Detecting state and trait anxiety from auditory and visual cues: a meta-analysis.

Observers' ability to detect state anxiety and trait anxiety in others was evaluated in a meta-analysis that also included a critical moderator variable, communication channel. The overall effect size (r) for accuracy was .39 for the 46 state anxiety studies and .26 for the 34 trait anxiety studies. However, the effect of communication channel was very substantial in magnitude and highly significant; it also interacted with the type of anxiety. State anxiety was correctly identified more often from audio-only stimuli (M r = .49) compared with video-only cues (r = .24), whereas the opposite was true for trait anxiety, where video-only cues resulted in greater accuracy (r = .31) compared with audio-only stimuli (r = .05). Results indicated that the moment-to-moment change in emotion (state anxiety) may be better communicated with vocal/verbal stimuli, whereas more stable features of personality (trait anxiety) may be more apparent from body and facial cues.