Abstract
Stereotype accuracy is a contentious topic. Part of the problem is that typically stereotypes are generic statements whose truth status is unclear due to the fact that they are ill-defined quantitatively. The article focuses on the epistemic aspect of stereotypical beliefs. In the ongoing debate, I side with those who argue against stereotypes being wrong or inaccurate by virtue of definition alone. I propose that, when possible, stereotype accuracy should be assessed in probabilistic terms by inspecting how likely a generic statement is to be true when applied to individual representative of the relevant group. This approach applies equally well to investigating the actual and the perceived accuracy of stereotypes.