Acoustic Measures of Word-Level Prosody in Childhood Apraxia of Speech: An Initial Validation Study

Le résumé fourni par la source

BACKGROUND: The "gold standard" of childhood apraxia of speech (CAS) diagnosis is expert clinical judgment based on perception of core features: inconsistent errors, impaired transitions, and impaired prosody. However, this standard has several limitations, which may be addressed with acoustic measures. PURPOSE: This retrospective study aims to provide initial validity evidence for nine acoustic measures of prosody and examine lexical stress production in CAS. METHOD: The study involves 33 children with CAS (4-8 years) imitating six bisyllabic words (three strong-weak, three weak-strong). For each word, nine acoustic measures of prosody were obtained: three pairwise variability index measures, two lexical stress ratio measures and their three component ratios, and word syllable duration. To address construct validity, we examined effects of stress pattern, age, and severity. To address convergent validity, we correlated acoustic measures with each other and with clinical-perceptual judgments of prosody. Finally, we examined the degree to which children with CAS differentiated stressed and unstressed syllables. RESULTS: Findings revealed that acoustic measures differed between stress patterns. Six of the measures had moderate and significant correlations with CAS severity for strong-weak words but not weak-strong words, and none of the measures correlated with age. All acoustic measures showed moderate or strong correlations with each other for strong-weak words but only some did for weak-strong words. None of the measures correlated significantly with clinical-perceptual measures of prosody. Children demonstrated equal stress on most measures for strong-weak words but clear evidence of stress differentiation for weak-strong words. CONCLUSIONS: This study provides qualified initial support for the validity of all acoustic measures. Findings replicate and extend prior research to show that children with CAS may have difficulty with production of lexical stress. Prospective research is needed to control for stimulus features with a larger sample that includes a range of diagnoses. SUPPLEMENTAL MATERIAL: https://doi.org/10.23641/asha.29289032.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.