The latest in the CAHRU/LIH research methods seminars was given by Dr Paul Leighton on 24 October 2017 on the subject of ‘Realist evaluation in health services research’. Paul Leighton is a senior research fellow at the University of Nottingham, deputy director of the NIHR Research Design Service for the East Midlands and an expert in health services research, qualitative methods and process evaluation. The presentation covered the philosophy of critical realism and how this related to realist methods, described a practical example of a realist evaluation and finished with concluding thoughts and comment on the method.
Critical realism was described as a philosophical approach developed by Roy Bhaskar that adopts an ontological position of being in which: the actual world is independent of the human mind and the mind is capable of perceiving the empirical world, but unable to directly perceive unseen forces in the real world. Gravity and social class were cited as examples of unseen forces in the real world.
“Realist” or “realistic evaluation” was described as a form of primary research methodology derived from critical realist philosophy, by Ray Pawson and Nick Tilley, which acknowledges the potential for unseen forces and accepts that interventions will not always work or work in the same way because of local contextual factors will influence how an intervention works and the impact it has. Realistic evaluation was summed up as “What works, for whom, in what circumstances?”
The process of realistic evaluation involves developing a programme theory, which theorises what should happen and how it might happen. This so-called mid-range theory is tested using qualitative and/or quantitative methods to explore the context, mechanisms (often hidden) and outcomes (C+M=O), revised iteratively and tested again, leading to a refined mid-range theory.
This was illustrated using the Falls in Care Homes (FinCH) study and summarised by concluding that interventions will not work all the time in all settings, but will work differently in different contexts.
A. N. Siriwardena