Mood effects on implicit and explicit memory.

Abstract

Three major effects of mood on memory have been identified including mood-dependent memory (MDM), mood congruent memory (MCM) and resource allocation (RA) effects. The results of studies examining these effects have been inconsistent. The majority of these studies have employed explicit memory tests; however, explicit tests provide the opportunity for subjects to self-generate cues for retrieval that might overpower mood as a cue. It was hypothesized that use of an implicit memory test would highlight mood by reducing the opportunity for subjects to generate relatively stronger cues for retrieval, resulting in intensified MDM and MCM effects, provided that the implicit memory test was conceptually-driven and, therefore, could be impacted by mood, and the nominal cues provided at test were reduced to a minimum. An implicit analogue of free recall was developed which met these conditions. It was also hypothesized that MDM would be most likely to be found if stimulus items were related to mood semantically as well as temporally. Subjects studied positive, neutral and negative words following either a happy (H) or sad (S) uninstructed musical mood induction. Half of the stimulus items were encoded elaboratively and half shallowly. Prior to test, subjects received either a happy or sad musical mood induction. Subjects were placed into one of four mood groups based on subjective reports of mood prior to encoding and retrieval (HH, HS, SH, SS). Each subject received an implicit memory test (free recall analogue) followed by an explicit memory test (free recall) for the studied words. No MDM effects were observed; however, when only items that were semantically related to encoding mood (mood congruent) were examined, there was a strong trend towards mood congruency in the implicit but not explicit condition. Mood congruent retrieval was found in the implicit but not explicit condition. No mood congruent encoding or resource allocation effects were observed. It was concluded that mood had a greater opportunity to affect retrieval from episodes when implicit memory tests were employed. Some caveats to this conclusion are discussed as well as potential methodological pitfalls in conducting this type of research.