Attention to prosody (intonation) and content in children with autism and in typical children using spoken sentences in a computer game

Abstract This study validated a video game paradigm to explore attention to prosodic and linguistic components of spoken sentences in nine moderate-to-low functioning children with autism and impaired verbal skills. Nine typically developing children were also included. The children listened to pre-recorded sentences varying with respect to content (e.g., “Max ate a grape” vs. “Tom threw a ball”) and prosody (i.e., intonation of statement vs. question). During training, children learned to select one of two sentences differing in both content and prosody. At testing, children listened to stimuli consisting of recombinations of the content and prosodic features of the training stimuli. Testing performance indicated that the children with autism attended to the content and prosodic features of the training stimuli equally, whereas the children with typical development showed a clear preference for content over prosody. Both groups showed accurate discrimination of the training stimuli from the recombined test stimuli. The findings are interpreted in light of three approaches to explain the unusual attention patterns in autism: stimulus overselectivity, weak central coherence, and enhanced perceptual functioning.