摘要：It is often not possible to realize ideal evaluation standards when it comes to evaluating modifications to educational settings. In this article theoretical and practical problems in the evaluation of modifications in educational settings are discussed. Based on these considerations the ENDIT model of evaluation is presented. It comprises five minimal standards necessary for a convincing evaluation: 1) effect establishment, 2) control of the novelty effect, 3) discriminant validation, 4) superiority over compared to implicit control groups, and 5) time-delayed control group. The five standards are explained and their utility for research is demonstrated, by way of an example, through the evaluation of a visualization tool that was introduced in order to increase participation in an e-mentoring community. Participants in the investigation comprised 231 female high-school students participating in the e-mentoring community CyberMentor that aims at increasing interest and participation in STEM (Science, Technology, Engineering, and Mathematics).
