Abstract
The existence and persistence of stereotypes on gender and science, as well as their influence on attitudes and behaviors, have been largely studied worldwide. Current measures of gender-science stereotypes are mainly descriptive and do not ask respondents their opinions about the perceived cause(s) of these gender differences. However, empirical evidence suggests that gender bias attribution, i.e., the difference in the causes to which gender differences are attributed, has heterogeneous consequences. Here, it was exploited the fact that Project Implicit includes both instruments of gender-science stereotypes and gender bias attribution to test whether and to what extent two components of gender bias attribution, i.e., causes attributed to personal characteristics and those attributed to social/contextual factors, were associated with the endorsement of implicit and explicit gender-science stereotypes. Both an exploratory and confirmatory factor analysis tested whether the instrument on gender bias attribution in Project Implicit could be decomposed into two components, while an SEM (Structural Equation Modeling) analysis tested the hypothesized association. The factor analysis confirmed that bias attribution should be decomposed into two distinct components, “internal factors” and “external factors”. Finally, the association between these two components and implicit and explicit gender stereotypes varied depending on the participant’s gender. Explicit gender stereotypes’ endorsement was positively associated with the external component in the case of women and with the internal component in the case of men. Conversely, the association between attribution and implicit gender stereotypes was null.