Degree Name

Doctor of Philosophy


Previous studies on depression have suggested that depression is accompanied by memory disturbances such as memory deficits (Hasher & Zacks, 1979; Ellis & Ashbrook, 1988; Hartlage, Alloy, Vazquez, & Dykman, 1993). The thesis aims to explore depressive memory deficits in terms of long-term and short-term memory, employing precision of elaboration and memory span tasks respectively.

The long-term memory deficits focused on elaboration deficits using a subjectgenerated elaboration method (Stein, & Bransford, 1979). Regarding elaboration deficits in depression, several studies suggested that memory deficits in recall occur in tasks requiring elaborative processing such as in semantic processing within the depth of processing paradigm, and processing of elaborated sentences (Weingartner, Cohen, Murphy, Martello, & Gerdt, 1981, study 1; Ellis, Thomas, & Rodrigez, 1984, study 1). Recall deficits due to these elaboration deficits were attributed to inadequate allocation of cognitive effort expended on the current task. In addition Hasher and Zacks (1988) and Abramson, Alloy, and Rosoff (1981) suggested that depression interferes with elaborative processing. These elaborative processing deficits were inferred from reduced performance in overall recall. However, the purpose of the first three experiments was to examine directly this elaborative processing deficit through precision of elaboration, as well as recall, using a subject-generated elaboration method. The results showed that both recall and the extent of precise elaboration depends upon the task instructions. Consistent with this hypothesis, dysphoric students generated less precise elaboration in a more controlled task instruction, and greater depression is associated with less precise elaboration (experiment IC), while in a less controlled task instruction (experiment 1A), such a difference in the precision of elaboration disappears. Reduced recall performance was found in experiment 1A, but not in experiment IC. These results were discussed in terms of a cognitive effort account of depressive memory deficits and the reduced initiative hypothesis. It seems that these results put a constraint on a general elaborative processing deficit.

The following chapter shifted the focus to the mood induction procedure from naturally occurring depression with aiming to develop a better mood induction procedure in terms of duration and intensity. The most popularly used mood induction procedure, the Velten Mood Induction Procedure is flawed by its short-lived induced mood, particularly when cognitive and behavioural measurement proceeds (Clark, 1983). To overcome this problem, this study examined the effects of three mood induction procedures (the Velten procedure, music mood induction, and the combined Velten-plus-music procedure) on the strength and duration of induced depression while undertaking a sentence completion task. Mood state was measured prior to, immediately after, and at 3 six-minute intervals after mood induction. The results showed that the Velten-plus-music and Velten procedures induced a large immediate increase in depression scores as measured by the Depression Adjective Check List (DACL) and a Visual Analogue Scale (VAS). Mood induced by the Velten-plus-music and music procedure was relatively stable through the post-induction period, but mood induced by the Velten procedure dissipated quickly over time. Anxiety level was increased by all three induction procedures, relative to a neutral music control condition. The results show that the transient mood change induced by the Velten procedure can be prolonged by playing depressing music.

The subsequent four experiments examined verbal short-term memory. The effect of dysphoric mood states on immediate serial recall was investigated from a working memory perspective. Experiments 3A and 3B found no differences between dysphorics and nondysphoric controls with respect to either serial recall of 6 or 7 digits, or speech rate measures. Self-report questionnaires indicated that the dysphorics had a higher incidence negative automatic thoughts and task-irrelevant thoughts while performing the memory tasks. However, this did not influence serial recall performance. Experiment 3C investigated serial recall for long and short words, and found similar levels of recall performance, word length effects and speech rates in natural dysphorics and three groups of nondysphorics who were subjected to either neutral, elated or dysphoric mood induction. The natural dysphorics reported more negative automatic thoughts than any of the mood-induced groups, but no differences in report rates for task-irrelevant thoughts were found. In experiment 3D, a running memory span task was used, in which subjects were required to recall the final six items of a list in which list length was 6, 8, 10, or 12 items. In this task, dysphorics showed memory deficits only when list length exceeded 6 items, and updating of the rehearsal set was required. The results of Experiments 3A - 3C show that the articulatory loop is functioning normally in dysphorics, but experiment 3D suggests that dysphorics have difficulty in adapting their rehearsal strategies, suggesting a deficit in central executive function. The results are discussed in terms of the distinction between automatic vs. effortful processing, and in terms of the distinction of maintenance vs. elaborative rehearsal.

The next two experiment address similar issues to verbal short-term memory in the visual recognition memory of matrix patterns with varying retention intervals (1, 8, 15 seconds). Experiment 4A using a 4 x 4 matrix pattern showed that recognition performance of dysphorics did not deteriorate more rapidly over retention intervals, compared to nondysphorics, nor did overall performance vary between both groups. The subsequent experiment, 4B, used a 5 x 5 matrix pattern with longer retention intervals (1, 15, 30 seconds). There was a significant interaction between group and retention intervals. However, this interaction effect was due to poor performance of dysphoric subjects in the immediate recognition test rather than in a longer retention interval, compared to nondysphorics. These results suggest encoding difficulties rather than difficulties in the maintenance rehearsal of stimuli. In terms of the working memory model, these results are similar to those in verbal short-term memory, suggesting that dysphoric students have some difficulties with central executive functioning rather than with the visuo-spatial scratch pad responsible for the maintenance rehearsal of visual stimuli.

In sum, the results reported appear to fit better with the initiative deficit hypothesis rather than the reduced attentional capacity or cognitive effort accounts of depressive memory deficits because performance of depressives was sensitive to task instruction and cognitive strategy rather than task difficulty.