Repetition Can Have Similar or Different Effects on Accurate and False Recognition

In two experiments involving the Deese/Roediger and McDermott (DRM) procedure, participants demonstrated false memory when they were presented with semantically related word lists, either 1 time or 5 or 10 times (Experiment 1) or 1 time or 5 or 25 times (Experiment 2), at rates of 20-ms or 2-s/word, followed by a recognition memory test. Following 20-ms exposures, accurate and false recognition increased monotonically with list repetition; following 2-s exposures, list repetition affected accurate and false recognition differently. Accurate recognition increased monotonically with list repetition, whereas false recognition first increased and then decreased with additional repetitions. These findings were consistent with expectations based on fuzzy trace theory and other dual-process accounts. But, they were notably inconsistent with the implicit activation hypothesis and difficult to reconcile with other single-process accounts. Depending on stimulus exposure duration, repetition can affect accurate and false recognition similarly or differently, and it can definitely contribute to false recognition.