An investigation of the depressive realism effect and its mechanisms

The depressive realism effect refers to a phenomenon in which depressed individuals are more realistic at assessing the relationship between two events (referred to as contingency learning) than non-depressed individuals (Alloy & Abramson, 1979). The main aim of this thesis was to investigate the depressive realism effect by: (i) replicating the effect and (ii) understanding its potential mechanisms.Experiments 1 to 5 (Chapter 2) aimed to replicate the depressive realism effect by comparing the accuracy of assessments of contingencies made by unselected participants who received either a positive or a negative mood induction. Despite the effectiveness of the mood induction procedures, across these five experiments, there were no differences between the assessment of contingencies provided by participants in the positive and negative mood induction conditions. These findings suggested that negative mood alone was not sufficient to elicit the depressive realism effect. Thus, consistent with the original study by Alloy and Abramson (1979), in Experiment 6 (Chapter 3) a high dysphoric sample was employed. The depressive realism effect was replicated in this experiment, such that high dysphoric participants made more accurate assessments of contingencies than did low dysphoric participants. This replication suggested that studying a high dysphoric sample may be critical to detect the depressive realism effect.Experiments 7 to 14 investigated the role of various mechanisms that have been theorised to account for the depressive realism effect. Specifically, the results of Experiment 7 (Chapter 3) suggested that information presented early in a task had no impact upon participants’ assessments of contingencies. Experiments 8 - 12 (Chapters 4 & 5) did not find evidence to support the proposal that information processing biases, in particular expectations and memory biases, play a role in the depressive realism effect. Further, the depressive realism effect was not replicated