Stability and validity of an automated measure of vocal development from day-long samples in children with and without Autism Spectrum Disorder

Individual difference measures of vocal development may eventually aid our understanding of the variability in spoken language acquisition in children with autism spectrum disorder (ASD). Large samples of child vocalizations may be needed to maximize the stability of vocal development estimates. Day-long vocal samples can now be automatically analyzed based on acoustic characteristics of speech likeness identified in theoretically driven and empirically cross-validated quantitative models of typical vocal development. This report indicates that a single day-long recording can produce a stable estimate for a measure of vocal development that is highly related to expressive spoken language in a group of young children with ASD and in a group that is typically developing.