汉语阅读技能诊断测验(CRSDT)的初步编制

Development of Chinese Reading Skill Diagnostic Test: Reliability and Validity

目的：为汉语儿童阅读障碍（ＲＤ）的研究与临床标准化诊断试编制一个《汉语阅读技能诊断测验（ＣＲＳＤＴ）》。方法：根据认知心理语言学与神经语言学理论，结合ＲＤ症状学特点编制了ＣＲＳＤＴ，在小学三～六年级２４５名阅读障碍与正常儿童中作一试验。结果：（１）在测验各项目及总分上，正常儿童与ＲＤ儿童差别显著（Ｐ＜．０００１）；正常儿童不同学绩水平间的差别大于不同年级之间的差别。（２）分半信度为．７７～．８１；重测信度为．９０（．６７～．９４）；α系数为．９０。（３）测验总分与语文成绩相关为．８５；临床按标准分７０划界诊断的灵敏度为９４．７％，特异度为９８．５％；因素分析所获三因子模型（字识别、词句理解、记忆－把握）可解释６４．４％的变异，符合汉语失读症的神经语言学层级理论。结论：ＣＲＳＤＴ的信度、效度初步达到心理测量学要求，具有临床实用价值，并与国外同类测验作了比较。

To develope the Chinese Reading Skill Diagnostic Test (CRSDT) for use in clinicaldiagnosis and research on Chinese children with Reading Disorder(RD). Method:Based on cognitive neurolinguistic theories and clinical features of RD among Chinese children, CRSDT was developed and tested on 245 children with and without RD at grades 3～6. Results:(1)Significant differences in total and subtest scores were found between RD and normal children. Among normal children, larger differences were obtained between superior and lower achievement levels as compared with differences between school grades. (2) The CRSDT showed satisfactory reliability, as revealed by split-half reliability (r=.77 to .81), test-retest reliability (r=.67 to .94), and alpha coefficient (α=.90). (3) Validity of CRSDT was also good, as shown by correlation between scores on CRSDT and Chinese achievement (r=.85), test sensitivity (94.74%) and test specificity (98.47%). (4) Factor analysis resulted in a three-factor model accounting for 64.4% of total variances. The three factors identified were character recognition, word and sentence comprehension, and memory-accuracy, which accord well with the neurolinguistic theory for acquired dyslexia in Chinese adults.Conclusion: The CRSDT attained satistifactory psychometric properties and pragmatic values. Comparison with similar tests developed abroad was discussed.