Nasalance measures in Cantonese-speaking women.

Abstract

To establish and evaluate stimulus materials for nasalance measurement in Cantonese speakers, to provide normative data for Cantonese-speaking women, and to evaluate session-to-session reliability of nasalance measures. One hundred forty-one Cantonese-speaking women with normal resonance who were students in the Department of Speech and Hearing Sciences, University of Hong Kong. Participants read aloud four speech stimuli: oral sentences, nasal sentences, an oral paragraph (similar to the Zoo Passage), and an oral-nasal paragraph (similar to the Rainbow Passage). Data were collected and analyzed using the Kay Nasometer 6200. Data collection was repeated for a subgroup of speakers (n = 28) on a separate day. Nasalance materials were evaluated by using statistical tests of difference and correlation. Group mean (standard deviation) nasalance scores for oral sentences, nasal sentences, oral paragraph, and oral-nasal paragraph were 16.79 (5.99), 55.67 (7.38), 13.68 (7.16), and 35.46 (6.22), respectively. There was a significant difference in mean nasalance scores for oral versus nasal materials. Correlations between stimuli were as expected, ranging from 0.43 to 0.91. Session-to-session reliability was within 5 points for over 95% of speakers for the oral stimuli but for less than 76% of speakers for the nasal and oral-nasal stimuli. Standard nasalance materials have been developed for Cantonese, and normative data have been established for Cantonese women. Evaluation of materials indicated acceptable differentiation between oral and nasal materials. Two stimuli (nasal sentences and oral paragraph) are recommended for future use. Comparison with findings from other languages showed similarities in scores; possible language-specific differences are discussed. Session-to-session reliability was poorer for nasal than oral stimuli.