论文部分内容阅读
本文应用概化理论,以义务教育阶段学生识字量测验为研究对象,探讨题组侧面和题目侧面对识字量测验信度的影响。研究分别采用随机单面交叉设计和混合设计进行分析。研究结果发现该识字测验的概化系数高于0.90,可靠性指数高于0.85,能够准确测量学生识字能力。测试用字数量的变化会导致概化指标的变化,提高识字量测验信度的有效方法是增加题组数。该测验的简版测试用字数分别为20和30时,即可保证可靠性指数高于0.8,达到测量学要求。
In this paper, the application of generalization theory, with compulsory education students in the amount of literacy test as the research object, to explore the side groups and the subject side of the impact of the test of confidence level. The study was conducted using random single-sided cross design and mixed design analysis. The study found that the generalized coefficient of the literacy test was higher than 0.90 and the reliability index was higher than 0.85, which could accurately measure students’ literacy skills. Changes in the number of test words will lead to changes in general indicators, to improve the reliability of the test method is to increase the number of questions group. The simple test of the test with the words were 20 and 30, respectively, to ensure that the reliability index higher than 0.8, to meet the measurement requirements.