Conditions for Successful Tutoring: A Reanalysis of LAU and KESS Data

在德国，超过20%的中学生接受私人辅导。在某些情况下，这些辅导班甚至由国家补贴。然而，目前的研究状况对辅导是否达到了其主要目标：提高学生的学习成绩尚无定论。特别是，以下问题仍然悬而未决：在学生、教师和课程的哪些条件下，辅导在提高学生的成绩和动机特征方面是成功的？基于对家长家庭作业协助和常规课堂教学质量的中心研究，我们开发了一个差异化的辅导效果模型。在联合研究项目MILES（纵向教育研究中的方法论问题）的背景下，我们将使用汉堡小组研究KESS和LAU的数据来评估该模型。我们将重点关注辅导课的持续时间和强度、导师的资格、教学内容、导师反馈学生时的行为以及学生的动机和成就相关先决条件对学业自我概念、表现焦虑和工作习惯等成就和动机变量的影响。在方法层面上，我们将使用倾向得分匹配程序来建立有和没有私人辅导的学生的平衡样本。课程特点和学生特点的影响将通过回归分析来估计。与此同时，将开展三个较小的研究，以检查该模型中无法用LAU和KESS数据检查的方面。这些研究将集中在辅导期间的家庭作业辅导质量、教学质量的差异化测量以及辅导对校外学习时间持续时间和变异性的影响。

In Germany, more than 20 percent of secondary school students receive private tutoring. Under some conditions, these tutoring classes are even subsidized by the state. However, the current state of research is inconclusive about whether tutoring reaches its main goal: to improve students' achievement. In particular, the following question remains open: Under which conditions on the part of the students, the teachers, and the lessons, is tutoring successful in terms of improving students' achievement and motivational characteristics? Building on central research on parental homework assistance and instructional quality in regular classes, we developed a differentiated model of the effects of tutoring. In the context of the joint research project MILES (Methodological Issues in Longitudinal Educational Studies) we will use data from the Hamburg panel studies KESS and LAU to evaluate this model. We will focus on the effects of duration and intensity of the tutoring classes, the tutor's qualification, the teaching content, the tutor's behavior when giving feedback to the students, and motivational and achievement-related preconditions of the students on achievement and motivational variables such as academic self-concept, performance anxiety, and work habits. On the methodological level, we will use propensity score matching procedures to build balanced samples of students with and without private tutoring. The effects of the characteristics of the lessons and the students will be estimated by regression analyses. At the same time, three smaller studies will be developed to examine aspects of the model which cannot be examined with the LAU and KESS data. These studies will focus on the quality of homework assistance during tutoring, on the differentiated measurement of instructional quality and on the effects of tutoring on the duration and variability of out-of-school learning time.