考研分数能否精准预测研究生学业表现——基于倾向值匹配(PSM)的实证分析

Can Postgraduate Academic Performance Be Accurately Predicted with the Entrance Examination Score:An Empirical Analysis based on Propensity Score Matching

学术界对研究生选拔方式长期存在矛盾冲突。一方面,考研分数作为研究生水平的预测指标饱受质疑,另一方面,如果采取本科学业表现作为选拔标准尚缺乏实证验证。使用研究生入学之后各个学业表现指标,通过PSM倾向得分匹配等多个模型从各方面验证影响因素。结果表明,考研分数只影响学生硕士阶段GPA成绩,而学生本科阶段学业表现影响研究生科研与竞赛表现,考研面试成绩影响研究生毕业论文成绩。研究反映出当前研究生选拔制度的弊端,并从相关角度提出政策建议。

There are long-standing disputes in the academic circle on postgraduate selection method.On the one hand,postgraduate entrance examination scores used as a predictor to judge the level of postgraduates is in question.On the other hand,if it is better to use undergraduate study performance as the selection standard still lacks empirical verification.This study uses several models such as PSM tendency score matching to analyze the academic performance indicators of the postgraduates matriculated so as to verify from various aspects the influencing factors.The results show that the postgraduate entrance examination score only influences the GPA score of postgraduates during the period when they study for master's degrees,the academic performance of postgraduates in their undergraduate study period influences the research and competition performance of postgraduates,and the interview score from the postgraduate entrance examination influences the score of dissertation for graduation.This study exposes the disadvantages of the current postgraduate selection system and puts forward some policy suggestions from relevant perspectives.