Anxiety-related attentional biases in the spatial and temporal dimensions

Numerous studies support the existence of attentional biases towards threatening information in anxiety. However, it is not clear whether this bias involves difficulties in disengaging attention from threatening stimuli causing interference to other tasks, or whether it involves facilitated detection of such stimuli. This is because findings from previous paradigms can be interpreted in either way. The current experiments were designed to clearly differentiate between these alternative interpretations, both in the spatial (Experiment 1) and in the temporal (Experiment 2) attentional dimensions. In Experiment 1, high (n = 31) and low (n = 28) anxious Ps performed a visual search task with a lexical decision component, in which Ps needed to rapidly decide whether an oddball letter string among a variable number of letter-strings was a word or a nonword. High anxious Ps were expected to be quicker in detecting threatening relative to neutral words in a facilitation condition and more distracted by threatening distractors in an interference condition. In Experiment 2, high (n = 30) and low (n = 27) anxious participants were exposed to streams of words containing two highlighted targets (T1 and T2) presented by rapid serial visual presentation. Experimental stimuli (threatening and neutral words) were used as T2's in the facilitation condition and as T1's in the interference condition. We expected that identification of T1would cause a 500-msec impairment in the identification of T2 (an attentional blink [AB]) in the facilitation condition for both types of T2 for low anxious Ps but only for the neutral words for the high anxious Ps; in the interference condition we expected that the AB effect would be stronger for the high anxious Ps when threatening words served as T1's. In both experiments the emotion-related attentional effects were equally evident in both groups, and therefore no anxiety-related attentional biases could be detected. In Experiment 1 participants were also tested for their recall of the experimental words used in the visual search task, and the results provided some support for the existence of explicit memory biases in anxiety. In Experiment 2, the cognitive effects were affected by the valence of the experimental words.