No abstract

The peak-end memory bias has been well documented for the retrospective evaluation of pain. It describes that the retrospective evaluation of pain is largely based on the discomfort experienced at the most intense point (peak) and at the end of the episode. This is notable because it means that longer episodes with a better ending can be remembered as less aversive than shorter ones; this is even if the former had the same peak in painfulness and an overall longer duration of pain. Until now, this bias has not been studied in the domain of anxiety despite the high relevance of variable levels of anxiety in the treatment of anxiety disorders. Therefore, we set out to replicate the original studies but with an induction of variable levels of anxiety. Of 64 women, half watched a clip from a horror movie which ended at the most frightening moment. The other half watched an extended version of this clip with a moderately frightening ending. Afterward, all participants were asked to rate the global anxiety which was elicited by the video. When the film ended at the most frightening moment, participants retrospectively reported more anxiety than participants who watched the extended version. This is the first study to document that the peak-end bias can be found in the domain of anxiety. These findings require replication and extension to a treatment context to evaluate its implications for exposure therapy.

Previous research suggests that state anxiety may sway political attitudes. However, previous experimental procedures induced anxiety using political contexts (e.g., social or economic threat). In a pre-registered laboratory experiment, we set out to examine if anxiety that is unrelated to political contexts can influence political attitudes. We induced anxiety with a threat of shock paradigm, void of any political connotation. All participants were instructed that they might receive an electric stimulus during specified threat periods and none during safety periods. Participants were randomly assigned to one of two conditions: Political attitudes (implicit and explicit) were assessed under safety in one condition and under threat in the other. Psychometric, as well as physiological data (skin conductance, heart rate), confirmed that anxiety was induced successfully. However, this emotional state did not alter political attitudes. In a Bayesian analytical approach, we confirmed the absence of an effect. Our results suggest that state anxiety by itself does not sway political attitudes. Previously observed effects that were attributed to anxiety may be conditional on a political context of threat.

No abstract