【摘要】 教育增值评价是对研究对象在一定时期的“净产出”进行评价。从增值评价模型、分析方法、实证研究和本土化实践方面展开评述。首先，根据增值评价方法基于的统计理论将其划分为基于描述统计、统计回归和机器学习算法三类。然后，对这三类增值评价方法的数据要求、分析软件、优缺点以及实践应用进行介绍和评价。最后，对所有增值模型进行综合比较，提出模型选择建议，并从深入增值评价理论研究、构建“以测助评、以评促教”良性循环、重塑增值价值理念三个维度讨论了增值评价未来研究方向。更多还原

【Abstract】 Educational value-added assessment is a kind of process evaluation, which evaluates the “net output” of the research object in a certain period. The research of value-added assessment includes value-added model, analysis method and practical application. Firstly, according to the statistical theory, value-added evaluation method has been divided into three categories based on descriptive statistics, statistical regression and machine learning algorithm. Then, the study introduced and evaluated the data requirements, analysis software, advantages and disadvantages as well as practical applications of these three value-added evaluation methods. Finally, it made a comprehensive comparison of all the value-added models, and put forward some suggestions on model selection, and discussed the future research direction of value-added evaluation from three dimensions: depth theoretical research of value-added evaluation, the need to build a virtuous circle of “evaluation and teaching”, and the reconstruction of value-added value concept.更多还原