摘要
目的对自行编制的单音节言语识别测听材料的录音版本进行等价性临床验证,以获得良好的测试信度。方法对编制好的30张单音节字表(每张表包含25个单音节字)进行数字化录音和处理,调整各测试表的均方根强度。选择60名听力正常人进行等价性实验。每位受试者在6个不同给声强度以随机顺序聆听30张字表。每张字表在言语听力级-8dB、-2dB、4dB、8dB、14dB、20dB等6个言语听力级上进行测试,每个强度测试10个不同的受试者。以百分比记录60名受试者对各张字表的正确识别率,通过logistic曲线拟合绘出每张字表的识别得分随言语声强度级变化的PI曲线。对30张字表的正确识别率进行双因素方差分析和Post-Hoc两两比较。结果筛选出22张相互间等价性较好的汉语普通话单音节测试字表(P=0.175)。
Objective To evaluate the psychometrical equivalence of a set of monosyllable word lists which we have developed in order that the lists have good reliability in speech discrimination test. Methods 30 lists （25 monosyllabic words in each） were recorded and post-processed digitally. RMS intensity of each list was adjusted to an average level. Six presentation levels were selected for evaluation： -8 dB、 -2 dB、 4 dB、 8 dB、 14 dB and 20 dB speech Hearing Level. 60 subjects with normal hearing served in our equivalent test. One list was presented at each intensity level. The orders of the presentation of the lists were randomized for each subject. Each list was presented an equal number of times at each intensity level across the entire subject population, and each subject listened to all the 30 lists one time at 6 presentation levels. Each response had to match the presented stimuli in both lexical tone and pronunciation to be scored as correct. Original data were expressed as percentile scores and a Performance-Intensity function for each list was plotted using logistic regression to obtain regression slope and 50% recognition threshold level. Two-factor ANOVA and Post-Hoc test were used to compare the recognition scores of the 30 lists. Results 22 lists were considered to be primarily equivalent with each other（P=0.175）. A further work was to be done to get a set of test materials with more reliability.
出处
《中华耳科学杂志》
CSCD
2008年第1期17-20,共4页
Chinese Journal of Otology
基金
科技部“十一五”支撑计划(2007BAI18B12)
关键词
测听法
言语
普通话
Audiometry
Speech
Mandarin
