Human beings can show preferentially attentional bias to different facial expressions. However, it is unclear whether the modulation of selective attention by facial expressions is based on the face itself (object-based attention) or its location (space-based attention). The current study aimed to test this problem by using faces with different emotional valences in the two-rectangle paradigm across three experiments. We found that there was no significant difference in space-based effect among the positive, neutral, and negative conditions. However, the object-based effect was larger for the negative condition than for the neutral and positive ones, because of its slower reaction times for the invalid different-object trials. The results indicated that the object-based attentional selection was modulated by facial expressions, and that faces expressing negative emotions hamper the disengagement of attention from the whole object (i.e., the face), instead of the certain location. Our study can further add support to the attentional prioritization hypothesis over attentional spreading hypothesis.

The current study investigated the mechanism underlying subliminal inhibition using the negative compatibility effect (NCE) paradigm. We hypothesized that a decrease in prime activation affects the subsequent inhibitory process, delaying onset of inhibition and reducing its strength. Two experiments tested this hypothesis using arrow stimuli as primes and targets. Two different irrelevant masks (i.e., a mask sharing no prime features) were presented in succession in each trial to not only ensure that primes were processed subliminally, but also avoid feature updating between primes and masks. Prime/target compatibility and prime background density were manipulated in Experiment 1. Results showed that under subliminal inhibitory condition, the NCE disappears when the density increases (i.e., pixel density in the prime's background of 25 %) in Experiment 1. However, when we fixed the prime's background at the density of 25 % and manipulated prime/target compatibility as well as inter-stimuli-interval (ISI) between mask and target in Experiment 2, behavioral results showed marginally significant NCEs in the 150-ms ISI condition. Electrophysiological evidence showed the lateralized readiness potential for compatible trials was significantly more positive than that for incompatible trials during the two consecutive time windows (i.e., 400-450 and 450-500 ms) in the 150-ms ISI condition. In addition, the NCE size was significant smaller in Experiment 2 than in Experiment 1. All of the results support predictions of the continuous subliminal inhibitory mechanism hypothesis which posits that decreases in prime activation strength lead to delay in inhibitory onset and decline in inhibitory strength.

Previous studies have shown that individual difference plays an important role in the object-based cueing effect observed in experiments on attentional selection. A wide range of studies have also used the theory of cognitive style to explain individual strategies in most cognitive processes. However, the characteristics of individuals that modulate object-based attentional selection are still unclear. To investigate the modulation of object-based attention by cognitive style, this study used a classic two-rectangle paradigm and compared space-based effects and object-based effects for individuals who had been preselected as wholists or analysts in terms of cognitive style. The results revealed that a space-based effect was obtained for both wholist and analytic individuals. However, an object-based effect was obtained only for analysts and not wholists, regardless of object orientation. The results further indicated that a wholist versus analytic cognitive style can modulate object-based attention by way of perceptual grouping. Our study provides the first evidence that object-based attention can indeed be influenced by individual characteristics and extends traditional model of cognitive style by indicating that wholist individuals tend to group two or more objects in a scene into one larger gestalt.