edu.

More by Kuheli Dutt