Dynamic Assessment of the Looking Patterns of Toddlers with ASD during Teaching

Friday, May 15, 2015: 11:30 AM-1:30 PM
Imperial Ballroom (Grand America Hotel)
A. Bean Ellawadi1, D. A. Fein2 and L. Naigles3, (1)The Ohio State University, Columbus, OH, (2)Psychology, University of Connecticut, Storrs, CT, (3)University of Connecticut, Storrs, CT

Tek and colleagues’ (2008) investigation of the emergence of early language learning biases in toddlers with ASD revealed that, although both toddlers with ASD and their typically-developing (TD) peers demonstrated a noun bias (that novel words refer to objects rather than actions), the toddlers with ASD did not demonstrate a shape bias (that novel labels are extended to objects of similar shape).  


In this study we compare the dynamic looking patterns of toddlers with ASD during the teaching trials of the noun and shape bias experiments to determine (a) whether there are differences in the average number or duration of fixations during teaching and (b) whether these differences are related to performance at test. 


This sample (from Tek et al. 2008) consists of 17 toddlers with ASD (MA = 33.10 months) and 16 TD toddlers (MA = 20.69 months) matched on expressive language.  The participants were seen for 2 visits, 4 months apart.  The experiments were presented using the Intermodal Preferential Looking paradigm (Naigles & Tovar, 2012).  Children’s frame-by-frame eye movements were coded during the teaching trials.  The average number of fixations (fixations) and length of the longest fixation  (duration) were calculated. 

Results:   Repeated Measure ANOVAs were conducted.  For the noun bias experiment the dependent variables were fixation and duration and the independent variables were group and time.  There was a main effect of group for the fixations (F (1, 29) = 6.61, p=.02); the ASD group produced significantly more fixations than the TD group.   No other results were significant.  For the shape bias experiment the dependent variables were fixation and duration and the independent variables were group, time, and condition (name versus no name).  There was a main effect of time for the fixations (F (1, 30) = 8.95, p<.01) with fewer fixations at Time 1 than 2, and for duration (F (1, 30) = 35.96, p<.001), with shorter durations at Time 2 than 1.  There was also a condition by group interaction (F (1, 30) = 6.2, p=.019); the TD group demonstrated a larger difference in looking durations across conditions as compared to the ASD group.   Correlation and regression analyses were conducted to investigate whether looking patterns were related to shape bias test performance.  For the TD group a significant correlation emerged between fixations and duration during teaching and test at Time 1; however, neither fixations nor duration predicted test performance.  For the ASD group there was a significant correlation between duration during teaching and test performance at Time 2.  The regression trended towards significance.          


Although differences in fixation during teaching were noted across experiments, only differences in duration were observed in the experiment testing the shape bias (the bias the ASD group did not demonstrate at test).  The group by condition interaction suggests that for the TD group hearing a label directs their attention to relevant information, which yields more efficient processing of stimuli.  Labels do not appear to have the same facilitative effect for the ASD group.