Semantically clustered words are stored with integrated context. Validating a Measurement Model for Source Memory, Storage, and Retrieval in Free Recall

Source memory (i.e., memory for context) has been studied with recognition tasks almost exclusively. However, encoding context affects recall stronger than recognition, presumably because of more complex retrieval strategies in the former task. An extension of Batchelder and Riefer (1980) pair-clustering model is proposed which is intended to measure the storage and retrieval of clusterable word pairs as well as the memory for the sources in which these were presented. In two experiments, the construct validity of the central model parameters is demonstrated. Furthermore, there was a strong stochastic dependency between recalling the sources of the first and the second word of a clustered pair, respectively, suggesting that not only semantic but also contextual features are bound together in clustered pairs. Advantages of using recall tests in source monitoring research are discussed.