Typical Patterns of Visual Filtering Among Children with ASD with Dynamic Presentation of Targets and Flankers

Background: Visual filtering is the ability to selectively focus on important information while ignoring task-irrelevant information. Filtering efficiency is often studied with some variation of the flanker paradigm (Eriksen & Eriksen, 1974), a computer-based task on which participants are asked to quickly and accurately identify a target among flanking stimuli. Among adults with ASD, impairments in both the narrowing (Burack, 1994) and broadening (Mann & Walker, 2003) of attention have been reported. In fine-tuned analyses of the effect of stimulus-distractor compatibility on performance, children with ASD showed a typical flanker compatibility effect (FCE) as flankers that were physically compatible with the target were associated with faster and more accurate response times (RTs), whereas incompatible flankers were associated with slower and less accurate RTs (Henderson et al, 2006; South et al., 2010). However, the implications of the traditional flanker task are limited as it entails the simultaneous presentation of the flankers and targets, thereby failing to depict the dynamic real-world in which distractions may occur before, during, or after the appearance of an attended-to stimulus or event.

Objectives: To investigate the spatial and temporal modulation of visual filtering among children with ASD in a study of the effects of asynchronous onset of visual stimuli on target identification.

Methods: Fifteen children with ASD and 15 MA-matched TD children completed a modified version of the flanker task. The targets appeared at the centre of a computer screen with flankers to their left and right either before (-200, -400 ms), simultaneous with (0 ms) or after (200, 400 ms) the target. Feature compatibility was also varied. In the Response Compatibility condition, the flankers and target were from the same response (RC; response compatible) or opposite response (RI; response incompatible). In the Feature Similarity condition, the flankers and target either shared a feature (FS; feature similar) or did not (FD; feature dissimilar). Response compatibility and feature similarity were varied across asynchronies. The dependent measure was response time (RT) and percentage of error (POE).

Results: The TD children responded faster than the children with ASD although a similar pattern of filtering emerged for both groups. The presentation of flankers before the target led to faster RTs, whereas the presentation of flankers after the target led to slower RTs. The highest POEs were committed when the flankers were RI or FD with the target. The presentation of the flankers after the target was associated with the smallest POE, whereas the flankers presented before and simultaneous with the target were associated with a higher POE.

Conclusions: Similar to the TD children, the children with ASD showed faster RTs, but reduced accuracy, when the flankers cued the onset and location of the target. Slower RTs were found when flankers appeared abruptly after the target along with increased accuracy. The findings suggest that, despite slower RTs, the children with ASD showed similar patterns to the TD children in their performance on a task of visual filtering with a dynamic presentation of stimuli.