Eye-Gaze direction triggers a more specific attentional orienting compared to arrows

Abstract

Numerous studies have shown that eye-gaze and arrows automatically shift visuospatial attention. Nonetheless, it remains unclear whether the attentional shifts triggered by these two types of stimuli differ in some important aspects. It has been suggested that an important difference may reside in how people select objects in response to these two types of cues, eye-gaze eliciting a more specific attentional orienting than arrows. To assess this hypothesis, we examined whether the allocation of the attentional orienting triggered by eye-gaze and arrows is modulated by the presence and the distribution of reference objects (i.e., placeholders) on the scene. Following central cues, targets were presented either in an empty visual field or within one of six placeholders on each trial. In Experiment 2, placeholder-objects were grouped following the gestalt's law of proximity, whereas in Experiment 1, they were not perceptually grouped. Results showed that cueing one of the grouped placeholders spreads attention across the whole group of placeholder-objects when arrow cues were used, while it restricted attention to the specific cued placeholder when eye-gaze cues were used. No differences between the two types of cues were observed when placeholder-objects were not grouped within the cued hemifield, or no placeholders were displayed on the scene. These findings are consistent with the idea that socially relevant gaze cues encourage a more specific attentional orienting than arrow cues and provide new insight into the boundary conditions necessary to observe this dissociation.

Fig 1. Schematic view of a trial sequence for both the gaze cue and the arrow cue conditions.

The example represents: A) gaze-cue/placeholder-absent/same-hemifield condition, and B) arrow-cue/placeholder-present/same-location/same-hemifield condition.

Fig 2. Illustration of the four types of cue-target relation of Experiment 1.

The images represent the gaze-cue in a placeholder-present condition. The cue-target relation for the placeholder-absent condition was the same, with the exception that no placeholder boxes were presented on the scene.

Fig 3. Reaction times (RTs) results from Experiment 1.

Results are shown separately for the general-cueing effect (Same-Location/Same Hemifield vs. Opposite-Location/Opposite Hemifield) and the hemifield-effect (Same-Hemifield vs. Opposite Hemifield). Mean RTs presented for each type of cue as a function of the cue-target relation in the placeholder-present and placeholder-absent conditions. Error bars represent the standard error of the mean, computed following Cousineau’s [31] method to eliminate variability between participants.

Fig 4. Schematic view of a trial sequence for both the gaze cue and the arrow cue conditions of Experiment 2.

The example represents: A) gaze/placeholder-absent/same-location/same-group condition, B) arrow/placeholder-present/opposite-group condition.

Fig 5. Illustration of the four types of cue-target relation of Experiments 2.

The placeholder-group tilted orientation shown here is -45˚ from vertical. The top images represent an example of gaze cue in a placeholder-present condition; the bottom images represent the arrow cue in a placeholder-present condition. The cue-target relation for the placeholder-absent condition was the same, with the exception that no placeholder boxes were presented on the scene.

Fig 6. Reaction times (RTs) results from Experiment 2.

Results are shown separately for the general-cueing effect and the grouping-effect. Mean RTs presented for each type of cue as a function of the cue-target relation in the placeholder-present and placeholder-absent conditions. Error bars represent the standard error of the mean, computed following Cousineau’s [31] method to eliminate variability between participants.