16593
Goal Anticipation in Toddlers with ASD and High-Risk Siblings of Children with ASD

Thursday, May 15, 2014
Atrium Ballroom (Marriott Marquis Atlanta)
S. Thomas1, J. Parish-Morris2, K. Spielman1, E. N. Cannon3, A. L. Woodward4, J. Pandey1, R. T. Schultz1 and S. Paterson1, (1)Center for Autism Research, The Children's Hospital of Philadelphia, Philadelphia, PA, (2)University of Pennsylvania and Children's Hospital of Philadelphia, Philadelphia, PA, (3)University of Maryland, College Park, College Park, MD, (4)University of Chicago, Chicago, IL
Background: Typically developing infants interpret actions by assigning intentions and goals to human characters, but not to inanimate objects, such as grasping claws (Cannon & Woodward, 2011). Although some research suggests that children with ASD are able to recognize goals and intentions (Carpenter et al., 2001; Hamilton, 2009), other findings suggest that their ability to use social information about intentionality in the service of other goals such as motor imitation or word learning is impaired (Parish-Morris et al., 2007; Somogyi et al., 2013). With few exceptions, studies of intentionality in ASD focus primarily on middle childhood; little research has explored the development of goals and intentions in toddlers.

Objectives: Using an established paradigm developed by Cannon and Woodward (2011), we aim to determine whether toddlers with Autism Spectrum Disorder (ASD) and those at high-risk for ASD attribute goals to a reaching action performed by a human arm vs. a mechanical claw.  

Methods: Toddlers with autism (N=8) and siblings of a child with autism (N=21; Mean age=30.80 months, range: 23.9-39.6) were randomly assigned to watch a short video of a reaching action performed by either a hand (N=16) or a mechanical claw (N=13). Gaze data was recorded using a Tobii infrared eye tracker. Four identical blocks included 3 familiarization trials (hand or claw reaches toward one of two toys) and 1 test trial each. In test trials, the toys switched places, and the video showed the hand or claw beginning to reach, but stopping short of choosing an object. An experimenter blind to diagnostic status coded children’s anticipatory looks during the test trials.  

Results: A multivariate GLM controlling for nonverbal mental age revealed that, in contrast to prior research with typical infants, toddlers across both groups looked equally at the goal object and the location of the familiarized reach during test trials in both conditions, F(2,25)=2.28, p=0.14. This suggests that children with ASD and high-risk siblings do not demonstrate differential goal attribution to human actors vs. mechanical claws, which is consistent with prior studies showing reduced attention to social information and less automated interpretation of actions as goal-oriented. Thus, this finding could relate to symptoms of ASD and features of the broader autism phenotype. An alternative explanation is that older children simply do not make anticipatory looks to goal objects in this paradigm; data on matched typical toddlers is currently being collected to tease apart these two possibilities. 

Conclusions: These results have implications for how we understand the development of social cognition and social skill in ASD and siblings at risk, as understanding goals and intentions is a building block for later theory of mind. Current data collection with a typically developing control group of the same age and developmental level will shed further light on our findings.