Abstract: Observing another person’s averted eye gaze led to automatic shift of attention in the corresponding direction, which facilitated subsequent early visual processing. This phenomenon was termed ‘eyes gaze cueing effect’, in short, ‘gaze cueing effect’ (GCE). Many researchers tried to interpret the processing mechanism underlying GCE. However, there were controversies on this topic. Friesen and Kingstone (2003) claimed that gaze cues triggered a kind of reflexive or exogenous shift of visual attention that was normally associated with the abrupt onset of a stimulus in the periphery of vision. Whereas Vecera and Rizzo (2004) found a patient EVR with frontal-lobe damage was impaired in using eye gaze cues to allocate attention, suggesting that GCE may involve frontal-lobe processes which control voluntary, not automatic, shifts of visuospatial attention. Both viewpoints have gained some evidences. Thus, whether gaze cues shifted spatial attention reflexively (exogenously) or voluntarily (endogenously) was still in debate. This issue would be clarified in the present study in two aspects. On the one hand, how gaze cues shifted observer’s attention. On the other hand, how gaze cues affected the processing of the target in the corresponding direction.
Two experiments were conducted in this study, which employed the spatial Stroop paradigm and the target search task respectively. In both experiments, subjects were presented with a gaze cue to the right or left visual field (VF) before the occurrence of a lateralized target to response. There were two conditions in this non-predictive cueing task: (1) targets were presented in the VF indicated by the eye gaze direction (congruent); or (2) targets were opposite to the eye gaze direction (incongruent). In the first experiment, 22 subjects were required to discriminate the direction to which the peripherally presented arrow points. While in the second experiment, another 16 subjects all were instructed to finish two different tasks: One involved the detection of single feature, and the other involved the search of feature conjunction. Stimulus presentation and data collection were conducted by E-prime software.
From the first experiment, it was found: (1) The lack of GCE was due to the processing conflict between gaze cue and other endogenous cues, which indicated that the two cues had similar processing mechanism; (2) Gaze cue produced larger spatial Stroop effects for valid than invalid trials, which suggested that eyes gaze led to the generation of an internal representation of the spatial location. From the second experiment, similar facilitation effects of the gaze cue for both single feature and feature conjunction search was found, which revealed that gaze cue affected feature extraction as well as later processing stages (decision and response processes), but it did not influence feature integration directly.
It was concluded that the intrinsic mechanism of GCE belonged to the endogenous attention system. Eyes gaze cue may induce observers’ attention shift through the generation of an internal representation of the spatial location indicated by the cue, and facilitate target feature extraction and encoding, thus speed up the response ultimately.

Key words: eyes gaze cueing effect, exogenous attention, endogenous attention, spatial Stroop effect, feature integration