Processing bias in anxious subjects and repressors, measured by emotional Stroop interference and attentional allocation.

Hypothesized that repressors (Ss high in defensiveness with low trait anxiety) would show cognitive avoidance of threatening information in an attention deployment task, but an attentional bias for the same information in an emotional interference task, while Ss high in anxiety would show a threat-related bias on both tasks. A modified Stroop task and a visual probe task (VPT) were used, with physical and social threat words, and social and general positive words. Ss were 51 undergraduates. Results show that high state anxiety was related to greater Stroop interference for physical threat words as well as for social words, both threat and positive. No group effects were found for the Stroop, in spite of sufficient power. In the VPT, high trait anxious Ss shifted attention only towards social threat words, especially when these words were presented outside their attentional focus. No difference involving the repressor group was present. There was a small positive inter-task relation for social threat-related bias. It is suggested that the emotional biases measured by the Stroop and the VPT reflect automatic decisions about cognitive resource allocation at subsequent phases in information processing, at which increasingly more specific aspects of the emotional information are deciphered and use