班级规模、教师学历如何影响学生学业成绩?——基于PISA 2018的国际比较研究

How do Class Size and Teachers’ Academic Qualifications Affect Students’ Academic Performance?——International Comparative Research based on PISA 2018

缩小班级规模和提高教师学历是当前世界各国提升学生成绩的两种重要举措。本研究基于PISA 2018中国京、沪、苏、粤四省市与OECD国家数据,对班级规模和教师学历的分布特征进行国际比较,并采用多层线性模型分析班级规模、教师学历及其交互项对学生成绩的影响。研究发现,我国四省市中学班级规模平均比OECD国家高出13人/班,中学教师硕士学历比例比OECD国家平均低25.8%;从教师学历对应的班级规模来看,我国四省市硕士学历师资集中于班级规模为21-30人、36-40人两个班额区间的学校,OECD国家则表现出"大班额低学历、小班额高学历"的特点,硕士学历师资集中在16-20人的小班化学校;从班级规模对学生成绩的影响来看,以50人以上班额为基准,我国四省市班级规模缩小至41-45人、46-50人时,学生成绩提升明显,而OECD国家缩小至16-20人、26-30人、31-35人时,学生成绩提升更明显;从教师学历对学生成绩的影响来看,硕士学历教师比例对学生成绩呈现显著的正向影响,教师队伍中硕士学历比例每提升10%,我国四省市学生数学学科成绩平均提升10.2分,OECD国家学生成绩平均提升5.6分。对OECD国家班级规模进行二次项检验发现,班级规模对学生成绩的提升呈非线性关系,测算班级规模对学生成绩影响的边际效益最佳范围是21-30人/班。此外,对班级规模与教师学历的交互项进行检验,发现我国班级规模与教师学历对学生成绩是"相互独立"的影响效应,而OECD国家班级规模与教师学历对学生成绩则是"相互替代"的作用关系,两者均有特定的作用阈限。

Reducing the class size and improving teachers’ education qualifications are two important measures to improve students’ academic performance. With the data of PISA 2018 from Beijing, Shanghai, Zhejiang, and Jiangsu in China and OECD countries, this paper uses an HLM model to analyze the impact of class size, teacher education, and their interaction terms on student performance. The study finds that the average size of middle school classes in the four provinces and cities in China is 13 students higher than that in OECD countries, and the proportion of middle school teachers with a master’s degree is 25.8% lower than that in OECD countries on average. From the distribution of teacher qualifications in schools, the master’s degree teachers in the four Chinese provinces and cities are concentrated in schools with a moderate class size of 36-40 students per class. On the contrary, OECD countries show the pattern of "Large class size with lower qualifications and small class size with higher qualifications", with master’s degree teachers concentrated in small-class schools with 16-20 students per class. Regarding the impact of class size on students’ performance, Chinese students in classes with 41-45 or 46-50 students performed significantly better than those with more than 50 students. In OECD countries, students in classes with 16-20, 26-30 or 31-35 have significantly higher scores. In addition, the proportion of teachers with a master’s degree is significantly positively correlated with student performance. For every 10% increase in the proportion of master’s degree holders in the faculty team, the average student score in the four Chinese provinces and cities will increase by an average of 10.2 points, and that in OECD countries will increase by an average of 5.6 points. A further test of class size in OECD countries shows that class size has a non-linear relationship with the improvement of student’s performance. The highest marginal benefit of class size on student performance occurs in classes with 21-30 students. At the same time, the test of the interactions between class size and teacher’s education qualifications reveals that the effects of class size and teacher’s education qualifications on student’s performance are independent of each other in Chinese schools, while they are substitutional in schools in OECD countries with specific thresholds of impact respectively.