Effects of the Attributes of Educational Interventions on Students' Academic Performance: A Meta-Analysis

This meta-analysis examined the influence of attributes related to the implementation of learning strategy instruction interventions on students' academic performance, and also examined how the attributes related to the method of testing the intervention effects affected the actual effects measured. Using metaregression, we analyzed the influence of the subject domain in which the intervention was implemented, the implementer, its duration and intensity, student cooperation, and research method aspects (including measurement instrument). Most attributes moderated the intervention effect. Using forward regression analysis, we only needed four attributes to obtain the best model, however. This analysis showed that the intervention effect was lower when a standardized test was used for evaluation instead of an unstandardized test. Interventions implemented by assistants or researchers were more effective than those implemented by teachers or using computers. Cooperation had a negative, and session duration a positive, contribution. Together, these attributes explained 63.2% of the variance in effect, which stresses the importance of emphasizing not only the instructional focus of an intervention but also its other attributes.