Facial recognition

Two experiments investigated the possible properties involved in facial recognition. In the first experiment, the overall ratings of faces were compared to the ratings of their component features. Certain features were found predictive thus indicating the use of features in facial recognition. It was suggested that the nature of the task could have dictated the strategy used. In the second experiment, similarity ratings were compared for faces seen in upright and inverted presentations with two time variations (5 and 8 seconds). There appeared some characteristics common to all conditions, but not enough across the two variables to produce significant correlations when both time and presentation mode changed.