Gender stereotypes are sticky, and even in virtual worlds, they carry over in unexpected ways. According to Motherboard, new research suggests that people carry assumptions about gender with them online—by treating digital women differently from digital men—but that their own behavior in virtual worlds is determined by the gender of their avatar. When a woman plays as a male character, her behavior actually changes.
In social sciences, there's a double-edged concept of “stereotype threat” and “stereotype boost.” People whose social group—man or woman, white or black, straight or queer—is stereotypically considered to act a certain way will actually be affected by that stereotype. So, for example, there's a stereotype that women are bad at math. Put to the test, women who think women are supposed to be bad at math will subsequently do worse on a math exam.
Stereotype boost is the opposite of stereotype threat. Where stereotype threat causes stress and anxiety, stereotype boost—the concept that you're supposed to be better at something—can give an actual boost in performance.
The weird thing is, stereotype threat and stereotype boost kick in just by putting on a digital mask, according to the new research. A woman who is randomly assigned to play as a digital man will get a stereotype boost to her math ability, because her new digital self is no longer subject to the stereotype-imposed anxiety. In the same way, a man playing as a woman will suddenly perform worse. Man or woman in real life, the gender of the players' avatars is what affected their math skills.
As the researchers explain in their study, donning a digital avatar can do all sorts of things to our behavior:
Research on the Proteus effect—named after the Greek god known for the ability to take on many different physical forms—has demonstrated that people conform to the stereotypes associated with the characteristics of their virtual self-representations, both behaviorally and cognitively. For example, those who were randomly assigned an attractive avatar acted more intimately and socially when interacting with a confederate than did those who were randomly represented by an unattractive avatar.
...More specific to the effects of gendered avatars, research showed that both females and males who were arbitrarily represented by a female avatar were more likely to conform to female-typed language norms (e.g., reference to emotions, use of apologies) when compared to those represented by a male avatar. Furthermore, analyses on game players' behaviors in multiplayer online gaming environments revealed that male avatar players were more likely to engage in killing, while players using female avatars were more likely to engage in healing activities. These patterns were observed regardless of the players' actual gender, indicating that people tend to conform to stereotypic expectations associated with the virtual gender identify of their avatars.