Abstract
In virtual environment, avatars are like identity masks, which allow the users to create their own avatars at will. The goal of this study is to investigate how avatar's gender differences influence the participants in stereotypes relevant task performance, based on the Proteus effect to explore whether stereotype threat and stereotype lift occurred in virtual group settings. Participants were randomly assigned to a 2 (participant gender：female vs. male) X 2 (avatar gender：lone-female vs. lone-male) between-participant design. We manipulated avatar’s gender in a group of three with two opposite-gendered avatars (co-actors), putting the gender of the participant avatar in a numerical minority status to cause stereotype threat and stereotype lift situation. The results showed that participants represented by different gender indeed influence wayfinding ability. Particularly, when the participants represented by a female avatar, exhibited a significant deficits in all the wayfinding ability tasks, compared with lone male participants. These results suggest that participant of female avatar associated with negative stereoptype in wayfinding ability, may cause performance impairment by stereotype threat. On the contrary, participant of male avatar associated with positive stereoptype in wayfinding ability, may cause performance boost by stereotype lift. Furthermore, the results also indicate that there are gender differences in specific wayfinding task. Although there is no gender difference in Euclidean-based instructions errors number, but male participants use less time to complete route learning compared to female participants. These results suggest that, male ares adept at using abstract strategy and Eulidean, such as NSEW, when navigating in the virtual environment, and that means they required less time to do the decision-making. On the other hand, male participants are outperformance on survey knowledge scores. These results suggest that, males are adept at using abstract strategy and Eulidean, so that they are good at requiring survey knowledge as well.