Mood-Dependent Memory Research Paper

The capriciousness of MDM would be cause for little concern were it not for the fact that the phenomenon has important implications for both clinical and cognitive psychology. Regarding clinical implications, it has long been recognized that severely depressed patients have trouble remembering events that they earlier experienced in a nondepressed state (see Teasdale 1983). Traditional psychoanalytic theory ascribes this problem to a dynamic process of denial, which implies that depressed individuals are unwilling to remember prior periods of wellbeing. However, as Reus et al. (1979) have conjectured, it may be that the patients are unable to remember, owing to the shift in their affective state (from a neutral or possibly positive affect to a decidedly negative one). In addition to depression, MDM has been implicated in such diverse clinical conditions as bipolar illness, traumatic amnesia, and dissociative identity disorder (see Johnson and Magaro 1987, Schacter and Kihlstrom 1989).

With respect to cognitive implications, Bower has admitted that when he began working on MDM, he was ‘occasionally chided by research friends for even bothering to demonstrate such an ‘‘obvious’’ triviality as that one’s emotional state could serve as a context for learning’ (Bower and Mayer 1989, p. 164). Though the criticism seems ironic today, it was incisive at the time, for many theories strongly suggested that memory should be mood dependent. These theories included the early drive-as-stimulus views held by Hull (1943) and Miller (1950), as well as such modern innovations as Tulving’s encoding specificity principle (Tulving and Thomson 1973) and Bower’s own network model of emotion (Bower and Cohen 1982). Thus, the frequent failure to demonstrate MDM reflects badly on many classic and contemporary theories of memory, and it blocks understanding of the basic issue of how context affects learning and remembering (see Davies and Thomson 1988).

Prompted by these clinical and cognitive implications, researchers in the 1990s focused their efforts on identifying why mood dependence sometimes comes and sometimes goes. These efforts have paid off, as it now appears that MDM can be demonstrated in a robust and reliable manner, provided that several critical factors are taken into account. These factors are of two types: one concerned with characteristics of the subjects’ encoding and retrieval tasks; the other with attributes of the moods they experience while performing these tasks. The following sections review evidence bearing on both types of factors; unless noted otherwise, this evidence stems from studies in which happy and sad moods were engendered experimentally in ‘normal’ subjects (university undergraduates or other healthy volunteers) by means of a specific affectinduction technique (examples of which will be given later).

1. Task Factors

Regarding task factors, one key consideration is the manner in which retrieval of the target events is tested. According to several accounts (e.g., Bower 1992, Eich 1995a, Kenealy 1997), MDM is more apt to arise when retrieval is mediated by ‘invisible’ cues produced by the subject than by ‘observable’ cues provided by the experimenter. Thus, free recall is a much more sensitive measure of mood dependence than are either cued recall or old/new recognition memory.

A second critical task factor is, in effect, the complement of the first. Just as the odds of demonstrating MDM are improved by requiring subjects to generate their own cues for retrieving the target events, so too are these prospects enhanced by requiring subjects to generate the events themselves. In support of this proposition, Eich and Metcalfe (1989) observed a significantly greater mood dependent effect in the free recall of verbal items that subjects had actively generated (e.g., guitar, elicited by the request: Name a musical instrument that begins with g) in contrast to items that the subjects had simply read (e.g., gold, embedded in the phrase: Gold is a precious metal that begins with g). This result, which was replicated by Beck and McBee (1995), occurs regardless of whether the overall level of generate-item recall is higher than that of read items—the prototypic ‘generation effect’ (Slamecka and Graf 1978)—or whether it is lower (as happens when subjects read some target items three times but generate others only once). Moreover, and in line with remarks made in the preceding paragraph, the results of a test of old new recognition memory, which was administered shortly after free recall, showed no sign of mood dependence for either type of target. Thus, it seems that the more one must rely on internal resources, rather than on external aids, to generate both the cues required to effect retrieval of the target events and the events themselves, the more likely is one’s memory for these events to be mood dependent.

This reasoning provided the impetus for a series of studies by Eich et al. (1994). During the initial or encoding session of one of these studies (Experiment 2), university students were asked to recollect specific episodes or events, from any time in their personal past, that were called to mind by neutral-noun probes such as ship and street; every participant recollected as many as 16 different events, each evoked by a different probe. After recounting the particulars of a given event, subjects rated its original emotional valence— that is, whether the event was a positive, neutral, or negative experience when it occurred. Half of the subjects completed this task of autobiographical event generation while they were feeling happy and half did so while sad—affects that had been induced by having subjects ponder mood-appropriate ideas and images while mood-appropriate music played softly in the background.

During the retrieval session, held two days after encoding, subjects were asked to freely recall (i.e., recall in any order, without benefit of any observable reminders or cues) the gist of as many of their previously generated events as possible, preferably by recalling their precise corresponding probes. Subjects undertook this test of autobiographical event recall either in the same mood in which they had generated the events or in the alternative affective state, thus creating two conditions in which encoding and retrieval moods matched (happy/happy, sad/sad) and two in which they mismatched (happy/sad, sad/happy).

Results of the encoding session showed that, on average, subjects generated more positive events (11.1 vs. 6.7), fewer negative events (3.3 vs. 6.8), and about the same small number of neutral events (1.2 vs. 2.0) when they were happy rather than sad. This pattern replicates prior work (e.g., Clark and Teasdale 1982, Snyder and White 1982), and it provides evidence of mood congruent memory—the ‘enhanced encoding and/or retrieval of material the affective valence of which is congruent with ongoing mood’ (Blaney 1986, p. 229).

Results of the retrieval session provided evidence of mood dependent memory. Relative to subjects whose encoding and retrieval moods matched, those whose moods mismatched recalled, on average, a smaller percentage of their previously generated positive events (26 percent vs. 37 percent), neutral events (17 percent vs. 32 percent), and negative events (27 percent vs. 37 percent). Collapsing across event types, totalevent recall averaged 35 percent in the happy/happy condition, 23 percent in the happy/sad condition, 27 percent in the sad/happy condition, and 34 percent in the sad/sad condition.

This effect does not appear to be a fluke: The same advantage was seen in two other studies in which happy and sad moods were induced through a combination of ideas, images, and music (Eich et al. 1994, Experiments 1 and 3) as well as in three separate studies in which the subjects’ affective states were altered by changing their physical surroundings (Eich 1995b). Moreover, similar results were obtained in a recent investigation of bipolar patients who cycled rapidly, and spontaneously, between states of mania or hypomania and depression (Eich et al. 1997). Thus it seems that autobiographical event generation, when coupled with event free recall, constitutes a useful tool for exploring mood dependent effects under both laboratory and clinical conditions, and that these effects emerge in conjunction with either experimentally induced or naturally occurring shifts in affective state.

2. Mood Factors

Up to now, discussion has focused on factors that determine the sensitivity of an encoding or a retrieval task to the detection of mood dependence. It stands to reason, however, that no matter how sensitive these tasks may be, the odds of demonstrating MDM are slim in the absence of an effective manipulation of mood. So, what makes a mood manipulation effective?

One consideration is mood strength. By definition, MDM demands a statistically significant loss of memory when target events are encoded in one mood and retrieved in another. It is doubtful whether anything less than a substantial shift in mood could produce such impairment. Indeed, the results of a meta-analysis by Ucros (1989) revealed that the greater the difference in moods—depression vs. elation, for example, as opposed to depression vs. a neutral affect—the greater the mood dependent effect. In a related vein, Bower (1992) has proposed that MDM reflects a failure of information acquired in one state to generalize to the other, and that generalization is more apt to fail the more dissimilar the two moods are.

No less important than the strength of the moods is their stability over time and across tasks. In terms of demonstrating MDM, it does no good to engender a mood that evaporates as soon as the subject is given something to do, like memorize a list of words or generate a series of autobiographical events. It is possible that some studies failed to find mood dependence simply because they relied on moods that were potent initially but paled rapidly (see Eich and Metcalfe 1989).

Yet a third element of an effective mood is its authenticity or emotional realism. Using the autobiographical event generation and recall tasks described earlier, Eich and Macaulay (2000b) found no sign whatsoever of MDM when undergraduates simulated feeling happy or sad, when in fact their mood had remained neutral throughout testing. Moreover, in several studies involving the intentional induction of specific moods, students have been asked to assess candidly (postexperimentally) how authentic or real these moods felt. Those who claim to have been most genuinely ‘moved’ tend to show the strongest mood dependent effects (see Eich 1995a, Eich et al. 1994).

Thus it appears that the prospects of demonstrating MDM are improved by instilling affective states that have three important properties: strength, stability, and sincerity. In principle, such states could be induced in a number of different ways; for instance, students might (a) read and internalize a series of self-referential statements (e.g., I’m feeling on top of the world vs. Lately I’ e been really down), (b) obtain false feedback on an ostensibly unrelated task, (c) receive a posthypnotic suggestion to experience a specified mood, or, as noted earlier, (d) contemplate mood-appropriate thoughts while listening to mood-appropriate music (see Martin 1990). In practice, however, it is possible that some methods are better suited than others for inducing strong, stable, and sincere moods. Just how real or remote this possibility is remains to be seen through close, comparative analysis of the strengths and shortcomings of different mood-induction techniques.

3. Conclusions

The preceding sections reviewed recent research aimed at identifying factors that play pivotal roles in the occurrence of mood dependence. What conclusions can be drawn from this line of work?

The broadest and most basic conclusion is that the problem of unreliability that has long beset research on MDM is not as serious or stubborn as most investigators believed it to be only a decade ago. More to the point, it now appears that robust and reliable evidence of mood dependence can be realized under conditions in which subjects (a) experience strong, stable, and sincere moods, (b) take responsibility for generating the target events themselves, and (c) also assume responsibility for generating the cues required to retrieve these events.

Taken together, these observations make a start toward demystifying MDM—but only a start. To date, only a few factors have been examined for their role in mood dependence; odds are that other factors of equal or greater significance exist, awaiting discovery. For instance, it is conceivable that mood dependent effects become stronger, not weaker, as the interval separating encoding and retrieval grows longer (see Eich 1995a). Also, given that individual differences in personality have already been shown to play an important part in mood-congruent memory (see Bower and Forgas 2000, Smith and Petty 1995), it is possible—indeed probable—that subject factors also figure prominently in mood dependent memory (see Eich and Macaulay 2000a). And although the literature is replete with clinical conjectures about the role of MDM in various forms of memory pathology, it is lacking in hard clinical data. To date, few studies have sought to demonstrate mood dependence in individuals who experience significant, sometimes extreme, alterations in mood as a consequence of a psychiatric condition, such as unipolar mania (Weingartner et al. 1977), bipolar illness, or dissociative identity disorder (Nissen et al. 1988). As we have seen, recent research involving induced moods in normal subjects has helped establish the reliability of MDM. By exploring MDM within the context of the marked mood swings shown by select clinical samples, it may be possible to establish the phenomenon’s generality.

Bibliography:

1. Beck R C, McBee W 1995 Mood dependent memory for generated and repeated words: Replication and extension. Cognition and Emotion 9: 289–307
2. Blaney P H 1986 Affect and memory: A review. Psychological Bulletin 99: 229–46
3. Bower G H 1981 Mood and memory. American Psychologist 36: 129–48
4. Bower G H 1992 How might emotions affect learning? In: Christianson S-A (ed.) Handbook of Emotion and Memory. Erlbaum, Hillsdale, NJ, pp. 3–31
5. Bower G H, Cohen P R 1982 Emotional influences in memory and thinking: Data and theory. In: Clark M S, Fiske S T (eds.) Affect and Cognition. Erlbaum, Hillsdale, NJ, pp. 291–332
6. Bower G H, Forgas J P 2000 Affect, memory, and social cognition. In: Eich E, Kihlstrom J F, Bower G H, Forgas J P, Niedenthal P M (eds.) 2000 Cognition and Emotion. Oxford University Press, New York, pp. 87–168
7. Bower G H, Mayer J D 1989 In search of mood dependent retrieval. Journal of Social Behavior and Personality 4: 121–56
8. Clark D M, Teasdale J D 1982 Diurnal variation in clinical depression and accessibility of memories of positive and negative experiences. Journal of Abnormal Psychology 91: 87–95
9. Davies G M, Thomson D M (eds.) 1988 Memory in Context: Context in Memory. Wiley, Chichester, UK
10. Eich E 1995a Searching for mood dependent memory. Psycho-logical Science 6: 67–75
11. Eich E 1995b Mood as a mediator of place dependent memory. Journal of Experimental Psychology: General 124: 293–308
12. Eich E, Macaulay D 2000a Fundamental factors in mood dependent memory. In: Forgas J P (ed.) Feeling and Thinking: The Role of Affect in Social Cognition. Cambridge University Press, New York, pp. 109–30
13. Eich E, Macaulay D 2000b Are real moods required to reveal mood-congruent and mood dependent memory? Psychological Science 11: 244–8
14. Eich E, Macaulay D, Lam R W 1997 Mania, depression, and mood dependent memory. Cognition and Emotion 11: 607–18
15. Eich E, Macaulay D, Ryan L 1994 Mood dependent memory for events of the personal past. Journal of Experimental Psychology: General 123: 201–15
16. Eich E, Metcalfe J 1989 Mood dependent memory for internal vs. external events. Journal of Experimental Psychology: Learning, Memory, and Cognition 15: 443–55
17. Hull C L 1943 Principles of Behavior. Appleton-Century-Crofts, New York
18. Johnson M H, Magaro P A 1987 Effects of mood and severity on memory processes in depression and mania. Psychological Bulletin 101: 28–40
19. Kenealy P M 1997 Mood-state-dependent retrieval: The effects of induced mood on memory reconsidered. Quarterly Journal of Experimental Psychology 50A: 290–317
20. Martin M 1990 On the induction of mood. Clinical Psychology Review 10: 669–97
21. Miller N E 1950 Learnable drives and rewards. In: Stevens S S (ed.) Handbook of Experimental Psychology. Wiley, New York, pp. 435–72
22. Nissen M J, Ross J L, Willingham D B, MacKenzie T B, Schacter D L 1988 Memory and awareness in a patient with multiple personality disorder. Brain and Cognition 8: 21–38
23. Reus V I, Weingartner H, Post R M 1979 Clinical implications of state-dependent learning. American Journal of Psychiatry 136: 927–31
24. Schacter D L, Kihlstrom J F 1989 Functional amnesia. In: Boller F, Grafman J (eds.) Handbook of Neuropsychology. Elsevier, New York, Vol. 3, pp. 209–30
25. Slamecka N J, Graf P 1978 The generation effect: Delineation of a phenomenon. Journal of Experimental Psychology: Human Learning and Memory 4: 173–84
26. Smith S M, Petty R E 1995 Personality moderators of mood congruency effects on cognition: The role of self-esteem and negative mood regulation. Journal of Personality and Social Psychology 68: 1092–107
27. Snyder M, White P 1982 Moods and memories: Elation, depression, and the remembering of the events of one’s life. Journal of Personality 50: 149–67
28. Teasdale J D 1983 Negative thinking in depression: Cause, effect, or reciprocal relationship? Advances in Behavior Research and Therapy 5: 3–25
29. Tulving E, Thomson D M 1973 Encoding specificity and retrieval processes in episodic memory. Psychological Review 80: 352–73
30. Ucros C G 1989 Mood state-dependent memory: A meta-analysis. Cognition and Emotion 3: 139–67
31. Weingartner H, Miller H, Murphy D L 1977 Mood-state-dependent retrieval of verbal associations. Journal of Abnormal Psychology 86: 276–84