The role of semantic relatedness in producing the revelation effect: A test of the global-matching model

The revelation effect is a robust episodic memory phenomenon where individuals are more likely to report that they recognise an item when it is judged after an interpolated task than when it is not. Westerman and Greene proposed a global-matching model explanation of the effect, according to which the recognition bias results from the interpolated task activating information in memory. It follows from this prediction that the magnitude of the revelation effect should vary linearly with the degree to which revealed and test stimuli are related. The present experiment found that the magnitude of the revelation effect varied monotonically as a function of the degree to which revealed items (words) were semantically related to test items, especially for false alarms, thereby providing empirical support for the global-matching account. The results are discussed in terms of familiarity-based and criterion-based explanations of the revelation effect.