Endogenous Visual Orienting with and without Saccades in Autism Spectrum Disorder: An Eye-Tracking Study

Background:

Endogenous visual orienting is the allocation of visual attention resources to a spatial location driven by symbolic cues or internal goals. Previous research suggests that people with ASD display differences in their use of symbolic cues in endogenous orienting compared with people who do not have ASD, and that temporal aspects of task performance may play a significant role in group differences (Landry et al., 2009; Landry & Parker, 2013). 

Objectives:

The study examined the temporal properties of endogenous orienting in the context of traditional endogenous orienting with a button press response, and with eye-tracking.

Methods:

Participants with ASD (n = 16, Males = 13, mean age = 10.79 years, age range = 6 to 21 years) and typically developing participants (TD) (n= 16, Males = 12, mean age = 10.92 years, age range = 6 to 22 years) completed the tasks. Both groups were matched on raw scores for the Raven’s Progressive Matrices, age and gender. Two experimental tasks were completed: a traditional Posner (1980) arrow cuing task with forced-choice button press responses, and an overt orienting task in which participants were instructed to saccade to the target. Both tasks included valid (75%) and invalid (25%) cues at different cue-target SOAs (150ms, 300ms, 450ms, 600ms, 750ms, 900ms). To maintain task engagement, the saccade task also included the instructions that participants should press a button if the target was an animal (and not press for an object). Eye-tracking was performed during both tasks. This was to ensure participants maintained central fixation during the traditional button press task, and to measure saccades during the saccadic overt orienting task.

Results:

A 2x2x6 Mixed ANOVA examined orienting effects (calculated as: (invalid-valid) / average RT across all trials). No differences were found between ASD and TD participants in the magnitude of orienting overall (ME Group p=.46), nor as a function of SOA (Group x SOA p=.33), nor did they differ in orienting as a function of whether the task was the traditional button press Posner task or a saccadic overt orienting task (Task x Group x SOA p=.24); Figure 1 shows the patterns of orienting for both groups. A 2x2x6 Mixed ANOVA was used to examine saccadic velocity during the saccadic overt orienting task. A significant Group by SOA interaction was found, F (3, 31) = 8.45, p<.001; Figure 2 shows the different SOA patterns shown by the two groups.

Conclusions:

In this sample, we did not replicate previous findings of differential orienting performance for ASD. We found very similar patterns of orienting across 6 SOAs in both a traditional button-press arrow-cued endogenous orienting task and when saccadic RT was measured in overt shifts following endogenous cues. We found different patterns of saccadic velocity across SOA between ASD and TD participants. This finding is intriguing, as orienting performance did not differ between groups on the saccadic task, despite different patterns of saccadic velocity across SOA between groups.