Familiarity and recollection in fictitious face recognition: implications for recognition models

Using the paradigm, participants were engaged in a fictitious face recognition experiment in which three between-subject conditions were manipulated by either adding or not adding wrinkles on the faces between the study and the recognition tasks (no change, single change, double change). Our hypothesis was that this procedure would maximize judgments based on familiarity, thus determining whether data provided better fit to the models based on the signal detection theory or to predictions of dual models. In global terms, our results support the signal detection model predictions, although we discuss whether these models better fit tasks based on familiarity or tasks based on recollection.