变异系数显著性检验在评价学生综合学习能力中的应用

Application of hypothesis test for coefficient of variation in the evaluation of the comprehensive ability and knowledge of medical students

目的 通过对考试题型得分变异程度的分析,为评价学生有关课程学习方式提供数据处理的方法.方法 对桂林医学院2008级临床医学专业92名本科学生医学统计学课程考题中单项选择题、填空题、名词解释题及分析题得分的变异程度进行比较分析.结果 确定变异系数显著性检验水准为α1=0.0167,综合分析题（32.92％）变异系数较高,而填空题（17.14％）、名词解释题（18.91％）和单项选择题（19.23％）变异系数均较低,前者分别与后三者比较,其差异均有统计学意义（P＜0.0167）.结论 综合分析题得分的个体差异较大,提示有相当数量学生灵活运用知识综合分析问题的能力不足.

Objective This study tends to establish a data processing method for the learning type of medical statistics course through the test item type analysis. Methods An investigation was carried out on 92 students graded 2008 from the department of clinical medicine. It compares and analyzes the variation of exam scores. Results The significance level of multi-coefficient of variation significance test was confirmed as a1 = 0.0167. Comparing with the items of problem analysis, the variations of single-choice, filling blank, and explanation conception items were significantly lower （P〈0.0167）. Conclusions Rote learning is the major method in learning this course.