An examination of two-process theories of false recognition

Contemporary theories of false memory suggest there are two processes that combine to produce false memory: one that increases false memory (error-inflating processes) and one that counteracts false memory (error-editing processes). Two experiments using the DRM paradigm (Deese, 1959; Roediger & McDermott, 1995) explored the influence of manipulating the number of associates studied, study item presentation frequency, backward associative strength, and study time on error-inflating and error-editing processes separately by examining speeded and unspeeded recognition decisions. The results of these studies indicate that (1) increasing the number of associates studied primarily influenced error-inflating processes; (2) increasing backward associative strength increased error-inflating processes and impaired error-editing processes; (3) increasing study item presentation frequency increased both error-inflating and error-editing processes; and (4) increasing study time had a weak effect on error-editing processes. Further, the results of these studies suggest that comprehensive theories of false memory phenomena must propose the existence of two different factors: one that increases false memory and is available early in memory retrieval, and one that usually, but not always, decreases false memory and is available later in retrieval.