The Role of Global and Local Visual Information during Gaze-Cued Orienting of Attention

Abstract

Gaze direction is an important social communication tool. Global and local visual information are known to play specific roles in processing socially relevant information from a face. The current study investigated whether global visual information has a primary role during gaze-cued orienting of attention and, as such, may influence quality of interaction. Adults performed a gaze-cueing task in which a centrally presented face cued (valid or invalid) the location of a peripheral target through a gaze shift. We measured brain activity (electroencephalography) towards the cue and target and behavioral responses (manual and saccadic reaction times) towards the target. The faces contained global (i.e. lower spatial frequencies), local (i.e. higher spatial frequencies), or a selection of both global and local (i.e. mid-band spatial frequencies) visual information. We found a gaze cue-validity effect (i.e. valid versus invalid), but no interaction effects with spatial frequency content. Furthermore, behavioral responses towards the target were in all cue conditions slower when lower spatial frequencies were not present in the gaze cue. These results suggest that whereas gaze-cued orienting of attention can be driven by both global and local visual information, global visual information determines the speed of behavioral responses towards other entities appearing in the surrounding of gaze cue stimuli.

Fig 1

Example of an unfiltered face stimulus (A) and its low-pass (B), mid-pass (C) and high-pass (D) filtered versions. The face in this figure is derived from the MacBrain Face Stimulus Set, the model approved publication in scientific journals.

Fig 2. The stimuli and timing sequence used in the EEG (500ms) and ET (300ms) task.

The illustration shows a valid trial. The face in this figure is derived from the MacBrain Face Stimulus Set, the model approved publication in scientific journals.

Fig 3. ERP waveforms and mean amplitude ± SE of the EDAN and ADAN for contralateral (contra) and ipsilateral (ipsi) gaze cues.

*P < 0.05.

Fig 4. ERP waveforms and mean amplitude ± SE of P1 amplitude, P1 latency, N200 amplitude, N200 latency, manual reaction times (RT) and saccadic reaction times (RT) for each cue-validity condition.

*P < 0.05; **P < 0.01; ***P < 0.001.

Fig 5. Mean ± SE of manual reaction times (RT) and saccadic reaction times (RT) for each filter condition.

*P < 0.05; **P < 0.01; ns = not significant.