A Child-Friendly Eye-Tracking Paradigm Reveals Impaired Implicit Repetition Learning in Adults and Children with ASD

Background: In the domain of memory adults with ASD tend to experience difficulties in repeating back sequences of stimuli in their order of presentation (Poirier et al., 2011; Gaigg et al., 2014). When tasks require participants to respond as quickly as possible to repeating sequences of stimuli children with ASD sometimes demonstrate typical learning rates (Barnes et al., 2008; Brown et al., 2010; Travers et al., 2010) and sometimes not (Motofsky et al., 2000; Gordon & Stark, 2007). This pattern is of considerable interest because explicit and implicit serial order memory processes play a critical role in children’s language development (e.g., Ullman, 2004). It is therefore possible that individual differences in this domain contribute to the heterogeneity in language development in ASD. To examine this issue fully it is necessary to device methods that can be used to study serial order memory in children of any age and ability level.

Objectives: Our primary objective was to develop a novel eye-tracking paradigm that is suitable for examining serial order memory processes in children and adults of any age and ability level.

Methods: Two experiments are reported that involved 38 adults (19 ASD; 19 TD) and 54 children (31 ASD; 23 TD) respectively. The adults in experiment 1 all had IQs within the typical range whereas experiment 2 included children with and without substantial learning difficulties. In both experiments participants were asked simply to ‘watch out for’ a rabbit that appeared in a repeating sequence of 5 (out of a possible 8) rabbit holes on a screen. A Tobii eye-tracker was used to monitor gaze latencies to the rabbit and to quantify the number of times participants anticipated the rabbit in a particular location.

Results: In both experiments gaze latencies to the rabbit and anticipatory fixations indicated significantly reduced implicit repetition learning in the ASD groups. Importantly, and in line with observations from explicit serial order memory tasks, both groups appeared to learn in which of the 8 locations the rabbit was likely to appear. In other words, both groups anticipated the rabbit progressively more frequently in one of the 5 locations that formed part of the sequence, even if they did not anticipate it in the correct location at the correct time. Moreover, both groups also demonstrated reliable primacy effects, whereby gaze latencies to the first rabbit in the sequence decreased significantly more over successive trials than gaze latencies to the other rabbits in the sequence.

Conclusions: Both experiments extend the findings from explicit serial order memory tasks (e.g., Poirier et al., 2011) in showing that adults and children with ASD experience relatively specific difficulties in learning about the order of events. Moreover, the observations suggest that eye-tracking paradigms such as these will proof fruitful for examining the developmental trajectory of serial order memory processes in ASD in the future.