As a woman scientist, my career choice resists gender norms. Even in my relatively female-heavy field of ecology, women publish only about 23 percent of journal articles, so whatever publications I eventually, finally produce will constitute a tiny nudge toward gender equality. My interest in the under-representation of women in the sciences goes beyond my own career, however. I read so many articles about the topic that the leaders of Duke Women in Science and Engineering tasked me with furnishing the group’s Facebook page. I also volunteer somewhat regularly in science education, most often with girls. So, I was surprised to discover that I, objectively speaking, am a sexist.

This exploration of my biases started with my reading an article about implicit biases against black people common to white Americans, demonstrated through Implicit Association Tests. I wanted to know if I share that prejudice, and I do; I am also a racist, mildly inclined to automatically associate blackness with badness, and whiteness with goodness. That I am less racist than most white people I found depressing rather than redeeming. Bad news had not ceased, however; there remained another test, this one on gender roles. Thus I learned that, in terms of implicit bias, I am more of a sexist than a racist. I have a moderate automatic association of women with families and men with careers.

In the weeks that followed, some mild soul-searching ensued. I wondered how I could have developed into such a contradiction, and uncomfortably thought about my own relatively privileged and un-sexist childhood. Whenever someone asks me about my own experience with sexism in school, I describe the only overt incident I recall: a male science teacher wrote in my eighth-grade yearbook, “If we could clone you and make you a male, we could rule the word.” I found this botched compliment more bizarre than discouraging at the time, perhaps because I knew I was generally better at science and school than the boys; everyone had told me so. My teachers were mostly women, and if they praised male students in a somewhat different way than they did female ones, they also fundamentally wished me well and supported my achievements, often gushingly so. Likewise, my parents always encouraged me to do well in school and in science, though I did get into a screaming match with my mother at age fifteen when I offhandedly declared that maybe I would just not have kids. Had I somehow subconsciously internalized the notion that women belong at home through years of such small slights, and never overcome it? As I struggle against society’s expectations for women through my work, am I also fighting myself?

More than I worried about the role of internalized sexism in my own life, I worried about its effects on the girls with whom I volunteer. I had read that teachers’ subconscious sexism could propagate self-doubt about STEM ability in female students. I hope to make the girls with whom I work more confident, about science and about life in general, than I am. Now I considered with some horror the notion that I might have exactly the opposite effect. As I judged a science fair, and then shepherded a group of sixth-grade girls around the Duke Females Excelling More in Math, Engineering, and Science (FEMMES) capstone event, I kept checking myself for discouraging behavior towards the girls and the students of color.

My self-distraction probably only encouraged me towards regrettable absent-mindedness, and I would certainly change some of my behavior if I could. Overall, however, volunteering reassured me. I met so many bright young people enthusiastic about science, of a much greater diversity than I see in the halls of my own school. The students clearly appreciated the events. However tainted, my small contribution to providing these kids with these experiences must do more good than harm, I concluded. Just watching the students reminded me of my own first half of grade school, before black students succeeding in science along with me became unusual, before I had twice as many male classmates as female ones in calculus. I recognized that these children still have a chance, if us grown-ups can just not screw it up. With luck, perhaps these kids are more likely to remember simply that I, a female scientist, existed, than whatever self-doubts I failed to hide from them. I hope my small failures will not contribute to an accrual of discouragement that, like some bioaccumulated toxin, will eventually make these children unable to succeed where they otherwise could have done. I hope the joy of discovery that I have seen in them endures.

In a final twist to this story, as preparation to write this post, I revisited the Implicit Association tests, and found a new one on gender and science. Guess what? I have “a strong association of Female with Science and Male with Liberal Arts,” a result shared by just 1 percent of test-takers. Of course, I promptly searched back through my deliberate anti-sexist behavior at the most recent science fair, newly concerned. Had I perhaps engaged in reverse sexism when I wanted to give only second place to the boy with heavy parental influence, but ultimately better science, than my seemingly more creative, female selection for first place? (I was outvoted anyway; the boy won.)

Then, belatedly, I calmed down and stopped wallowing in self-blame. I realized that this test result, more than any of the others, makes sense. Every week, I read articles about women in science, math, and engineering; I have several news feeds with keywords like “women science.” (About a fifth of the articles returned have titles like, “Science explains what women really want,” but that’s another story.) So, even though most of the PhD-level scientists I know are men, of course the words “women” and “science” go together in my head more easily than “women” and “English” or even “men” and “engineering.” This test most likely reflects my interest in women in science, and not a lack of faith in male scientists.

In other words, perhaps the Implicit Association Tests indicate what I’m used to thinking about. In that case, perhaps I can manage my biases through what I choose to read and do. Watching diverse children experience science might actually make me less biased. Maybe these experiences with the kids are even more mutually beneficial than I realized. I will continue to try to interact with children, and adults too, with more self-awareness about my own flawed thought patterns than I once did. Hopefully, together, we will move towards a fairer future.