Change of anxiety states by probabilistic processing

In contrast to the vast literature on the influence of affect on cognitive processing, this report raised the neglected question whether cognitive pro cessing can influence anxiety states. Based on network and proceduralizatio n models, the hypothesis was that anxiety states are prolonged only when co ngruent with compatible types of probabilistic processing. Anxiety states w ere induced in 66 female university students with an autobiographical recol lection method. To manipulate style of processing, the participants judged the likelihood of events happening to them using probability or categorical response formats. Results showed that different types of anxiety states (i .e., threat to self-esteem. physical threat) are sustained by different kin ds of processing (i.e., graded or yes-no likelihood judgments, respectively ). These findings point to the reciprocal influence of cognitive processing on affect. The results are discussed with regard to network and procedural ization models and the mood repair hypothesis.