大学生学业成绩的宿舍同伴效应分析

Roommates' Peer Effects on Undergraduate Students' Academic Achievement

大学本科期间是大学生专业成长的关键时期,接触最为密切的宿舍同伴之间的互动对大学生个体有重要影响。本文借助某一本院校3460名本科生的数据,利用新生入学时随机分配宿舍这一项自然实验,使用普通最小二乘法和面板数据模型来检验大学生学业成绩的宿舍同伴效应。研究发现,同伴平均标准化GPA每增加1个标准差,大学生个体的标准化GPA大约增加0.03个标准差。尽管学业成绩的宿舍同伴效应平均来看显著为正,但不同学习能力个体受同伴影响的效果不同,低学习能力者的学业成绩随着同宿舍高或低学习能力者比例提高而下降。当将低学习能力者的一位中等学习能力的同伴换成高或低学习能力者时,该生标准化GPA将降低约0.13个标准差。本研究发现将为在教育第一线从事学生工作的人提供更具有可操作性的建议,分配宿舍时尽量安排低学习能力和中等学习能力的学生同宿舍,高学习能力者同宿舍。

College life is a crucial period for undergraduate students' professional growth, and the interaction of intimate roommates has important impact on individual students. Based on the data of 3460 undergraduates in a certain institution, using the ordinary least squares method and the panel data model,this paper estimates peer effects on college students' academic achievement in a university context where freshmen are randomly assigned to dorms. The results show that while peers' average standardized GPA increased by one standard deviation, the individual's standardized GPA increased by 0.03. Even though on average, the peer effect on academic achievement seems significantly positive,the effect varies for individual with different learning ability, and the academic achievement of the low learning ability decreases with the increase of the proportion of the peers with high or low learning ability. If the middle-achieving peer of a low-achieving student is replaced by a high-achieving or low-achieving peer, the student's grade will decrease by 13 percent of a standard deviation. This study will provide more practical suggestions for those who work at the forefront of education field that when dormitories are assigned, students with low learning ability should be arranged with middle learning ability in the same dormitory, and students with high learning ability could be arranged in the same dormitory.