CHARACTERISTICS OF VOCAL INTERACTIONS BETWEEN PARENTS AND CHILDREN WITH AUTISM SPECTRUM DISORDER IN RELATION TO HIERARCHICAL TEMPORAL CLUSTERING

Abstract

The current study explores the vocal interactions of parents and children with and without Autism Spectrum Disorder (ASD) through a novel measure of hierarchical acoustic clustering in order to better understand the specific characteristics of parent and child speech that differ within ASD communication. Interactions between the parent and child were video-recorded and coded for the frequency and timing of parent and child vocalizations. To measure hierarchical temporal clustering, audio recordings of the dyadic interactions were analyzed across twelve timescales ranging approximately from the phoneme-level scale to the phrase-level scale and quantified using Allan Factor (AF) variances. There were three main findings of the study. First, significant relationships were found between frequency of interpersonal turns and ASD toddler language and developmental assessment scores. Second, ASD dyads exhibited significantly greater hierarchical temporal clustering compared to TD dyads. Third, the vocal characteristics which most correlated with hierarchical clustering in ASD dyads were frequency of total vocalizations and turn-taking, particularly when considering total interpersonal turns and total turns. These findings call attention to the importance of turn-taking in communication, and the reduced quality of turn-taking in interactions of parents and children with ASD.