Developing and testing intervention theory by incorporating a views synthesis into a qualitative comparative analysis of intervention effectiveness. (20th February 2019)

• Record Type: Journal Article
• Authors: Melendez‐Torres, G.J.; Sutcliffe, Katy; Burchett, Helen E.D.; Rees, Rebecca; Thomas, James
• Is Part Of: Research synthesis methods. Volume 10:Number 3(2019)
• Page Start: 389
• Page End: 397
• Publication Date: 2019-02-20
• Subjects: intervention theory -- qualitative comparative analysis -- qualitative meta‐synthesis -- systematic review; Research -- Methodology -- Periodicals; Research -- Statistical methods -- Periodicals; 507.2
• DOI: 10.1002/jrsm.1341
• Languages: English
• ISSNs: 1759-2879

Abstract

Abstract : Qualitative comparative analysis (QCA) was originally developed as a tool for cross‐national comparisons in macrosociology, but its use in evaluation and evidence synthesis of complex interventions is rapidly developing. QCA is theory‐driven and relies on Boolean logic to identify pathways to an outcome (eg, is the intervention effective or not?). We use the example of two linked systematic reviews on weight management programs (WMPs) for adults—one focusing on user views (a "views synthesis") and one focusing on the effectiveness of WMPs incorporating dietary and physical activity—to demonstrate how a synthesis of user views can supply a working theory to structure a QCA. We discuss how a views synthesis is especially apt to supply this working theory because user views can ( a ) represent a "middle‐range theory" of the intervention; ( b ) bring a participatory, democratic perspective; and ( c ) provide an idiographic understanding of how the intervention works that external taxonomies may not be able to furnish. We then discuss the practical role that the views synthesis played in our QCA examining pathways to effectiveness: ( a ) by suggesting specific intervention features and sharpening the focus on the most salient features to be examined, ( b ) by supporting interpretation of findings, and ( c ) by bounding data analysis to prevent data dredging.