Rehearsal and recognition of human faces.

The question of whether or not photographs of human faces may be rehearsed was addressed in two experiments. The faces varied in perspective from the presentation to the recognition phases of the experiments. In Experiment 1, two instructional sets which favored different rehearsal modes (i.e., visual image versus verbal description) and two durations of the rehearsal period were compared. In Experiment 2, the continued observation of pictures was compared to the rehearsal of images of those pictures for equivalent periods of time. The results from both experiments demonstrated that rehearsal increased both recognition accuracy and confidence, and decreased response latency. Thus, previous research which had neglected the use of appropriate control groups had consistently underestimated the magnitude of rehearsal effects upon recognition accuracy. Further, compared to a control condition, the continued observation of a picture and the rehearsal of an image of that picture in Experiment 2 produced equivalent gains in recognition accuracy. The results were interpreted in terms of the schematic nature of visual images and the facilitatory effects of rehearsal for the generation of retrieval of such a schematic.