An Acoustic Inspection of Vocalizations in Young Children with Autism Spectrum Disorders

Background: Few investigations have focused on the vocal quality used by children with autism spectrum disorders (ASD) during the prelinguistic stage of communication development. Previous research, using a limited range of vocal categories and not based on acoustic inspection, found that children with ASD produced significantly more syllables with atypical vocalizations than children with developmental delays (DD) (Sheinkopf et al., 2000). This finding could indicate a positive symptom for ASD thus enhancing the sensitivity of screening tools that currently rely heavily on negative symptoms. A core communication deficit in ASD is directing communicative behaviors to conversational partners, so this study also will explore relationships between the vocal quality and directedness of each vocalization.

Objectives: To reanalyze the data collected by Sheinkopf et al. (2000) by: 1) using acoustic displays and a more refined set of 20 categories to assess vocalizations produced by children with ASD and children with DD, 23-64 months of age, and 2) evaluating the relationship between vocal quality and communicative directedness.

Methods: The participants’ (9 DD, 10 ASD) vocalizations produced during administration of the Early Social-Communication Scales were digitized and viewed with software designed to display videos with concurrent spectrographic information. Measurements included two sets of codes: 1) dynamics of vocal fold vibration (“vibratory regimes”; Buder et al., in press) and 2) auditory impressions (e.g., pressed, breathy). The sets are mutually exclusive and non-redundant. Both sets incorporated inspection of acoustic displays, and the regime codes were produced with special attention to narrowband spectrograms. The video of each vocalization was observed and identified as being “directed to person” or “other”.

Results: Preliminary findings suggest that children with ASD produce different phonatory regimes and vocal qualities than children with DD.

Conclusions: Implications for early diagnostic procedures and acquisition of socially appropriate voice will be discussed.