Issues in synthesizing research in education

Abstract

Background Systematic reviews of educational research require synthesis of the results from a range of studies. When these studies are diverse in nature and methodological approach, synthesis can be problematic.

Purpose To offer a critique of the process of synthesizing systematic reviews by considering the stages of the process and the problems raised at each stage.

Source of evidence A brief review of the literature on the role of systematic reviews in education is presented, together with a description of the staged process of a systematic review.

Main argument Synthesis of systematic reviews is difficult on account of the fact that they are heterogeneous across different aspects of the research process. Moreover, synthesis of such systematic reviews may be neither feasible nor desirable.

Conclusions Clear and manageable research questions are an important prerequisite for synthesis; Bayesian synthesis should be used where appropriate; and the minimization of bias must be maintained through the stages of synthesis in any research.

Keywords:

• Systematic reviews
• Meta-analysis
• Synthesis
• Narrative synthesis
• Educational research

Acknowledgements

This article was first presented as a paper at the Department for Education and Skills Research Conference 2004, ‘Research in Education: What Works?’, Queen Elizabeth II Conference Centre, London, 19 November 2004. The authors are grateful to Stephen Gorard and Carole Torgerson for comments on an earlier draft of the article.