Realist methodology in practice: translational findings from two realist syntheses

The author is currently conducting two rapid Realist Syntheses, one to identify the theoretical bases of closed-circuit television (CCTV) to reduce alcohol-related assault in the night time economy, and the other to identify dimensions of evaluation to improve the effectiveness and efficiency of a number of services in northern Australia which address homelessness and alcohol-harm reduction. The CCTV project grew out of a "completed" Realist Evaluation; the homelessness and alcohol-harm project is the foundation for a future Realist Evaluation. This paper will examine how the Realist Synthesis protocols have been applied both retrospectively, and to inform the future Realist Evaluation. Each evaluation aims to understand how specific interventions work, or don't work, using the explanatory structure of generative causation. Key findings are: that precise definitions of the programs' outcomes are crucial to retrospectively applying the Realist Synthesis methodology; that the realist methodology can embed a continuous improvement process in the funding organisation once these outcomes are defined, making research engagement more effective; that the outcomes (and causal mechanisms) lie at different systemic levels, both internal and external to the organisation; and that this last point is something people within funding organisations intuitively grasp, but have difficulty understanding.