Research in psychology indicates that individuals often make inferences regarding unknown individual qualities based on potentially irrelevant (but socially observable) information. This paper explores occupational choices when individuals receive imprecise signals regarding ability and use the observable characteristics of previously successful individuals to infer own ability. Individuals who fail to observe successful predecessors of their same type may underestimate their potential for success in the occupation. We discuss the role of these biases in light of the literature on affirmative action and firm incentives.