How Simplified Language Affects Comprehension and Learning in Young Autistic Children (PALS)

How Single-Word and Telegraphic Simplification Affects Language Processing and Word Learning in Young Children With Autism Spectrum Disorder

The long-term study goal is to experimentally evaluate the components (and likely active ingredients) of early language interventions for young children with ASD. The overall objective is to determine how single-word and telegraphic simplification affects real-time language processing and word learning in young children with ASD (relative to full, grammatical utterances). The proposed project will investigate three specific aims: 1) Determine how single-word and telegraphic simplification affects language processing. 2) Determine how single-word and telegraphic simplification affects word learning. 3) Evaluate child characteristics that may moderate the effects of linguistic simplification on language processing and word learning. Aim 1 will test the hypothesis that children with ASD will process full, grammatical utterances faster and more accurately than single-word or telegraphic utterances. Aim 2 will test the hypothesis that full, grammatical utterances will support word learning better than telegraphic or single-word utterances. Aim 3 will test the hypothesis that language and cognitive skills significantly moderate the effects of linguistic simplification on language processing and word learning in young children with ASD.

language, autism, intervention, processing, learning, early childhood, treatment, early intervention, telegraphic, language input, looking-while-listening, eye tracking, comprehension, neurodevelopmental disorders

This study uses a within-participant experimental manipulation. All participants will be exposed to all utterance types (across trials).

Children will participate in screen-based language processing and word learning tasks in which they hear utterances with different types and amounts of linguistic simplification (i.e., a within-group manipulation).

LWL involves a child looking at a screen with two images, one on each side of the screen, while being provided an auditory cue that includes the name of the target image. The primary outcome is a dichotomous variable indicating whether the child is fixated on the target image or the distracter image; gaze directed at neither image will be considered missing. It will be determined whether gaze is to the left or right image, using trained coders who are blind to target side. Gaze location will be observed every 33 ms from 300 ms after onset to 2000 ms after target word onset, for a total of 51 observations per trial. Analyses will focus on differences across trial types and child characteristics in the trajectory of gaze location over the course of the trial.

A secondary outcome measure will be gaze location during the Teaching Phase of the Fast Mapping task (Study 2b). The variable is a dichotomous variable indicating whether the child is fixated on the novel image (only a single image is presented in each trial during the Teaching Phase) or not. It will be determined whether gaze is directed to the novel image, as our prior attempts to manually code single-object trials have not been successful. Only children who successfully calibrate and provide adequate automatic eye-tracking data in the Fast Mapping task will be included. Gaze location will be observed every 33 ms during the Teaching Trials. Analyses will focus on differences across trial types (single-word vs. full, grammatical teaching utterances).

Inclusion Criteria: Existing or suspected autism spectrum disorder, confirmed through ADOS-2 English as primary language 1-4 years old Exclusion Criteria: Known genetic condition (e.g., Down syndrome, fragile X) Cerebral palsy Acquired brain injury Uncorrected vision or hearing impairment

Dataset will include eye-gaze data, assessment scores, and transcriptions. De-identified analyzed data will be shared with other scientists for non-commercial research purposes in accordance with Michigan State University policy. Pending informed consent, de-identified data will be uploaded to the National Database for Autism Research, and transcripts will be shared with ASDBank. Scientific resources generated will be made available prior to publication if the work is different from our purposes. Following publication, resources will be made available to the broader scientific community on request. Findings and de-identified data will be shared through a variety of avenues, including scientific and clinical conferences, webinars, newsletters and scholarly manuscripts. Identifiable data may be shared if a family provides consent. All peer-reviewed manuscripts will be submitted to NIH's PubMed Central, to be made available at least 12 months after the official publication date.