一组海南话双音节言语测听词表的编制

Development of Disyllable Lists for Hainan Dialect Speech Audiometry

目的编制一组海南话双音节言语测听词表。方法根据常用原则,参照当地报刊杂志与地方志,选出海南话中常用的双音节词652个,由熟悉海南话的播音员诵读并进行录制。去掉录音效果差(过度削峰、有明显杂音)的词,最终由500个词组成10张词表,每表50词。通过对20名听力正常受试者的测试,记录每个词在2、6、10、14dB HL四个强度的正确识别率。用Statistica11.0统计软件拟合出每个词的P-I曲线,并计算出阈值和阈值处斜率。根据各词阈值与平均阈值的差值,对每个词的声级都进行强度调整,使得每个词的识别阈都一样。结果从中筛选出300个双音节词,这些词的阈值和斜率均服从正态分布,编制出6张词表,每表50个词。结论一组海南话双音节言语测听词表已编制完成,但仍需进行词表间等价性验证。

Objective To create a set of Hainan dialect speech audiometry materials.Methods Six hundred and fifty two disyllables commonly used in Hainan dialect were identified,read and recorded by an announcer familiar with Hainan dialect.After removing poorly recorded words(excessive peak clipping or noise),500 words were placed in 10 phoneme balanced word tables,each containing 50 words.These were tested in 20 normal hearing subjects,and the rate of correct recognition of each spondee word at 2,6,10 and14 dB HL was recorded.The Statistica11.0 statistical software was used to fit the P-I curve for each word and to calculate the threshold and slope of threshold.The sound level for each word was adjusted and rectified for recognition threshold based on the difference between the average threshold and the threshold of each spondee word.Result Three hundred disyllable words with threshold and slope both in normal distribution were selected to generate 6 glossaries.Conclusion A set of Hainan dialect speech audiometric table has been generated,although the equivalence between individual glossaries remains to be confirmed.