| Three schema hypotheses (framework, integration, and attention hypotheses) were tested to determine if schemata for faces show the same properties as schemata in other domains. A new theory of levels of schemata for faces was introduced. Experiment 1 showed that superior recognition of famous over unknown faces (as demonstrated in previous work) was due to subjects attaching an episodic tag to the famous person schema and not to the retention of specific episodic information within the schema framework. Experiment 2 showed schema integration effects. Confusion of old and new schema information occurred for famous faces, but not for unknown or inverted faces. Experiment 3 supported the attention hypothesis. Subjects were better at recognizing features that had been presented in a schema inconsistent position rather than a schema consistent position on unknown faces. The results from Experiment 4 contrasted with standard schema findings. Subjects were unable to detect substituted features that were inconsistent with schemata for famous faces, suggesting that an overall Gestalt is used in recognizing famous faces. Overall, the results from these experiments are in agreement with results from other schematic domains. |
