The Cost of Attentional Engagement: Target Switching during Visual Search in 2-Year-Old Toddlers with ASD

Background: Children with ASD exhibit over-focused attention at times and excessive distractibility at other times. The development of attentional factors that contribute to these seemingly contradictory behaviors are not clear. Earlier, we found that young toddlers with ASD are able to deploy their visual attention in a highly goal-driven manner (Blaser et al., 2014, Sci Rep; Smith et al., 2014, VSS), yet they may be inefficient in responding to changes in the environment (Greenaway & Plaisted, 2005, Psych Sci). 

Objectives: We investigated the cost of increased attentional engagement during visual search. We modified the classic visual search paradigm to test whether toddlers with ASD and age-matched typically developing toddlers (TYP) are able to switch their task goal from one search target to another.

Methods: 15 ASD and 28 TYP toddlers participated (ASD: 14 males, mean age: 28.3 months, MSEL Composite: 63.1; TYP: 11 males, mean age: 26.6 months, MSEL Composite: 103.5). Participants sat on their caregivers’ lap while a Tobii T120 eye-tracker recorded their eye movements. A block of 20 trials were presented: two familiarization trials, three single-feature training trials, then a sequence of 15 feature-conjunction test trials. Stimuli were 8 items defined by a feature-conjunction of color (isoluminant orange/green) and shape (apple/carrot). Feature-conjunction displays consisted of a target and a non-target (a green apple [A] and orange carrot [C]), plus 6 distractor items (green carrots and orange apples). Importantly, the 15 feature-conjunction trials were broken into three five-trial Phases defined by which object was highlighted as the target (counterbalanced across participants [ACA or CAC]). Each trial started with the target flying in and jumping up-and-down in the center of the screen (highlighting). Then, the search display appeared for 4 s, after which the target rotated back-and-forth for 2 s, accompanied by a sound effect; acting as both feedback and reward.

Results: In Phase 1, ASD and TYP toddlers found (fixated) the target more often than the non-target, and looked longer at the target than any other item in the display; an indication of task understanding. After the target switch (Phase 2), search performance of both groups decreased, and looking duration to the non-target increased. Interestingly, in Phase 3, when the target was again the same as in Phase 1, TYP toddlers recovered their initial search performance, whereas toddlers with ASD did not. This difference was also reflected in total looking durations.

Conclusions: Toddlers with ASD are capable of exercising goal-driven attention during visual search, yet this attentional focus appears to come at a price. ASD toddlers demonstrated an inability to flexibly change task goals, even when the initial task rules were reinstated. Our finding suggests that in young children with ASD, attention is over-focused at the level of task goal selection.