Gaze Following Atypicalities in Children with Autism: The Role of Motivation

Background:  Although impairments in joint attention are common in ASD, several studies have found intact gaze following accuracy. In  a recent study (Falck-Ytter et al, 2014), we found that typically developing (TD) children displayed a first fixation duration (FFD) bias towards an object being attended to by another person, as compared to an unattended object. In an ASD-sample, no such FFD bias was found. This suggests that the FFD measure might detect subtle gaze following atypicalities that are not reflected by the more common accuracy measure. Given the explorative nature of the previous study, independent replication is required. More knowledge about the mechanisms underlying the altered response in ASD is also needed.

Objectives:  The aim of the present study was twofold. First, we wanted to replicate the previous finding of a weaker processing bias for attended vs. unattended objects in ASD as compared to TD-children. Second, we aimed to explore the potential role of motivational factors.

Methods:  The study included a group of high functioning children with ASD (n = 16, mean age = 81.6 months) and a group of typically developing children (n = 18, mean age = 73.6 months). Using eye tracking technology, the children’s gaze patterns were recorded as they watched videos of a model looking at one of multiple objects. In the Circumscribed Interest (CI) condition, the objects were model trains and toy cars (common circumscribed interests, expected to be particularly motivating for children with ASD). The objects in the non-CI condition were plants. Two measures were used: (1) Gaze following accuracy was measured as a difference score (DS) with the number of incongruent gaze shifts subtracted from the number of congruent gaze shifts; (2) FFD bias was measured as a DS with the mean FFD at the unattended objects subtracted from the mean FFD at the attended object. 

Results:  The groups did not differ in terms of gaze following accuracy in neither condition, and no interaction effect was found (p:s > 0.05). More importantly, the previous finding of a reduced FFD bias for attended objects in ASD was replicated in the non-CI condition (p < 0.05). However, in the CI condition, no group difference was found.  A 2x2 ANOVA confirmed that the manipulation modulated the responses differently in the two groups (p < 0.05). 

Conclusions: The previously found greater FFD bias for attended objects in a TD- as compared to ASD-sample was replicated in the non-CI condition, suggesting that the FFD measure reliably tracks gaze following atypicalities in children with ASD. However, no group difference was found in the CI-condition, showing that the altered FFD response in ASD does not generalize across all object types. Rather, this demonstrates that when objects known to be particularly interesting for children with ASD are included, the performance of these children does not differ from that of their typically developing peers. This finding speaks to the importance of motivational factors in joint attention behaviors in ASD.