Creating fair lineups for suspects with distinctive features

In their descriptions, eyewitnesses often refer to a culprit's distinctive facial features. However, in a police lineup, selecting the only member with the described distinctive feature is unfair to the suspect and provides the police with little further information. For fair and informative lineups, the distinctive feature should be either replicated across foils or concealed on the target. In the present experiments, replication produced more correct identifications in target-present lineups--without increasing the incorrect identification of foils in target-absent lineups--than did concealment. This pattern, and only this pattern, is predicted by the hybrid-similarity model of recognition.