项目反应理论在课程教学质量评价中的应用  

APPLICATION OF ITEM RESPONSE THEORY IN THE ASSESSMENT OF COURSE TEACHING QUALITY

作　　者：崔文倩 张睿[1] 张志华[1] 吴天刚[1] CUI Wenqian;ZHANG Rui;ZHANG Zhihua;WU Tiangang(School of Physics Science and Engineering,Tongji University,Shanghai 200092)

机构地区：[1]同济大学物理科学与工程学院,上海200092

出　　处：《物理与工程》2020年第6期61-67,共7页Physics and Engineering

基　　金：上海市教育科学研究项目(C20122);大学物理互动式电子书设计与教学实践以及同济大学公共基础课教改项目;大学物理公共基础课程改革的支持。

摘　　要：测试是测量学生学习效果的重要手段,基于测试结果的教育数据挖掘有助于了解课程教学质量。本研究以大学物理课程为例,根据项目反应理论中的三参数逻辑斯蒂模型,计算出不同教学班级选择、填空与计算三种题型的难度系数、区分度和猜测系数,分析结果发现填空题猜测系数能较好地反映班级教学质量。研究表明无论是采用箱图法还是聚类分析法,通过填空题猜测系数这一指标都能发现教学效果出现异常的班级。通过精准化的教学质量管理,教学的整体质量将逐步提高。As an important means to measure the learning effect of students,educational data mining based on test results will help to understand the quality of course teaching.This study takes the university physics course as an example.Based on the three-parameter logical model in item response theory,we calculate the difficulty coefficients,discrimination and guessing coefficients of the three types of questions:Multiple choices,fill in the blanks and calculation in different teaching classes.The analysis results show that the guess coefficient of fill-inblanks can reflect the quality of class teaching.Our research shows that with either the box plot method or the cluster analysis method,the guess coefficient of filling-in-blanks can be used to detect the outlier classes.Through accurate teaching quality management,the overall quality of teaching will be improved.

关 键 词：项目反应理论 教学质量评价 教育数据挖掘 

分 类 号：G642[文化科学—高等教育学] O4-4[文化科学—教育学]