Developing an Observational Measure of the Flexible Use of Pre-Linguistic Vocalizations in Preverbal Children with Autism

Background: Why approximately thirty percent of children with Autism Spectrum Disorder (ASD) persist in using minimal spoken language despite intervention is unknown. Previous studies have suggested that differences in pre-linguistic skills are useful in predicting the extent to which children with ASD will develop spoken language. A foundational pre-linguistic skill that has not been investigated in children with ASD who are not using spoken language (i.e., preverbal children with ASD) is the ability to combine vocalizations with different types of affect (i.e. pre-linguistic vocalization functional flexibility).

Objectives:  (1) To determine if preverbal children with ASD demonstrate the full range of vocalizations and affect evident in language-matched TD peers, and (2) To develop a measure of pre-linguistic vocalization functional flexibility (PVFF) in preverbal children with ASD to ascertain whether children with ASD are able to use vocalizations with affect flexibly.

Methods: This study used an existing longitudinal data set from The Useful Speech Project (e.g., Yoder, Watson & Lambert., 2015). The children with ASD enrolled in this study were between 20-48 months at baseline and were examined at five total measurement points separated by four month intervals. For the present study, Time 2 multimedia recordings of each child were selected during two different structured assessments with built-in communication presses: Communication and Symbolic Behavior Scales and Early Symbolic Communication Scale. A partial-interval coding scheme was used to code these multimedia recordings for the occurrence of vocalization and affect types. Vocalizations were classified in five categories based on pitch and phonation: squeal, vocant, growl, cry and laughter. Facial affect was categorized as positive, negative or neutral.

Results: The behavioral coding method developed for this study was found to be reliable for both vocalizations (Protophones: mean: 0.88 and range: 0.80-0.98) and affect (mean: 0.92 and range: 0.67-1) and stable (R= 0.97 for protophones, R=0.85 for negative affect, R= 0.85 for neutral affect, and R= 0.93 for positive affect). A significant minority demonstrated limited variety in vocalizations and affect. Six participants (30%) did not demonstrate the full range of protophones, and five participants (25%) did not demonstrate the full range of affect. PVFF was operationalized using operant contingency values (OCV) to quantify the degree of association between vocalization types and affect types. The analysis revealed predominately negative OCVs for neutral affect and protophones in this sample (i.e., vocalizations were more likely to be produced in the absence of neutral affect). OCVs for negative affect and protophones revealed predominately positive values (i.e., vocalizations were more likely to be produced with negative affect than chance).

Conclusions: The results of the analysis indicate that preverbal children with ASD demonstrate limited flexibility in their use of vocalizations with different affect. In a social communication context, this could limit the types of messages that children with ASD are sending to communication partners. If these results are replicated, PVFF could be integrated as an early target in an intervention, due to its early developmental emergence.