Science is all about accounting for and eliminating biases, which skew results and mask the real effects being tested. Biases can be physical (friction, gravity or wind resistance) or psychological (a patient trying a new wonder drug might over-report her rate of recovery). Often, they're problems with instrument design or technology. This is why well-designed experiments, such as double-blind studies, are so important. But biases can only be eliminated if they're known, and as Arielle Duhaime-Ross writes at the Verge, a huge swath of science has been plagued by a systematic bias that we're only now becoming aware of.
Duhaime-Ross reports on a new study, which found that mice are scared of men. When a male researcher works with a mouse, the mouse's body courses with stress hormones. This doesn't happen when a woman scientist is doing the work. The difference in how mice respond to male and female researchers could potentially skew everything from behavioral studies to cell research.
It's not so much that mice are scared of male researchers as it is that mice are scared of male mammals. A whiff of testosterone from any male mammal is enough to trigger this fear, says Jef Akst for The Scientist. “In all likelihood, mice just haven't developed a way to discriminate between the smell of a male mouse and the smell of other male mammals, so men also elicit a fear response,” says Duhaime-Ross.
A huge chunk of modern biological science is done using mice, and the possibility of such a widespread, unacknowledged systematic bias is troubling. That this source of bias wasn't found until now, however, isn't so surprising. Though women now outnumber men in earning biology degrees, that's a fairly recent trend. In the 1960s, women represented roughly a quarter of professional biologists. It's hard to notice the different reactions to men and women if there aren't any women in the room.
Now that scientists know about the gender bias on mouse research, they can figure out how much of an effect it has on mouse behavior and physiology. Theoretically they can control for it moving forward, and maybe even correct for it in previous research.