目的 应用项目反应理论(item response theory,IRT)对医学心理学课程考试试卷质量进行分析,为教育教学质量评价提供参考。方法 2022年5月,对山西医科大学2016级临床医学专业330名学生医学心理学试卷中的60道选择题进行质量评价,检验了IRT模型的核心假设和模型-数据拟合情况,分析了试题参数和信息量。结果 试题数据符合IRT理论假设,模型拟合良好,选择了双参数逻辑回归模型进行参数估计,有59道(98.3%)试题符合区分度参数估计标准,有54道(90.0%)试题符合难度参数估计标准,有55道(91.7%)试题能够提供最低要求信息量(0.16),最终质量良好的试题有49道(81.7%),试卷能够提供的信息量达到25。结论 试卷整体质量良好,绝大部分试题质量较高,小部分试题需要修改或删除。本研究可以为高校部分课程考试试卷依据IRT进行定量评价提供借鉴。

Objective Applying item response theory (IRT) to analyze the quality of medical psychology course examination papers provides a reference for the quality of the evaluation of education and teaching. Methods In May 2022, we assessed the quality of 60 multiple-choice questions in the Medical Psychology test paper for 330 clinical medicine students in the class of 2016 at Shanxi Medical University. We tested the core assumptions of the IRT model and analyzed the model-data fit, as well as the item parameters of the test questions and the amount of information. Results The test data were consistent with the theoretical assumptions of IRT, and the model was well-fitted. The two-parameter logistic regression model was selected for parameter estimation. 59 (98.3%) of the test questions met the criteria for estimating the discrimination parameter, 54 (90.0%) of the test questions met the criteria for estimation of the difficulty parameter, 55 (91.7%) of the test questions were able to provide the minimum required amount of information(0.16), and 49 (81.7%) of the test questions were of satisfactory final quality, with the amount of information provided by the examination paper reaching 25. Conclusions The overall quality of the question papers was good, with the vast majority of the items being of high quality. However, it was necessary to modify or delete a small number of questions. This study can serve as a reference for the quantitative evaluation of examination papers in some university courses based on IRT.