Quality of remotely-collected gaze data in autistic and nonspectrum children

Many developmentalists have shifted to remote research. This project uses secondary data to evaluate the quality of eye-gaze data from 30 autistic children and a language-matched sample of 30 nonspectrum children (mean ages 48 and 27 months, respectively). All children completed an experimenter-moderated preferential looking paradigm via Zoom. Frequency of co-occurring child and household events, rates of missing data, and percent agreement between gaze coders were assessed. Results indicated that co-occurring events were minimal, with no diagnostic group differences. Missing data rates were low overall and were unrelated to diagnostic group, age, or language level of participants; however, higher rates of co-occurring child behaviors were associated with higher rates of missing data. Agreement between coders for eye gaze data was comparable to in-lab studies. Results affirm the usefulness of remote, experimenter-moderated gaze-based research with autistic and nonspectrum children.