On the relationship between repetition priming and recognition memory: Insights from a computational model

A single-system model of repetition priming and recognition memory is presented, which is conceptually similar to signal-detection theory. Key assumptions of the model are (a) that the same memory source contributes to both priming and recognition performance and (b) that variance of the noise associated with priming measures is greater than recognition. To test the model, four experiments were conducted examining the effects of a manipulation of attention at study on priming and recognition performance. The model predicted that (1) effects of attention will be observed on priming and recognition, albeit larger for recognition, (2) the magnitude of priming will not exceed recognition in any condition, and (3) priming and recognition performance will be weakly correlated. Predictions (1) and (2) were confirmed by the experiments, and some evidence for (3) was obtained, providing support for the model. (c) 2006 Elsevier Inc. All rights reserved.