Developing complexity-consistent theory in a realist investigation

Abstract

This is the second of a two part article examining the use of complexity-consistent theory in a realist investigation. The first article (Evaluation 18[4]) introduced the idea of complexity-consistent theory. It argued that complexity-consistent theories are likely to be useful for understanding complex processes of change and therefore useful in many kinds of evaluation. Further, it argued that theories can be organized within a hierarchy to reflect the different levels of reality involved in a change program. This was described as ‘layering’ theories. This second article describes the use of these concepts in a realist investigation. The purpose of the investigation was to develop a realist, middle-range theory to explain how and why some early intervention programs which ‘work’ on average for disadvantaged families do not work for the most disadvantaged. The investigation combined a small-scale realist evaluation of a family support program in Adelaide, South Australia with a modified form of realist synthesis. The work was undertaken as part of a PhD, supervised by Professor Nick Tilley, co-author of Realistic Evaluation.