Cognitive Processing of Scrambled Faces: Effects of Instructions and Task

The present study tests Rakover and Cahlon's (2013) face-checking model, which grades 7 regular and scrambled faces on a scale of similarity to an upright regular face, by predicting the results of 2 experiments in upright and inverted orientations: Experiment 1, which uses the interest choice task (to choose from a pair of faces the one most interesting), and Experiment 2, which uses the old/new recognition task. The main results of these 2 experiments show that in comparison to Rakover and Cahlon's (2013) findings, the face-checking model preserves its ability to predict satisfactorily the order of the 7 faces in the 2 experiments especially in the upright orientation; however, the model's success in making accurate point predictions is reduced significantly in both the upright and the inverted orientations.