摘要
目的 本研究对北京协和医学院选修医学统计学课程的2014级研究生的122份期末考试试卷进行评价分析,为评定教学效果和教学质量提供依据.方法 运用教育测量学理论和教育统计学方法对试卷进行评价,采用难度、区分度、信度（Cronbach α系数）和覆盖度对122份医学统计学期末考试试卷进行质量分析.结果 该试卷的笔试平均成绩为（52.5±8.4）分（满分70分）,软件操作考试平均成绩为（20.4±4.1）分（满分30分）.总分平均为（72.9±10.7）分,最低分42分,最高分92分,近似服从正态分布.平均难度为0.73,区分度为0.31,信度为0.71.22个考题的相关系数均未超过0.4,其中19个题目的相关系数小于0.2.结论 该考卷平均难度适中,具有较好的区分度、信度和总体覆盖度,适于进行医学研究生期末测评.此外,医学研究生对统计学软件应用的动手能力较差,应该进一步加强统计学软件应用的教学.
Objective To learn the overall results of the final test paper on medical statistics of medical graduates of Peking Union Medical College,and to provide the theoretical support for teaching quality assessment.Methods The difficulty coefficient,the discrimination index,the reliability （Cronbach α coefficient） and the degree of coverage were analyzed based on education measurement methods.Results The mean scores of the written examination section and the software operation examination were 52.5±8.4 and 20.4 ±4.1 respectively.The mean total score was 72.9 ± 10.7,ranged from 42 to 92,which was normally distributed.The difficulty coefficient,the discrimination index and the Cronbach α coefficient were 0.73,0.31 and 0.71 respectively.The correlation coefficients of 19 examination items were smaller than 0.2 and other coefficients were smaller than 0.4.Conclusions The test paper held the moderate difficulty and the good discrimination,reliability and the degree of coverage,which was fit for the final test of medical graduates.The teaching of statistical software should be strengthened in future.
出处
《中华医学教育杂志》
2015年第5期795-797,共3页
Chinese Journal of Medical Education
关键词
医学院
研究生
医学统计学
试卷分析
Medical college
Graduate
Medical statistics
Test paper analyses
