Teacher Effectiveness and its Determinants in Germany: A Deepened Analysis

教育经济学中关于教师有效性的知识状态可以简单地概括为两句话：1.教师在学生的发展中起着关键作用。2.教师的具体特征对学生的表现有何影响，目前尚不清楚。许多研究，特别是在美国进行的研究表明，教师素质的差异是相当大的，只有少数教师特质，如经验、学科知识或相同性别，才能适度解释这种差异。目前还不清楚这些研究在外部是否适用于德国的教育系统，因为德国的学校系统和教师培训都有很大的不同。其鲜明特点的例子是有组织的教师培训计划、早期跟踪以及私立和公立学校的有限竞争。国家教育小组研究首次将计量经济学方法应用于德国，得出了关于教师素质变化和教师特质与学生表现之间联系的可信结果。该项目由三部分组成。首先，我们希望在前一个资助期工作的基础上，使用在第二个资助期将获得的数据，以加强我们估算结果的统计能力，从而能够对德国的教师效率及其决定因素得出更可靠的结论。此外，我们希望将之前的研究扩大两倍。在第二部分中，由于非净资产数据的特殊新兴结构，我们可以将以前依赖于两年业绩收益的估计方法扩展到基于一年业绩收益的方法。从方法论的角度来看，这一点特别有趣，因为文献的标准是基于一年的收益。此外，这两种估计方法的比较对于评估我们对第一个资助期的结果和潜在的政策含义至关重要。在第三部分，我们希望在第二个资助期结束时利用面板数据显著延长的时间维度来改进估计方法。到那时，NINS数据将允许清理学生表现差异的混杂因素。我们希望将这一清洁措施分为教室和教师部分，这将允许更好地识别教师对学生成绩的因果影响。在他们的实体中，这三个项目部分与第一个资助期的结果相关联，将为德国教师及其特征对学生成绩的影响提供关键证据。根据这些结果，可以为当前的德国教育政策提供深刻的政策建议。

The state of knowledge in the economics of education on teacher effectiveness could briefly be summarized in two sentences: 1. Teachers play a critical role in the development of their students. 2. The specific teacher characteristics that account for their particular effects on their students' performance remain mostly unclear. Many studies, particularly conducted in the US, show that teacher quality variation is quite substantial and that only few teacher traits such as experience, subject knowledge or the same gender can moderately explain this variation. It is yet unclear whether these studies are externally valid for the German education system as it differs quite considerably in its school system as well as in its teacher training. Examples of its distinctive features are the structured teacher training programs, the early tracking and the limited competition of private and public schools. The National Educational Panel Study (NEPS) makes it possible for the first time to apply econometric methods for Germany that yield credible results about the variation of teacher quality and about the link between teacher traits and student performances.This project consists of three parts. First, we want to build upon the work of the previous funding period and use the data that will become available over the second funding period to strengthen the statistical power of our estimation results, thereby allowing for more robust conclusions on teacher effectiveness and its determinants in Germany. Furthermore, we want to extend our previous study twofold. In the second part, due to the specific emerging structure of the NEPS data, we can extend our previous estimation techniques that relied on two-year achievement gains to methods that will be based on one-year achievement gains. This is particularly interesting from a methodological point of view, as the standard of the literature is based on one-year gains. In addition, the comparison of these two estimation methods is critical for the assessment of our results of the first funding period and the potentially developed policy implications.In the third part, we want to exploit the considerably extended time dimension of the panel data by the end of the second funding period to improve the estimation methods. By then, the NEPS data will allow the cleansing of the variation in student performances of confounding factors. We want to divide this cleansed measure into a classroom and a teacher component, which will allow for a better identification of causal teacher effects on student achievement.In their entity, these three project parts in association with the results from the first funding period will provide critical evidence on the effects of German teachers and their traits on student achievement. Profound policy advice can be made on the basis of these results for current German education policy.