普通话语句测听句表等价性测试

目的对普通话测听词汇表(mandarin speechtest materials,MSTM)中的语句测试材料进行等价性测试,以期得到难度平衡的若干张句表。方法选取听力正常青年受试者61名(61耳)采用听说复述方式,进行30组句表(每组10句)的测试,以关键词记分,然后对各表得分进行统计学分析。结果30组句表大体上可分为三套,初步认定两套29组句表(290句)分别具有等价性。结论初步认定经实验筛选出难度平衡的两套29组句表可以用于临床言语评估测试。     

听力正常成人普通话语句测听复测信度评估

目的：评估汉语普通话言语测听材料（Mandarinspeechtestmaterials,MSTMs）中的等价语句测听材料在听力正常成人中的复测信度,以进一步完善该材料的标准化设计。方法随机抽取40例18～30岁听力言语正常的受试者,采用MSTMs中的等价语句测听材料（共15张测试表）,根据预实验结果测试受试者相对好耳在9dBHL时的言语识别率,间隔14～28天后在完全相同的条件下复测言语识别率,对两轮测试结果进行配对t检验,并计算临界差值。结果除表6外,其余14张MSTMs语句测听材料两轮言语识别率得分差异均无统计学意义;取95％置信区间的临界差值为18．93％,对应关键词数10个。结论本实验测得MSTMs语句测听材料复测信度好;使用该语句测听材料进行言语听觉能力评估时,若同一受试者前后两次识别率差异超过18．93％、对应关键词数超过10个,则认为干预有效。     

普通话言语测听单音节词表的编辑与初步等价性评估

目的根据目前听力学发展、尤其是人工耳蜗植入评估的的需要，编辑、录制普通话言语测听词表（mandarin speech test materials，MSTM）中的单音节词表，并进行初步的等价性分析。方法在建立专家小组的基础上，设定单音节词表的目的。依据普通话语音声、韵、调三维平衡和言语测听要求编辑了10张词表，每表50词。选词符合常用口语词和便于临床使用的原则，词源为普通话学界和国家公认的素材。然后由男声发音，录制合成磁带。对72例正常听力受试者进行单耳耳机测试，在其50％正确率水平评估各词表的等价性。结果通过试验证实有7张单音节词表具有良好的难度等价性，可以作为进一步验证和临床使用的基本材料。结论严格按照国际惯例和普通话规则编辑的普通话言语测听单音节词表，其中具有等价性者可以基本满足听力学临床、科研的需求。     

安静条件下语句识别表的等价性研究

目的 通过对解放军总医院研制的12张普通话儿童语句测听词表的难度等价性研究,判断该套材料是否符合言语测听材料的标准,以期挑选出具有难度一致性的词表应用于临床.方法 选取36名听力正常青年受试者,采用听说复述的方式,进行12组句表的测试(每组句表有50个关键词),以关键词计分,然后对各表得分进行统计学分析.结果 进行双因素(表号和强度)方差分析和Post-Hoc两两比较,12张测试表无统计学差异.结论 12张语句识别表具有较好的难度等价性,可应用于临床测试和实验研究.     

普通话言语测听材料的数字化录制与等价性分析

目的：对已经编制并通过初步等价性测试的普通话言语测听材料（MSTM）进行数字化录制，并进行第2次等价性测试和分析，为全面推广和进一步的多中心试验做准备。方法：根据第1次等价性测试的结果调整词表内容，增加练习用表。按照国际标准数字化录制所有材料，最后形成3张CD。每张CD均含有校准纯音、练习表、测试表。测试了60名青年正常听力受试者，单音节词、双音节词和句表的给声强度不尽相同。结果：统计分析后得到了具有等价性的两套单音节词表（每套9表）、9张双音节词表和27张句表。临床应用中如果不需要如此多的词表，可以选用最具有等价性的词表、句表用于同一（组）受试者。结论：具有等价性的普通话言语测听词表已经制备成CD，可以直接提供临床应用。     

正常人单音节词及语句识别-强度函数测试

目的:对难度等价性一致的普通话言语测听材料(MSTMs)中的单音节词表和句表进行分析,以期描绘出基于此测试材料的听力正常人识别-强度函数曲线.方法:选取18～25岁具有大专以上学历、且日常生活中以普通话作为交流方式的40例听力正常人作为受试者,对其中8例受试者进行预试验,从而获得正式测试时所需的给声强度;对其余32例受试者进行正式的言语识别测试.采用MSTMs中经过等价性评估的7张(每张50词)单音节词表和15张句表(每表50个关键词)作为测试材料.使用SPSS11.0软件对数据进行分析.结果:MSTMs单音节词表的言语识别率与给声强度间的函数关系的线性部分为Y=3.194X-46.147,识别率为50%时的给声强度为30.1 dBSPL;MSTMs句表的言语识别率与给声强度间的函数关系的线性部分为Y=7.419X-121.744,识别率为50%时的给声强度为23.1 dBSPL.结论:本研究初步建立了基于具有难度等价性的7张普通话单音节测听词汇表和15张普通话语句测听词汇表的正常人识别-强度函数曲线,为临床应用提供了基于MSTMs的正常听力者的鉴别标准.     

语音学理论在普通话单音节小词表编录中的应用

目的 探讨以汉语语音学理论为依据编制普通话言语测听材料 (mandarin speech test materials, MSTMs) 单音节小词表,从全新角度建立更适用于临床应用的评估材料.方法 以语音学理论为基础,将声母、韵母进行归类缩减后拼合.按照拼合后的音节选择日常生活中常用字,编制完成单音节小词表.对编制材料进行数字化录音以及音频处理,形成CD测试材料,并对词表进行初步等价性评估.结果 编制出体现语音学特征的汉语普通话言语测听小词表24张,每表20个单音节字.结论 体现语音学特征的汉语普通话言语测听小词表可以明显缩短测试时间,但在临床推广前尚需等价性评估.     

嘈杂语噪声下汉语语句测听中的学习效应

目的 验证采用噪声下语句识别表进行言语测听时是否存在"学习效应"(learning effect),并确定在使用本测试材料时需要几张练习表才能抵消学习效应.方法 选取年龄18～25岁、听力/言语发育正常且日常以汉语普通话为交流方式的64名受试者,在-3、-5、-7、-9 dB四种信噪比(signal-to-noise ratio, SNR)条件下进行语句识别率测试.采用32张嘈杂语噪声下汉语普通话语句识别表(每表10句,每句3～7个     

普通话言语测听双音节词表的编辑与初步等价性评估

目的根据言语测听以及普通话语音学的要求,编制汉语（普通话）言语测听材料（mandarin speech test materials,MSTM）中的双音节词（扬扬格）表,作为临床常规测试言语识别阈以及言语评估的基本材料.方法组织各相关学科专家共同制定选词标准,对于具体细节以中文、普通话的规则为准.共编辑了10张词表,每表50词,基本做到声、韵、调三维平衡.录制合成为磁带后进行了正常听力受试者的等价性测试,测试采用单耳耳机给声.并采用3种统计方法分析.结果相关分析结果显示各表间除表5以外相关系数很高,聚类分析将10张表分为2类,但一致性检验Kappa时显示各配对组中Kappa值≥0.40组很少,一致性不高.结论双音节词表作为临床最常用的言语测听材料之一,其编制、录制、等价性评估受到各种影响因素的作用而需要多学科的合作.本次编制的材料尚需要进一步调整录制、测试方法,以期在符合音素平衡的前提下,适合临床、研究使用.     

普通话单音节词言语测听材料的等价性分析

目的评估按照语音平衡和音素平衡的原则编制出的汉语(普通话)测听词汇表(MandarinSpeechTestMaterials,MSTM)中单音节词言语测听词表,以适用于临床诊断和康复效果的评估。方法分析60例正常听力者接受词表测试的结果,比较表间的难度,寻找难度一致的单音节词表作为测试用表。结果10张单音节词测试表可以分成两部分,测试表一、二、三、五、六、七、八、九、十难度具有等价性;测试表一、三、四、五、六、七、八、九、十难度具有等价性。结论分成的这两部分所包含的九张测试表之间具有等价性,可以初步应用于临床,但在推广前尚需更大样本测试。     

Recognition of lexically controlled words and sentences by children with normal hearing and children with cochlear implants

OBJECTIVE: The purpose of this study was to investigate the combined lexical effects of word frequency and neighborhood density (acoustic-phonetic similarity) on the recognition of words spoken in isolation and words spoken in sentences for children with normal hearing and children with cochlear implants. DESIGN: Lexically controlled sentences were created from a subset of words obtained from the spoken vocabulary of children between the ages of 3 and 5 yr. Two sentence lists were generated, applying the definitions and procedures of Kirk, Pisoni, and Osberger (1995) in accordance with the Neighborhood Activation Model ( Luce, 1986; Luce & Pisoni, 1998). One list was composed of lexically "easy" words (those high in frequency of occurrence but phonemically dissimilar to other words) and the other list was composed of lexically "hard" words (those low in frequency of occurrence but phonemically similar to other words). Each list consisted of five practice and 20 test sentences that were syntactically correct but semantically neutral (low in predictability). Three key words were used in constructing each of the 5- to 7-word sentences, resulting in 15 practice and 60 key words per test list. In the first of three experiments, 48 normal-hearing children between the ages of 5 and 12 yr were asked to repeat the words and sentences at one of six presentation levels to establish performance-intensity functions. In the second experiment, 12 normal-hearing children between the ages of 5 and 14 yr repeated the words and sentences under spectrally degraded conditions. Twelve children with cochlear implants ages 5 to 14 yr repeated the unprocessed stimuli in a third experiment. RESULTS: The lexically easy stimuli were recognized with greater accuracy than the lexically hard stimuli for the children tested in all three experiments. Sentence scores were significantly higher than word scores for the normal-hearing children (Experiments 1 and 2) and nine high-performing children with cochlear implants (Experiment 3). Word scores were higher than sentence scores for the three low-performing children with cochlear implants. There was a statistically significant relationship between chronological age and sentence score for the normal-hearing children listening under spectrally degraded conditions. For the children with cochlear implants, the relationship between language quotient and sentence and word scores was statistically significant. CONCLUSION: Sensitivity to the combined lexical properties of word frequency and neighborhood density was evident both for words and sentences. Lexically easy stimuli were recognized with greater accuracy than lexically hard stimuli across groups, affirming the robustness of this effect and verifying that words were being organized in relation to the frequency and acoustic-phonetic properties of other words. Syntactic context facilitated word recognition for the children with normal hearing and the high-performing implant group. The three low-performing children with cochlear implants recognized words more accurately than sentences, reflecting limitations in linguistic and cognitive capacity.     

Development of the Cantonese Hearing In Noise Test (CHINT)

OBJECTIVE: To develop a Cantonese version of the Hearing In Noise Test (CHINT) with the same features as the English Hearing In Noise Test (HINT) (Nilsson, Soli, & Sullivan, 1994). DESIGN: The CHINT was developed in five separate studies: (1) evaluation of initial materials; (2) creation of sentence materials; (3) equalization of sentence difficulty; (4) creation of sentence lists; and (5) evaluation of response variability, inter-list reliability, and establishment of norms. Using the CHINT material, reception thresholds for sentences were measured under four headphone test conditions: Quiet, and in noise with noise simulated as originating from 0 degrees (noise front), 90 degrees (noise right), and 270 degrees (noise left). The speech source was located at 0 degrees in all conditions. The locations of the speech and noise sources were simulated using virtual audio processing, as with the English HINT. The noise conditions consisted of listening with noise fixed at 65 dBA with the level of speech varied in an adaptive procedure. A total of 142 subjects with normal hearing thresholds participated in the five studies. RESULTS: Two versions of the test materials, twenty-four 10-sentence lists and twelve 20-sentence lists, were created from a single set of 240 sentences containing 10 syllables per sentence. Using the twenty-four 10-sentence lists, mean thresholds under earphones in quiet were measured at 19.4 dBA and reception thresholds for sentences of -3.9 dB for noise front, -10.6 dB for noise right, and -10.5 dB for noise left. Similar results were obtained using the 20-sentence lists (19.4, -4.0, -10.9, and -11.0 dB, respectively, for quiet, noise front, noise right, and noise left conditions). There was low response variability within each list. High inter-list reliability suggests that consistent results could be obtained using any list. Confidence intervals are reported. The CHINT norms for listening in quiet and noise conditions were comparable to those for the English HINT. CONCLUSIONS: The CHINT is the first standardized Cantonese sentence speech intelligibility test. The CHINT was developed using the same rationale as the English HINT, allowing norm reference results from the two tests to be compared directly across languages. Results showed the CHINT is a reliable test. The CHINT would benefit from further evaluation of validity.     

Development of equally intelligible Telugu sentence-lists to test speech recognition in noise

Objective: To develop sentence lists in the Telugu language for the assessment of speech recognition threshold (SRT) in the presence of background noise through identification of the mean signal-to-noise ratio required to attain a 50% sentence recognition score (SRTn). Design: This study was conducted in three phases. The first phase involved the selection and recording of Telugu sentences. In the second phase, 20 lists, each consisting of 10 sentences with equal intelligibility, were formulated using a numerical optimisation procedure. In the third phase, the SRTn of the developed lists was estimated using adaptive procedures on individuals with normal hearing. Study sample: A total of 68 native Telugu speakers with normal hearing participated in the study. Of these, 18 (including the speakers) performed on various subjective measures in first phase, 20 performed on sentence/word recognition in noise for second phase and 30 participated in the list equivalency procedures in third phase. Results: In all, 15 lists of comparable difficulty were formulated as test material. The mean SRTn across these lists corresponded to ?2.74 (SD?=?0.21). Conclusions: The developed sentence lists provided a valid and reliable tool to measure SRTn in Telugu native speakers.     

The development of Cantonese Lexical Neighborhood Test: a pilot study

OBJECTIVE: This study aimed at developing a theoretically driven open-set speech recognition test for pediatric clinical population of cochlear implant and/or hearing aid users, with Cantonese Chinese as their first language, to track progress in speech recognition performance as an outcome measurement of their rehabilitation. METHODS: Six monosyllabic and six disyllabic word lists were generated from the Cantonese CHILDES language database, constructed according to the Neighborhood Activation Model. There were three lexically "easy" and three lexically "hard" word lists in each sub-test, with 25 items in each list. Four pediatric cochlear implant users and 10 hearing aid users, with bilateral congenital severe to profound sensorineural hearing impairment and below the age of 10, participated in the study. Their performances on word recognition and phoneme recognition with the new test lists, as well as the inter-list equivalency, inter-rater reliability, and face validity of the new materials, were investigated. RESULTS: Word recognition was higher among disyllables than monosyllables. Lexically "easy" disyllabic words were better recognized than their "hard" counterparts and the monosyllables. No significant difference was noted among the three lists in each combination of syllable structure and lexical property. High inter-rater reliability, as well as high correlation between Cantonese LNT score and a receptive vocabulary test score, were revealed. CONCLUSIONS: These newly developed test lists provided reliable information on spoken word recognition of pediatric hearing prosthesis users with severe to profound hearing impairment. Inter-list equivalency and inter-rater reliability allowed monitoring of rehabilitation progress on such specific pediatric clinical population with this new test. (255).     

Design,optimization and evaluation of a Danish sentence test in noise

The Danish sentence test DANTALE II was developed in analogy to the Swedish sentence test by Hagerman and the German Oldenburg sentence test as a new Danish sentence test in noise to determine the speech reception threshold in noise (SRT, i.e. the signal-to-noise ratio (SNR) that yields 500% intelligibility). Each sentence is generated by a random combination of the alternatives of a base list. This base list consists of 10 sentences with the same syntactical structure (name, verb, numeral, adjective, object). The test sentences were recorded and segmented in such a way that the coarticulation effects were taken into account in order to achieve a high perceived sound quality of the resynthesized sentences: 100 sentences were recorded, each coarticulation between each word and the 10 possible following word alternatives were recorded, and the correct coarticulation was used to generate the test sentences. Word-specific speech recognition curves were measured for each recorded word to optimize the homogeneity of the speech material and the measurement accuracy. Level corrections of particular words and a careful selection of the test lists produced a noticeable reduction in the variation in the distribution of word-specific SRT (standard deviation 1.75 dB instead of 3.78 dB). Therefore, the slope of the total intelligibility function was expected to increase from 8.30%/dB (raw test material) to 13.2%/dB (after modification). These theoretical expectations were evaluated by independent measurements with normal-hearing subjects, and, for the most part, confirmed. The reference data for the DANTALE II are: SRT=-8.43 dB SNR; slope at SRT, s50 = 13.2%/dB. The training effect was 2.2 dB and could be reduced to less than 1 dB if two training lists of 20 sentences were performed prior to data collection.