多群体多元概化理论在高考数学试卷质量分析中的应用  

The Application of Multi-Group Multivariate Generalizability Theory to the Quality Analysis of College Entrance Examination Papers in Mathematics

作　　者：范晓玲[1] 刘洪涛[2,3] 伍慧[2] 

机构地区：[1]湖南师范大学教育科学学院,认知与人类行为湖南省重点实验室,长沙410081 [2]湖南师范大学教育科学学院,长沙410081 [3]邵阳学院

出　　处：《教育测量与评价》2017年第3期48-56,共9页Educational Measurement and Evaluation

基　　金：湖南省教育厅科学研究项目(13A057);湖南省哲学社会科学基金项目(14YBA284)阶段性成果

摘　　要：采用多群体多元概化理论分析高考数学试卷的质量,可以为改进和提高高考命题质量提供实证依据。文章以某年某省高考数学试卷中的部分主观题和客观题为分析对象,对全体考生进行随机等距抽样,结果发现:(1)项目区分度和难度在不同数学素养人群中的效应不同。(2)信度方面,无论题量如何变化,就全体测量对象而言主观题的可靠性是理想的,且低数学素养组的可靠性高于高数学素养组,变化幅度较小;客观题在全体测量对象中的可靠性也较为理想,低数学素养组的可靠性同样高于高数学素养组,但变化幅度较大。(3)效度方面,就全体测量对象而言中区分与高区分试题的效度较好,低数学素养组的效度很差,高数学素养组的效度较低;客观度区分度的拟合情况较好。因此,提高客观题效度的有效方式是改善考生方差,提高主观题效度的有效方法是降低误差方差。It is possible to provide an empirical basis for improving the quality of college entrance examination by analyzing the quality of college entrance examination papers in Mathematics through using multi-group multivariate generalizability theory. The article takes some subjective and objective items of college entrance examination paper in Mathematics in a certain province in some year as the analysis of object, having all the candidates for systematic sampling. The results show that ： （1 $Item discrimination and item difficulty in groups of different Mathematics quality are different.（2）As to the reliability. The reliability of subjective items is ideal in all measuring objects regardless of the number of examination items. Meanwhile, the reliability of lower Mathematics literacy group is better than that of the higher Mathematics literacy group,which represents a small rangeability. The reliability of objective items is good in all measuring objects. At the same time, the reliability of lower Mathematics literacy group is also better than that of the higher Mathematics literacy group,which represents a big rangeability. （3）As to the validity. The degrees of higher and medium discrimination have better validity in all measuring objects. The validity of lower Mathematics literacy group is worse, and that of higher Mathematics literacy group is relatively lower. The degree of fitting between objective degree and discrimination is better. So,changing candidates’ variance is a good method for improving the validity of objective items, and decreasing the error variance is another valid way for improving the validity of subjective items.

关 键 词：高考数学 质量分析 多元概化理论 

分 类 号：G632.474[文化科学—教育学]