An analysis of theory building in realist synthesis: perspectives from the literature

Article type
Authors
Hunt H1, Hardwick R1
1University of Exeter Medical School, University of Exeter, UK
Abstract
This presentation will be made jointly by Rebecca Hardwick and Harriet Hunt from the University of Exeter Medical School, UK.

Background: Realist reviews offer an approach to evidence synthesis which moves beyond describing whether something works, to explaining who it works for, in what circumstances, in what respects and why. Theory-driven in approach, Realist review sees evidence synthesis as an iterative process of theory building and refinement. Whilst the theoretical and methodological approach to theory building has been comprehensively described elsewhere, the practical tasks of theory building and refinement are less well defined within the literature.

Objectives: We aim to analyse existing realist reviews to explore how authors describe what they did when theory building, and provide examples and guidance to others wanting to apply the method.

Methods: We analysed a list of sources previously assembled by the authors of the RAMESES publication standards for realist syntheses and meta-narrative reviews (N = 29). We were looking for evidence of whether and how authors describe the process of theory building within their review. We based the data extraction form on the identified sources and piloted the form on one source.

Results: Very few sources described theory building within their abstracts, although many more detailed the product of their theory building activities, i.e. the theoretical frameworks that were developed via an undefined process. We provide a detailed breakdown of the descriptions of theory building within the identified sources.

Conclusions: Practical theory building takes place in a multitude of ways, yet here we have shown that within a purposive sample the descriptions of these steps not only lack detail but are often not mentioned at all. For realist reviewers this omission is unhelpful in both designing and evaluating realist syntheses. For this reason we intend to create a companion to theory building within realist synthesis, using two practical examples from the authors’ own work.