Perceptual and acoustic analysis of speech intelligibility in Mandarin-speaking young adults with cerebral palsy

This research investigated the relationship between specific acoustic features and perceptual judgements of speech intelligibility for utterances produced by 20 Mandarin-speaking young adults with cerebral palsy. Minimal phonetic contrasts were used to assess the relationship between acoustic features and intelligibility. Selected phonetic contrasts included two vowel contrasts (vowels differed in height and frontness) and four consonant contrasts (consonants differed on aspiration-unaspiration, affricate-fricative, stop-nasal, and place of stop articulation dimensions). Perceptual judgements from both item identification and scaling tasks reflected the diversity of speech intelligibility among these young adults with cerebral palsy. Individual phonetic contrasts correlated well with overall intelligibility; however, a multiple regression analysis revealed that aspiration-unaspiration, affricate-fricative and vowel frontness contrasts accounted for 99% of the variance in judgements of intelligibility. Seven acoustic parameters were examined (i.e., F1, F2-F1, VOT, friction duration, possibility of initial burst, nasality, and burst spectrum). Results indicated that cerebral-palsied speech was more varied than speech produced by normal age-matched subjects. F1, F2-F1, nasality scaling, and burst spectrum differentiated the acoustic patterns of the cerebral-palsied speakers from the normal speakers. Moreover, a multiple regression model predicted the intelligibility score with 74.8% of the variance accounted for by F2-F1, VOT, and possibility of initial burst. These findings are discussed in relation to production deficiencies and implications for improving intelligibility in the population of speakers with cerebral palsy.