In educational research, there is the general trade-off that empirical evidence should be generalizable to be applicable across contexts; at the same time, empirical evidence should be as specific as possible to be localizable in subject-specific educational interventions to successfully transfer the empirical evidence to educational practice. This trade-off is further increased by the fact that the diverse instructional contexts, such as school or student characteristics constrain the applicability of empirical evidence. Several approaches have been proposed to address this issue, however, emphasized the different problems (i.e., localization, generalization, transferability) rather in an isolated manner. To this end, in this article, we introduce a synergistic approach, the LoGeT (localize, generalize, transfer) model, which systematically integrates co-design (localization strategies) and ManyClasses principles (generalization strategies) with co-constructive transfer activities, to generate empirical evidence that may be applicable in educational practice. To illustrate the LoGeT model, we present three long-term projects, covering different granularities and durations of educational interventions across different fields of education (teacher education, adaptive teaching, non-interactive teaching) that successfully applied the LoGeT approach. Finally, we outline further directions for future iterations of the LoGeT model. We hope that the LoGeT approach may be a stimulus to guide researchers as well as practitioners alike to design generalizable and evidence-based educational interventions that are rooted in localized instructional contexts.

在教育研究中，存在一个普遍的权衡：经验证据应该具有普遍性以适用于各种情况；同时，经验证据应尽可能具体，以便在特定学科的教育干预中进行本地化，从而成功地将经验证据转移到教育实践中。由于学校或学生特征等不同的教学环境限制了经验证据的适用性，这一事实进一步加剧了这种权衡。已经提出了几种方法来解决这个问题，然而，它们以孤立的方式强调了不同的问题（即本地化、泛化、可转移性）。为此，在本文中，我们介绍了一种协同方法，即 LoGeT（本地化、泛化、迁移）模型，该模型系统地将协同设计（本地化策略）和 ManyClasses 原则（泛化策略）与共建迁移活动相结合，以产生可应用于教育实践的经验证据。为了说明 LoGeT 模型，我们提出了三个长期项目，涵盖成功应用 LoGeT 方法的不同教育领域（教师教育、适应性教学、非交互式教学）的不同粒度和持续时间的教育干预。最后，我们概述了 LoGeT 模型未来迭代的进一步方向。我们希望 LoGeT 方法能够成为一种激励，引导研究人员和实践者设计植根于本地教学环境的普遍且基于证据的教育干预措施。