普通话儿童词汇相邻性多音节词表编制研究  

The development of Mandarin Multisyllabic Lexical Neighborhood Test

作　　者：张宁[1,2] 刘莎[1,2] 盛玉麒[3] 郭连生[1,2] 杨宜林[1,2] 徐娟娟[1,2] 

机构地区：[1]北京市耳鼻咽喉科研究所,首都医科大学附属北京同仁医院,北京100005 [2]耳鼻咽喉头颈科学教育部重点实验室(首都医科大学) [3]山东大学中文信息研究所,济南250100

出　　处：《中华耳科学杂志》2008年第1期30-34,共5页Chinese Journal of Otology

基　　金：国家自然科学基金资助项目(30371533);首都发展基金资助项目(首都ZD199906);北京市自然科学基金资助项目(7063074)

摘　　要：目的以心理语言学言语听辨领域的邻域激活模型（Neighborhood activation model,NAM）为理论指导,编制开放式的普通话儿童词汇相邻性多音节词表（Mandarin Multisyllabic Lexical Neighborhood Test,M-MLNT）,用于跟踪调查使用助听器或植入人工耳蜗儿童的言语识别能力的改善。方法根据邻域激活模型,邻居词汇数目较少但在口语中出现频率高的目标词定义为“易词”,邻居词汇数目较多但在口语中出现频率低的目标词定义为“难词”。计算儿童语料库中双音节词的出现频率和邻居词汇数目,获得频率和邻居数目的中位数,出现频率大于中位数而邻居数目小于中位数的词定为易词,出现频率小于中位数而邻居数目大于中位数的词定为难词。编制出汉语普通话儿童词汇相邻性双音节词表。选择3~5岁听力正常儿童30名为受试者,进行词表的初步验证。结果儿童日常口语语料库中包括2745个双音节词,出现频率从1至1878,中位数为2;目标词邻居数目从0至5,中位数为0。根据邻域激活模型定义出945个易词和223个难词。为使每个词表含有均衡的音位种类并难度相当,最终选择120词作为测试条目,编制易词表3张,每张20词,难词表3张,每张20词,共6张MLNT词表。经统计学分析,三张易词表间各测试条目的出现频率、邻居数、30名受试者测试得分均无显著性差异（P〉0.05）,三张难词表间各测试条目的出现频率、邻居数、30名受试者测试得分均无显著性差异（P〉0.05）,易词表和难词表间测试得分具有显著性差异（P〈0.01）。结论汉语普通话儿童词汇相邻性双音节词表（M-MLNT）设计过程中没有音素平衡的限制,测试条目更为儿童所熟悉。用于评价听力损失儿童的口语词汇辨识能力,有利于获得个体间的真实差异。M-MLNT可作为使用助听器或人工耳蜗植入儿童词汇辨识能力的长期追踪测试,用于评价助听装�Objective Base on the Neighborhood activation model （NAM）, to develop the Mandarin Muhisyllabic Lexical Neighborhood Test （M-MLNT）, which is exoected to be useful for evaluating speech perception performance in children with hearing loss. Methods Test items based on the lexical characteristics of word frequency and neighborhood density which addressed in the neighborhood activation model （NAM）. M-MLNT consists of two parts： （1） lexically ＂easy＂ words with high word frequency, which were low phonemically similar to the target word and （2） lexically ＂hard＂ words with low word frequency, which were high phonemically similar to the target word. Choose 30 children of 3-5 year old with normal hearing as subjects, to verify easy word and hard word lists. Results 2745 words for children contain 945 easy words and 223 hard words. Three easy word lists and three hard lists were developed to estimate the performance of word recognition among normal-hearing children. There are no differences among scores of three easy words lists （P 〉 0.05）, and no difference among scores of three hard words lists （P 〉 0.05）. But there are significant differences between scores of easy and hard words lists （P 〈 0.01）. Conclusion Without the phonetic balance constrain, more familiar and more conventional words can be selected as test items in M-MLNT. M-MLNT can be used to evaluate spoken word recognition of hearing-impaired children who use hearing aids and/or cochlear implants.

关 键 词：儿童 邻域激活模型 言语测试 

分 类 号：R764.04[医药卫生—耳鼻咽喉科] R764.5[医药卫生—临床医学]