Poster #53 - Comparison of Autistic and Typically Developing Children’s Vocalizations via the International Phonetic Alphabet

• In Event: Poster Session 10

Abstract

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder that is characterized by social communication challenges (American Psychiatric Association, 2013). Atypical vocalizations may serve as an early indicator of ASD, and the International Phonetic Alphabet (IPA) provides a method for distinguishing these speech-related vocalizations. The IPA utilizes symbols to represent each individual speech sound of a word (e.g., f/i/sh→fɪʃ). Using IPA, studies have found that toddlers with ASD produce fewer speech sounds than typically developing (TD) children (e.g., Moffit et al. 2022). However, heterogeneous language profiles have been identified among children with ASD, such that some children are low verbal (i.e., language is below age-appropriate language years), middle verbal (i.e., language is delayed compared to their TD peers), or high verbal (i.e., language is comparable to their TD peers; Tager-Flusberg, Paul, & Lord, 2005). The objectives of the current study are 1) to compare the vocalizations of consonants and vowels produced by children with ASD and TD children with a new dataset, and 2) to determine whether the International Phonetic Alphabet (IPA) has the ability to capture the heterogeneity within ASD.

Twelve participants (4 TD, 4 low-verbal with ASD, and 4 middle-verbal with ASD) were selected from a longitudinal study tracking the language abilities of children with ASD and TD children (Naigles & Fein, 2017; see Table 1). Recordings from 30-minute parent-child play sessions were transcribed via IPA, with the transcriber listening for distinct phonemes produced by the child. Consonant (e.g., cat→kæt) and vowel (e.g., book→bʊk) types and tokens were tallied.

Descriptive and qualitative analyses of the data yielded the following results: TD children produced more consonant types (M = 9.25, SD = 4.92) than vowel types (M 5.50, SD = 2.38). The middle-verbal children also produced more consonant types (M = 15.5, SD = 7.59) than vowel types (M = 9.00, SD= 3.37). On the other hand, low-verbal children produced roughly the same number of consonant types (M = 3.00, SD = 2.16) and vowel types (M = 3.25, SD = 1.71; Figure 1a). In terms of consonant and vowel tokens, middle-verbal children produced more consonant (M = 97.25, SD = 104.01) and vowel (M = 85.5, SD = 87.88) tokens than the TD and low-verbal children. Furthermore, the TD children produced more consonants (M = 54.75, SD = 38.96) and vowel (M = 37.25, SD = 33.95) tokens than the low-verbal children (Figure 1b).

Overall, this study replicates previous findings that TD children produce more consonants than vowels (e.g., Woynaroski et al., 2017). Within the ASD group, however, middle-verbal children also produced more typical vocalizations (i.e., more consonants than vowels), as compared to the low-verbal children, highlighting the variability within ASD. These preliminary patterns suggest that transcribing via IPA has the ability to capture differences between children with ASD and TD children, as well as the heterogeneous language profiles within ASD at an early age. Future directions include transcribing and coding the vocalizations of more children per group.

Authors

• Sarah Marie Cintron, University of Connecticut
  Presenting Author

• Cynthia Boo, University of Connecticut

• Letitia R Naigles, University of Connecticut