Abstract: Posner and Cohen (1984) discovered the inhibition of return (IOR) phenomenon, which refers to the individual’s response to targets that appear in previously cued location, is slower than it is to the uncued location when the stimulus-onset asynchrony (SOA) is longer than 300ms. Psychologists agree that IOR is an adaptive and evolved mechanism that enhances the efficiency of attention during a visual search and allows people more easily to process targets that appear in novel positions. The current research uses event-related potentials (ERPs) and adopts a cue-target paradigm to investigate the IOR process during an emotion recognition task and the underlying mechanisms. We conducted a 2 (cue: valid cues, invalid cues) × 3 (target emotional: positive, negative, neutral) within-subject design, and selected the emotional face pictures (40 positive faces; 40 negative faces; 40 neutral faces) as the target stimuli of IOR. Sixteen participants of university students were instructed to complete the task of judging the target as emotional face or neutral one. Electroencephalogram (EEG) signals were recorded using a 64-electrode elastic cap and the Neuroscan ERP workstation. Based on the overall average map and the research literature, we divided the ERP components by the time windows in which they occurred: P1, 80~110 ms; N1, 110~140 ms; and N170, 140~180 ms. We used six electrodes (Po5, Po6, O1, O2, Poz and Oz) to detect the P1 and N1 components in the occipital region and two electrodes (P7 and P8) to detect the N170 component in the temporal occipital region. Repeated measure ANOVAs were conducted on the behavioural data and the measurements derived from ERP waveforms. The results demonstrate that the amplitude of P1 of the valid cues was smaller, while N1 amplitude was larger, compared with those of the invalid cues. N170 amplitudes for positive and negative faces were significantly larger than those for neutral ones. These results indicate that the response to the emotional stimulus are specialized and separated from IOR progress, and this separation consequently leads to the precedence of emotion processing when participants response to the valance of the targets, suggesting that IOR and emotional bias are two adaptation mechanisms which occur in two different neural pathways. Our findings also provide electrophysiological evidences for the research on the correlation of attention and emotion.

Key words: inhibition of return, emotional face, attention bias, Event Related Potentials (ERPs)