Depressive realism and the effect of intertrial interval on judgements of zero, positive, and negative contingencies

Abstract

In three experiments we tested how the spacing of trials during acquisition of zero, positive, and negative response–outcome contingencies differentially affected depressed and nondepressed students' judgements. Experiment 1 found that nondepressed participants' judgements of zero contingencies increased with longer intertrial intervals (ITIs) but not simply longer procedure durations. Depressed groups' judgements were not sensitive to either manipulation, producing an effect known as depressive realism only with long ITIs. Experiments 2 and 3 tested predictions of Cheng's (1997) Power PC theory and the Rescorla–Wagner (1972) model, that the increase in context exposure experienced during the ITI might influence judgements most with negative contingencies and least with positive contingencies. Results suggested that depressed people were less sensitive to differences in contingency and contextual exposure. We propose that a context-processing difference between depressed and nondepressed people removes any objective notion of “realism” that was originally employed to explain the depressive realism effect (Alloy & Abramson, 1979).

Acknowledgments

We thank Esnath Sibanda and Louise Neville for collecting data for Experiment 1. The data from Experiments 2 and 3 formed part of the first author's PhD dissertation.

Notes

¹ Effect sizes like Cohen's d and Hedges's g are based on the standardized difference between the sample means. (Hedges's g is preferred because it uses a pooled within-sample estimate of the population standard deviation, rather than simply the control group standard deviation.) One advantage of this method is that the effect size is expressed in standard deviation units, is easily interpretable, and is subject to the effect size conventions given by Cohen (small ≅ 0.2; medium ≅ 0.5; large ≅ 0.8). Like all effect sizes, the size of Hedges's g is not determined by the scale of the dependent variable and is useful as a tool for comparing effects across experiments, including those with different sample sizes.