Mislocalization of visual stimuli: independent effects of static and dynamic attention

Abstract

Shifts of visual attention cause systematic distortions of the perceived locations of visual objects around the focus of attention. In the attention repulsion effect, the perceived location of a visual target is shifted away from an attention-attracting cue when the cue is presented before the target. Recently it has been found that, if the visual cue is presented after the target, the perceived location of the target shifts toward the location of the following cue. One unanswered question is whether a single mechanism underlies both attentional repulsion and attraction effects. We presented participants with two disks at diagonal locations as visual cues and two vertical lines as targets. Participants were asked to perform a forced-choice task to judge targets' positions. The present study examined whether the magnitude of the repulsion effect and the attraction effect would differ (Experiment 1), whether the two effects would interact (Experiment 2), and whether the location or the dynamic shift of attentional focus would determine the distortions effects (Experiment 3). The results showed that the effect size of the attraction effect was slightly larger than the repulsion effect and the preceding and following cues have independent influences on the perceived positions. The repulsion effect was caused by the location of attnetion and the attraction effect was due to the dynamic shift of attentional focus, suggesting that the underlying mechanisms for the retrospective attraction effect might be different from those for the repulsion effect.

Figure 1. Possible positions for the top line.

The top line appeared at one of eleven possible positions. The distance between possible positions was 0.1°. Five were at the left side of the bottom line. The others were at the right side of the bottom line. The leftmost (rightmost) position was 0.5° away from the bottom line in horizontal orientation. The top line was equally likely to appear in one of the eleven positions.

Figure 2. Schematic representation of the trial events in Experiment 1.

The fixation point appeared for 1000 ms and participants were instructed to keep their fixation on it. After a 100-ms blank, the cue and target were presented in two different conditions. Under the cue-target condition, the cue was presented for 50 ms. After a 150-ms blank, two vertical target lines were presented for 100 ms. Under the target-cue condition, the target was presented for 100 ms first, followed by a 100-ms blank. Then the cue was presented for 50 ms. The cue-target and target-cue SOAs (stimulus onset asynchrony) were always 200 ms in all conditions.

Figure 3. Results of Experiment 1.

The vertical axis represents the proportion of “right” response for the left-diagonal cue and “left” response for the right-diagonal cue in each possible target position both under cue-target and target-cue conditions. Positive values on the horizontal axis mean that the top line was at the right of the bottom line for left-diagonal cue and at the left of the bottom line for right-diagonal cue and vice versa. The point of subjective equality (PSE), defined as the intersection of the cumulative Gaussian curves with the line that marked P = 0.5, was −0.054° (dotted curve) for the cue-target condition and 0.096° (solid curve) for the target-cue condition.

Figure 4. Schematic representation of the trial events in the cue-target-cue condition in Experiment 2.

The fixation point appeared for 1000 ms. After a 100-ms blank, the cue and target were presented. The cue was presented for 50 ms. After 150 ms of blank, the target was presented for 100 ms. After another 100 ms of blank, the cue appeared for 50 ms again. The cue-target and target-cue SOA (stimulus onset asynchrony) were always 200 ms.

Figure 5. Results of Experiment 2.

The mean of PSEs in Experiment 2 was 0.045°. Pooled data provided the data points and fitting curves.

Figure 6. Schematic representation of the trial events of the double-cue condition in Experiment 3.

The cues appeared at the same time and duration as the target lines. The cue-target and target-cue SOA were always 200 ms.

Figure 7. Results of Experiment 3.

The positive values on the vertical axis indicated that the perceived location of the target was away from the cue (attentional repulsion effect) and negative value implied that the perceived location of the target was shifted toward the cue (attentional attraction effect). Error bars represented the standard error of mean.

Figure 8. Differential magnitudes of attentional shifts in single-cue and double-cue condition under target-cue temporal order in Experiment 3.

Thick and thin arrows represented strong and weak magnitudes of attentional shifts. Shifts of attention were attenuated in double-cue condition due to attention had already been attracted to cues' location at the first presented frame.