SEARCH

SEARCH BY CITATION

Objective: This paper reviews 59 studies looking at cognitive, individual differences and physiological correlates of the repressive coping style, as defined by Weinberger, Schwartz, and Davidson (1979). A central aim is to evaluate the relative importance of the anxiety and social desirability components of repression. Thus, the empirical study investigates the relationships between repression and a number of relevant, but hitherto unexamined, constructs, including trait emotional intelligence (trait EI), self-estimated intelligence, functional and dysfunctional impulsivity, and stoicism. It was hypothesized that repressors would score higher than the other three groups on trait EI, self-estimated IQ and functional impulsivity, but lower on dysfunctional impulsivity.

Method: In total, 259 (174 females) participants from three British universities completed questionnaires measuring the dependent and independent variables. Participants were divided into four groups (truly low anxious, non-defensive/high anxious, defensive/high anxious and repressors) based on their scores on anxiety and social desirability. Analyses (moderated multiple regressions and ANOVAs) were conducted both on the total sample as well as on ‘extreme-scoring’ individuals.

Results: Where there were significant differences, the hypotheses were supported, particularly with respect to trait EI, self-estimated IQ and impulsivity. Using ‘extreme-scoring’ groups did not effectively change the results. The regressions revealed an absence of significant interactions between anxiety and social desirability.

Conclusion: Results are discussed in terms of the now replicated effect that repressors present a highly positive and optimistic self-image, despite cognitive and behavioural data suggesting that their coping style is psychologically unhealthy. In addition, it is argued that many findings in the repressive coping style literature can be parsimoniously explained through main effects of anxiety or social desirability alone (i.e., without invoking a construct that combines the two).