Science education research feeding into policy and practice in South Africa: The strengths of narrative and systematic reviews

This paper responds to the Marang Centre's narrative-systematic review of recent science education research in South Africa, and its power to inform policy and practice. This response explores the question to what extent the review findings would be different if a systematic review method would have been used instead. The following five stages of a reviewing process are used to structure a comparison: the identification of the review questions; the development of inclusion criteria; the identification of the literature search strategy; the application of keywords and descriptors to categorise research papers; and the strategies for synthesising research findings. For both review methods the development of the review questions, the selection of inclusion criteria and key-words yield very similar results. The wider angle of review questions in narrative-systematic reviews increases opportunities for identifying 'drivers' and 'connectors' underlying the reviewed body of research, whereas the explicit nature of the inclusion criteria and key-words in a systematic review facilitates the process of mapping the existing research. The major differences between both review methods lie in the adopted search strategy, the nature and detail of the descriptors used, and the method for synthesising research outcomes. For systematic reviews, quality judgements are constructed for each research report which allow weighing its relative contribution to the synthesis of the research findings. In the narrative-systematic method implicit, and external, quality judgements are included by using the refereeing process as a major searching tool. The high trustworthiness of systematic review findings as a basis for policy decisions and changes in educational practice contrasts the considerably higher demand on time and cost, and the more limited scope of review findings.