Covariation bias in women with a negative body evaluation: How is it expressed and can it be diminished?

Women with a negative body evaluation display covariation bias: They overestimate the relation between their own body and negative social feedback. This study aimed to develop a more fine-grained understanding of this covariation bias and to determine whether it could be diminished.

Seventy women completed a computer task wherein three categories of stimuli--pictures of their own body, a control woman's body, and a neutral object--were followed by (nonverbal) negative social feedback or nothing. Participants' estimates of the relation between each stimulus category and negative social feedback were assessed throughout the task.

Before starting the task, women with a more negative state body evaluation expected their body to be followed by more negative social feedback (demonstrating a priori covariation bias). During the task, when the relation between stimulus category and negative social feedback was random, women with a more negative trait and state body evaluation perceived at the present moment (online covariation bias) and retrospectively (a posteriori covariation bias) that their body was followed by more negative social feedback. When contingencies were manipulated so that women's own body was rarely followed by negative social feedback, covariation bias was temporarily diminished; this coincided with improvements in state body evaluation.

The task did not incorporate neutral or positive social feedback and focused only on undergraduate women.

Covariation bias exists preexperimentally and occurs when situational information is ambiguous. It is possible to (temporarily) diminish covariation bias. This might be a technique for improving body evaluation.