Assessment of cleft palate articulation and resonance in familiar and unfamiliar languages: English, Spanish, and Hmong.

Abstract

Linguistic diversity is increasing in the patients seen for cleft palate treatment and there are not enough providers who speak multiple languages. There are no published studies which directly investigate the ability to assess cleft palate articulation and resonance in a language not spoken by the examiner. The aim of this study was to determine whether listeners could make accurate judgments about articulation and resonance in languages they do not speak and to determine how experience level and familiarity with a language affect these ratings. Binary (presence/absence) and visual analog scale (VAS) judgments were obtained for hypernasality, misarticulations, speech acceptability, and overall velopharyngeal dysfunction (VPD) of English, Spanish, and Hmong samples from naïve listeners, generalist speech-language pathologists (SLPs), and specialist SLPs. The speech samples were obtained from 22 speakers, nine with a history of VPD and 13 controls. The ratings were completed by 24 native English listeners, eight at each level of experience (naïve, generalist SLP, specialist SLP). Overall, the listeners were more accurate for determining the presence/absence of misarticulations, speech acceptability, and VPD in English compared to Hmong. Hypernasality and VPD ratings in English were more accurate than in Spanish and ratings of misarticulations were more accurate in Spanish than Hmong. VAS ratings of hypernasality were highly correlated with the nasalance values from oral phoneme reading passages. Statistically significant correlations were present for overall and group ratings in English. Less consistent correlations were observed in Spanish and no significant correlations were present in Hmong. Overall, listeners judged English ratings to be easier to make, and were made with more confidence, compared to Hmong. In addition, the SLP specialists tended to find the ratings in all languages easier to make and were more confident than naïve listeners. Many of the expected differences for ratings based on listener experience and language familiarity were observed. There were advantages for all listener groups in English when compared to Hmong. These differences were inconsistent and weaker when Spanish was compared to English. The experience advantage for listeners was most apparent in English and Hmong.