L’affidabilità dei criteri di inclusione nelle meta-analisi in educazione: una rassegna di studi

Research syntheses, such as meta-analyses and systematic reviews, are methods for combining results of different primary studies on a certain theme. These methods have been widespread in the early eighties in educational research with the purpose of giving more reliable information to the teaching practice. As primary studies, not all the reviews carried out are reliable to inform practice on programs and strategies that are effective for learning. Although some systematic reviews and meta-analyses have weaknesses, it is possible to identify which procedures and standards are more valid and reliable for carrying out metaanalyses. This article reviews and examines studies that have evaluated methodological factors that affect effect sizes in meta-analyses of educational practices. The studies of this review have showed that the following methodological factors affect effect sizes: publication bias, sample size, study design, outcome measures and intervention duration. The conclusion specifies which inclusion criteria, based on the review results, are more reliable to carry out meta-analyses that have the objective to inform educational practices.