基于概化理论的新开课程试卷可靠度优化研究  

On How to Optimize the Dependability of a New Course Test Based on Generalizability Theory

作　　者：王天剑 Wang Tianjian(Foreign Language School,Guizhou University of Finance and Economics,Guiyang,Guizhou 550025,China)

机构地区：[1]贵州财经大学外国语学院

出　　处：《保定学院学报》2019年第3期108-112,共5页Journal of Baoding University

基　　金：2016年贵州省教育科学规划课题“概化理论在大学英语测试可靠度评价中的应用研究”(2016B236)

摘　　要：课程测试方式不仅受课程性质影响,也受学生的认知水平等因素影响。为考查某高校一门新开课程测试的可靠度,教研人员进行了一次模拟测试和概化理论研究。被试由10名修课学生构成,试卷由题库中随机抽取的不同题型构成。利用mGENOVA对成绩进行了多元概化分析。结果显示,模拟测试可靠度远远偏离理想水平。通过多次研究发现,在合乎逻辑且可操作范围内,合理调整题量与赋分权重,可使测试可靠度提升到理想水平。The method of course testing is not only influenced by the nature of the course, but also by factors such as the student’s cognitive levels. In order to examine the dependability of the test method of Introduction to Translation, a new course in a university, the teaching and research personnel conducted a simulation test and a generalizability theory study. The subjects consisted of 10 students, and the test papers were composed of different questions randomly selected from the question bank. The results were submitted to a multivariate generalizability theory analysis using mGENOVA. The analysis shows that the simulation test’s dependability is far from the ideal level. Through repeated decision studies, it is found that the test dependability can be improved to an ideal level by adjusting the number of questions and score weights.

关 键 词：概化理论 课程测试方式 试卷可靠度 

分 类 号：G40-058.1[文化科学—教育学原理]