LAUSR

Unfamiliar-face matching requires an identity comparison of two simultaneously presented faces that are unknown to the viewer. This can be a difficult task, even for police and security professionals who perform such comparisons routinely. This study examined whether the provision of example face pairs, presente on either side of a target face pair and clearly labelled as indentity matches and mismatches, improves matching accuracy. Examples aided performance at the group level, but analysis of individual differences indicates that this arises from improvement in lower-performing observers, who were initially the least accurate. This effect generalised to previously unseen faces from the stimulus set from which target and example pairs were drawn, but not to face pairs from a new stimulus set. We suggest that examples aid performance by providing criteria to distinguish identity-matches from mismatches, which observers would otherwise have to deduce by their own judgement during participation.