Evaluative priming reveals dissociable effects of cognitive vs. physiological anxiety on action monitoring

Performance monitoring enables the rapid detection of mismatches between goals or intentions and actions, as well as subsequent behavioral adjustment by means of enhanced attention control. These processes are not encapsulated, but they are readily influenced by affective or motivational variables, including negative affect. Here we tested the prediction that worry, the cognitive component of anxiety, and arousal, its physiological counterpart, can each influence specific processes during performance monitoring. In two experiments, participants were asked to discriminate the valence of emotional words that were preceded by either correct (good) or incorrect (bad) actions, serving as primes in a standard evaluative priming procedure. In Experiment 1 (n = 36) we examined the influence of trait worry and arousal. Additionally, we included a face priming task to examine the specificity of this effect. Stepwise linear regression analyses showed that increased worry, but not arousal, weakened the evaluative priming effect and therefore the rapid and automatic processing of actions as good or bad. By contrast, arousal, but not worry, increased posterror slowing. In Experiment 2 (n = 30) state worry was induced using an anagram task. Effects of worry on action monitoring were trait but not state dependent, and only evidenced when actions were directly used a primes. These results suggest a double dissociation between worry and arousal during performance monitoring.