Emotion guided threat detection: Expecting guns where there are none.

Abstract

Five experiments examine whether the ability of emotions to influence judgments of threat extends to a very basic process inherent in threat detection: object recognition. Participants experiencing different emotions were asked to make rapid judgments about whether target individuals were holding guns or neutral objects. Results across 4 experiments supported the hypothesis that anger increases the probability that neutral objects will be misidentified as ones related to violence, but not the converse. Of import, the findings demonstrate that this bias is not a simple function of the negative valence of an emotional state, but stems from specific threat-relevant cues provided by anger. Direct manipulation of participants' expectancies for encountering guns in the environment is shown not only to remove the bias among angry individuals when set to be low but also to produce a corresponding bias among neutral participants when set to be high. A 5th study demonstrates that the bias is amenable to correction given sufficient ability.