I have a bias against women in science. Please don't hold this against me. I am a woman scientist, mentor and advocate for women in science, and an associate dean in my school's Office of Diversity, with a budding field biologist as a daughter. Yet my performance on the Implicit Association Test (https://implicit.harvard.edu/implicit/demo), which measures unconscious associations between concepts, revealed that I have a tendency to associate men with science and career, and women with liberal arts and family. I didn't even need to wait for my score; I could feel that my responses were slower and that I made more mistakes when I had to group science words such as 'astronomy' with female words such as 'wife' rather than male words such as 'uncle'.

Neurobiologist Jennifer Raymond (front) and her colleagues in their lab at Stanford University in California. Credit: CODY PICKENS

The results from hundreds of thousands of people indicate that I am not an outlier — 70% of men and women across 34 countries view science as more male than female1.

WOMEN IN SCIENCE The gender gap and how to close it nature.com/women

Gender bias is not just a problem in science. A host of studies shows that people tend to rate women as less competent than men across many domains, from musical abilities to leadership2, and that many individuals hold biases about competency on the basis of other irrelevant attributes, such as skin colour, body weight, religion, sexual orientation and parental status.

Such biases have important consequences in the workplace. One study showed that mothers are 79% less likely to be hired and are offered US$11,000 less salary than women with no children3. By contrast, the same study shows that parenthood confers an advantage to men in the workplace.

A 2012 study by Jo Handelsman of Yale University in New Haven, Connecticut, and her colleagues shook the scientific community by reporting that science faculty members have a pervasive bias against female scientists4. This prevents us from doing our job of promoting the best scientists, and society is paying a price in terms of the advancement of science.

There is now sufficient evidence to move us beyond the denial phase of dealing with gender bias. Yet in talking to colleagues around the world, I find continued resistance to the idea that scientists, who take pride in being rational and objective, could be influenced by bias. One colleague was convinced that gender bias could affect the hiring of a lab manager, but he still doubted that it would affect a faculty-level hiring decision or the evaluation of a manuscript, even though the evidence suggests otherwise5. And I have seen junior colleagues shake their heads disapprovingly at the gender bias of older science faculty members, yet resist the idea that their generation might also have such bias.

Unfortunately, young people are not immune to gender bias. Many studies have been conducted on college-age subjects, and gender bias has even been reported in preschool children6. I tried to protect my own children from gender bias by doing things such as changing the gender of the characters in the children's books I read to them to reverse gender stereotypes, and using the feminine pronoun wherever possible — “Look at the elephant; she is so strong.” Despite these efforts, my daughter had a bias against women in leadership positions by the age of three. One day in the park, she announced, “I am the captain; I'm a girl captain,” suggesting that she knew she had to violate a gender stereotype to assume that leadership position. And although she has a scientist mum who runs a lab full of women, when my daughter took the implicit association test at age 8, it revealed a bias against women in science. My presence as a role model and other efforts at countering gender stereotypes were not enough to overcome the powerful cultural transmission of bias. Thus, it seems unlikely that unconscious gender bias will be eradicated any time soon, and the best we can do in the near term is to suppress its symptoms.

If we are vigilant, we can reduce the influence of bias on our decisions. Unconscious biases are mental habits that tend to dominate our gut reactions, but we also have more-rational decision processes, which compete with our biases for control of behaviour. Just as one can overcome physical habits such as biting one's fingernails or saying 'umm' when one speaks, one can suppress undesirable mental habits such as gender bias through deliberate, conscious strategies (see 'Ways to conquer gender assumptions'). By enabling more women to succeed, despite the existence of unconscious bias, this will gradually eliminate the stereotype of the successful scientist as male, which is the root of gender bias.

However, if left unrecognized and unchecked, bias can commandeer both our behaviour and our rational thought processes. Our brains are skilful at creating seemingly rational justifications for our behaviour, even when it is driven by bias. People who had to rate two 'applicants' for police chief — one who had more education and the other who had more experience — always chose the man over the woman, but justified their choice as arising from the value they placed on either education or experience, whichever factor was assigned to the man7.

Denial that bias exists gives it more power. I am not proud of my unconscious bias against women in science. However, I know that I must first recognize my own bias to overcome it with deliberate practices that suppress its effects. I urge you to join me.