Haven’t I seen you before? Prior familiarity can impair or enhance face recognition memory

Why do we remember familiar faces better than unfamiliar ones? Recognition memory for faces with which we have prior familiarity is superior to recognition of unfamiliar faces, even under very controlled circumstances. Prior familiarity with a face seems to substantively change the way we encode and recognize later instances of that face. However, most prior work has drawn comparisons between completely unfamiliar faces and faces with very strong pre-existing memory representations that contain semantic, person-related information such as a name, occupation, or personality traits. This approach ignores the varying levels and types of familiarity we experience with different faces. In addition, that prior research has almost entirely ignored the potential confounding effects of familiarity on response bias. To draw conclusions about the effects of prior familiarity for important scenarios, like eyewitness identification, a clearer understanding of these effects is needed. This dissertation examines the effects of varying levels of prior familiarity and conceptual knowledge on face recognition memory in nine experiments. In all experiments, prior familiarity was manipulated using a 3-phase (familiarization, study, and recognition test) procedure. Participants were presented with static face pictures and were asked to make various conceptual judgments (e.g., How friendly do you think this person is?), tried to learn their names, or passively viewed them during the familiarization phase. Familiarized and novel (unfamiliar) faces were then studied and tested. Across experiments, an increase in prior familiarity led to a more liberal response bias both when familiarity was gained through conceptual processing or through passive exposures. Discriminability, on the other hand, was enhanced by prior familiarity only when the level of familiarity was high and involved conceptual processing (Experiments 1-3). Familiarity engendered by passive exposures affected response bias equivalently, but reduced discriminability both in a standard Yes/No recognition test (Experiment 4) and in a lineup identification task (Experiment 5). Higher levels of prior familiarity accompanied by conceptual knowledge led to an increase in discriminability, and a shift towards a more liberal response bias both when the images of each face were identical (Experiments 1-5) and when they varied across exposures (Experiment 6). Experiments 7 and 8 examined the effect of study time and context reinstatement, respectively, on recognition memory performance for familiar and unfamiliar faces. Longer study time and context reinstatement were equally beneficial for the recognition of both types of faces. Finally, Experiment 9 examined the effect of prior familiarity on the encoding of the surrounding local context (a background scene), and showed that prior familiarity with faces facilitates the binding of faces with their context. This benefit was observed both for faces familiarized through learning of conceptual information and through passive exposures. These findings suggest that prior familiarity, regardless of the presence of conceptual, person-related information, leads to higher-quality encoding. The detrimental effect of prior familiarity emerges at test despite this particular encoding advantage. Overall, the current set of experiments suggest that prior familiarity has facilitative effects on discriminability when it is at relatively higher levels or involves conceptual knowledge. The lack of such conceptual representations, on the other hand, hinders the ability to localize the source of familiarity, and impairs face recognition memory. Implications of these findings for eyewitness identifications of familiar suspects are discussed.LimitedAuthor requested closed access (OA after 2yrs) in Vireo ETD syste