发展性阅读障碍是儿童在小学阶段最普遍的学习困难之一，许多发展性阅读障碍儿童将会在一生经历持续的读写困难，这将使他们面临更高的失业、自卑、焦虑和抑郁的风险。这种阅读能力发展落后个体的特点是智力正常、无神经系统障碍疾病、个体的生活教育环境因素良好，但其阅读能力的发展落后于同龄儿童。相比同龄的阅读水平正常的儿童，发展性阅读障碍儿童在阅读理解能力、识字能力、朗读能力，以及涉及需要阅读参与的课业上，都表现出显著落后。早诊断、早干预将有助于他们更好的适应学习以及调整心理状态， 因此，编制合适的工具筛查出这些阅读落后的儿童是非常重要的。 目的：寻找最合适的形音捆绑测验的材料类型，初步编制有效的小学生形音捆绑能力测验。 方法：参照国外成功的形音捆绑能力测验编制模式，结合汉语形音特点，确定适合的能够预测形音捆绑能力的视觉材料（片假名、平假名/图形）和语音材料（真音/假音）， 采用 2×2 双因素设计，编制 4 套形音捆绑测验，每套测验均有 25 道题目；随后通过专家 访谈等方式进行材料评定，调整 4 套试题的难度，编制形音捆绑测验的计算机测验软件。 正式测验的样本来自北京两所小学的 1169 名学生，使用识字量测验和汉字识别测验作为分组标准，并根据瑞文智力测验结果排除智力落后儿童及 SNAP-Ⅳ量表排除 ADHD 儿童， 总共有 85 名阅读落后组学生，和 120 名正常能力组学生。运用 SPSS_25.0 进行项目分析、信效度检验，并针对测验材料的鉴别能力进行分析确定一种最具有预测性的测验材料。针对最有效的测验进行 ROC 曲线分析，得到筛选阅读落后儿童的区分分数，使用区分分数进行判别分析，检验测验的鉴别能力。 结果：四种形式的形音捆绑能力测验的分数整体随年龄呈现递增趋势。项目分析表明四套测验的整体内部一致性系数为 0.815，平均难度在 0.3~0.69 之间，鉴别力指数在0.20~0.29 之间，均达到了尚可水平。按照年级进行测验成绩的 t 检验，发现图形×假音的材料组合的测验二最具有预测性。针对测验二在不同年级进行 ROC 曲线分析，确定了一至五年级判断为阅读能力落后的区分分数为：12.5、9.5、13.5、13.5、13.5。使用划分方式进行落后判定后，进行判别分析，测验二可以正确分类 71.2%的学生，因此该区分分数对阅读障碍的筛选有一定准确性。使用区分度、克隆巴赫α信度系数、ROC 曲线面积进行题目剔除后，删除测验二的第 1、4、8、11、13、22、25 题。对剔除以后的剩余的 18 个题目进行信度检验：克隆巴赫α系数为 0.66，内部一致性信度增大。使用 t 检验的方法对剔除题 1 目后两组被试进行区分，发现测验二的效应值从 0.80 增加到 0.89。 结论：形音捆绑能力测验的项目难度、区分度均较好，信度和效度达到了测量学的基本要求。该测验可以作为阅读落后儿童的初步筛选工具。

Dyslexia is one of the most common learning difficulties in childhood. Many children with dyslexia will experience continuous dyslexia in their lifetime, which will make them face higher risks of unemployment, inferiority, anxiety and depression. Developmental dyslexia is a type of dyslexia, and it is not caused by individual intelligence, main living and educational environment and various nervous system disorders. Compared with children of the same age with normal reading level, children with developmental dyslexia show significant backwardness in reading comprehension ability, literacy ability, reading ability, as well as in classes involving reading participation. Early diagnosis and intervention will help them better adapt to learning and adjust their mental state. Therefore, it is very important to develop appropriate tools to screen children with dyslexia. Objective: Preliminary preparation of the primary school student's ability to form and tone binding test, analysis of the test items, explore and screen the best auditory and visual stimulus materials for the development of the ability to form and tone binding of primary school students, determine the best test, judge the screening accuracy of the test. Methods: Referring to the successful compilation model of the Form-sound binding ability test in foreign countries and combining with the characteristics of Chinese Form-sound binding ability, the suitable visual materials (katakana, hiragana/figure) and voice materials (true/false sound) which can predict the ability of Form-sound binding were determined. Four sets of Form-sound binding tests were designed with 2×2 two-factor design. Each set of tests identified 25 items with 0/1 score. Subsequently, expert interviews were conducted. By means of material evaluation, adjustment of the difficulty of four sets of questions, compilation of computer testing software for shape and sound binding test. The sample of the formal test was 1170 students from two primary schools in Beijing. The literacy test and Chinese character recognition test were used as the grouping 3 criteria. According to the results of Raven intelligence test, the children with mental retardation and the children with ADHD were excluded from the Chinese version of the Swanson, Nolan and Pehlam version IV. Finally, 100 students in the reading retardation group and 140 students in the normal ability group were determined. Name. Four sets of sound binding ability tests were conducted on 240 students, and 840 valid test results were recovered. A total of 85 students in the reading backwardness group and 120 students in the normal ability group were enrolled. SPSS_25.0 was used for item analysis, reliability and validity test, and the identification ability of test materials was analyzed to determine the most predictive test materials. For the most effective test, ROC curve analysis was carried out to obtain the discriminant scores of children with reading disabilities. The discriminant analysis was carried out by using the discriminant scores to test the discriminant ability of the test. Results:The scores of the four forms of shape and sound binding ability test showed an increasing trend with age as a whole. Item analysis shows that the overall internal consistency coefficient of the four sets of tests is 0.815, the average difficulty is between 0.3 and 0.69, and the discriminant index is between 0.20 and 0.29, all of which reach the acceptable level. According to the t-test of the test results of grade, it was found that the second test of the combination of figures and falsetto was the most predictive. According to the ROC curve analysis of the second test in different grades, the distinction scores of the first to fifth grade judging reading ability as backward are 12.5, 9.5, 13.5, 13.5 and 13.5. After using the method of partition to determine the backwardness, the discriminant analysis is carried out. Test 2 can correctly classify 71.2% of the students, so the score has certain accuracy in the screening of dyslexia. Questions 1, 4, 8, 11, 13, 22 and 25 of the second test were deleted after using discrimination, clonal Bach alpha reliability coefficient and ROC curve area. The reliability of the remaining 18 items was tested: the clonal Bach alpha coefficient was 0.66, and the reliability of internal consistency was increased. T-test was used to distinguish the two groups of subjects after eliminating the questions. It was found that the effect value of test 2 increased from 0.80 to 0.89. Conclusion: The difficulty and discrimination of the test items are good, and the reliability and validity meet the basic requirements of surveying. The test can be used as a preliminary screening tool for children with dyslexia.