主页 | 中文 | English

第一届

特邀发言

时间 2014-06-29 23:27:08

《儿童语言(普通话)能力检测量表》研发报告

宁春岩 (Ning, Chunyan)

天津师范大学 语言研究所  

 

    《儿童语言(普通话)能力检测量表》(以下简称量表”)是国家社科基金重点项目,(项目批准号为08AYY001,由宁春岩主持在湖南大学和天津师范大学实施,于20116月结项完成。

    该量表的研发对北方方言区(石家庄、新乡、西安、河北静海县、天津)2,534名学前3-6岁儿童语言能力进行的检测,并使用了Winsteps SPSS对得到的数据、检测题、检测方法进行了多方面的统计分析,经过修改、实验、再修改、再实验后,最后研制而成。

    该量表由三类七个部分组成:第一类是基于常模的(Norm-Referenced)语言能力检测量表,其中包括语音部分、句法部分、语义部分和语用部分;第二类是基于单纯性语言障碍(Specific Language Impairment, SLI)症状的(Criterion-Referenced)语言障碍检查量表,其中包括SLI检查量表-句法部分和SLI检查量表-音系部分;第三类属语言障碍筛查(Language Screening)。

   其中,语言筛查依据SNEL模型而建立,经过多次实验和Winsteps在分离信度、整体分离指数等方面的评估检测,最后确认得到一个含有21个儿童语言发展里程碑的问卷表。

该量表其他部分的特点可概括如下:

(1)               结构特点:基于常模的检测、基于SLI症状的检测和家长报告筛查三者结合进行;

(2)               样本特点:研制所依据的样本规模比较巨大,多至2,543名受试儿童。

(3)               检测方式上的特点:计分全部使用了双值(正确/错误)积分法,使量表使用中的主观任意性降低到最低的程度;

(4)               统计上的特点:在检测题设计和结果统计分析中,使用了项目反应论模型进行测试评估和检测题调整,从而集中界定出语言能力正常与语言能力异常的临界值。

(5)               使用上的特点:

a)      具有临床使用上便捷性(检测实施平均约30分钟),检测实施者用前培训简单。

b)      具有使用上的广泛性:除可供语言治疗师专业临床使用外,量表可改编成家长自测版” (已完成),为家长监护子女语言发展用;也可以编写成网络版(已完成),让更多人参与研究,让更广泛的人使用;也可作为语言习得数据库,为儿童语言习得和语言障碍研究者使用。

该量表业已得出了普通话学前儿童语言能力常模和标准方差,从语言学角度诊断了5-7%的有疑似SLI症状的儿童。该量表初步被证实是可行的,值得试用的。期望能得到我国儿童医疗界、幼教工作者,幼儿语言教师、特殊教育学校、儿童家长、语言习得研究者和语言障碍研究者的共享试用,以求的更加完善。

 

关键词:语言障碍、SLI、语言习得、语言障碍筛查、儿童语言发展

 

 

Tone Perception and Music Training

 Peggy Mok

Chinese University of Hong Kong

 

This study investigates whether musical training can facilitate lexical tone perception of native speakers of a tone language. Cantonese has a complex tone system with six lexical tones (T). Some tone pairs, T2/T5 (rising), T3/T6 (level), T4/T6 (falling vs level) are merging in Hong Kong in recent years. A previous study shows that merging native Cantonese speakers have poorer general tone perception of Cantonese tones than their non-merging native counterparts, not only for the merging tone pairs. This suggests that the merging speakers are less sensitive to linguistic pitch differences in general, which concur with the idea that sound change may be caused by poor perception of the listeners. Other studies have established that musical training facilitates lexical tone perception of non-tone language speakers. However, it is still unclear if musical training can also influence tone perception of native speakers of tone-languages who are merging tones. Given the functional role of pitch in both linguistic and musical tones, it is legitimate to ask whether musical training can also facilitate the perception of these merging speakers with poorer lexical tone perception than control speakers, and whether the same perceptual mechanism is used in both linguistic and musical pitch perception. Three groups of listeners (30 normal Cantonese, 28 merging Cantonese, 34 foreign) with and without advanced musical training (defined as having seven or more years of formal training in any instrument or vocal singing) were invited to participate in tone perception experiments. AX discrimination tasks of Cantonese lexical tones and pure tones were conducted. The pure tones were resynthesized from canonical Cantonese tones, i.e., they have the same pitch contours as Cantonese tones but with no segmental information. Both accuracy and reaction time data were collected using E-prime. Both types of data show that while musical training can enhance lexical tone perception of foreign listeners, it has little influence on merging Cantonese listeners. The results indicate that different perceptual mechanisms may be involved in linguistic and musical tone perception, and that the linguistic use of tones is more fundamental and more robust than musical training. If the linguistic domain is not activated, then musical training can enhance lexical tone perception, but not vice versa. The present study provides empirical data to address important psycholinguistic issues related to speech perception and processing.