<基本认知能力要素测试系统>的编制与信度、效度分析

目的：建立一种新型的认知能力评价系统，用其进行重复测查时能有效地避免（消除）学习效应．方法：采用普遍认可的量表，根据随机化原则编制计算机化测试系统，该系统与研究任务和对象相适应，测试程序规范，测试项目较多并可任意选择和组合，测试结果记录自动化．为用户提供一种方便友善的人机界面．结果：５４名受试者相隔２ｗｋ的重测可靠性第数为０．６２￣０．８５，符合设计要求；分半信度系数为０．６５￣０．８４，表明测试

Establishment of and analysis of its reliability and validity

AIM: To establish a new type of assessment that evaluates cognitive ability and behavioral attribute, and avoid learning bias effectively, too. METHODS: Based on randomization, a computer administered basic cognitive ability testing system with accredited scales was established. This system is canonical, all purpose, auto recorded, and its multi items can be selected and combined freely. RESULTS: 54 subjects were measured in 1st and 3rd week, the test retest reliability coefficients were 0.62~0.85, which achieved the levels of test reliability; the split half reliability coefficients of all results were 0.65~0.84, proved interior coherence is up to grade; using NES C performance as the criterion, coefficient of criterion related validity is 0.77~0.85, the two are well correlated. CONCLUSION: The reliability and validity of this cognitive ability assessment instrument is high. It gives a new method to assess affections of abnormal environments on cognitive ability.