Objectives: Study 1 explores whether eye-tracking can be used to test children’s comprehension of nouns. Study 2 investigates the comprehension of verbs and the ability to use verbs to anticipate a direct object.
Methods: Children (ages 2;11-6;0) with and without autism participated in these experiments; within the autism group, there was a wide range of language abilities on the PPVT. Each child completed the PPVT, while the parent filled out a shortened version of the MCDI. The children then sat in front of the TOBii eye-tracker, which recorded their eye-movements as they looked at two pictures while listening to prerecorded sentences. In Study 1, the sentence said “Look at the [target word]!”, while in Study 2, the verb varied, sometimes biasing toward the target noun (“Drink the coffee!”) and sometimes remaining ambiguous (“Touch the milk!”). Analyses were conducted to examine looking-time and latency to look at the target picture.
Results: Both Study 1 and Study 2 showed that children look longer at the target picture than at the foil, but only for words that parents reported as known (MCDI). Study 2 also demonstrated that typically-developing children were able to use the semantics of verbs to predict the direct object, resulting in a shorter latency to target noun, while the children with autism did not show this pattern, regardless of PPVT score.
Conclusions: Eye-tracking may be a promising way to test language comprehension in children with autism. Further, it may elucidate ways in which children with autism may have impaired processing, such as their lack of anticipatory looking in response to verbs.