Complex realism in social research

In recent years, both realism and complexity have begun to have methodological influence in social research. Yet for the most part, these have existed separately and have had limited impact on empirical research. In this article, we develop a theoretical argument for complex realism, grounded in an ontology of probability, that may be operationalised to demonstrate the reality of social change at a micro- and meso-level. We apply our conception of complex realism to an example using the method of longitudinal case–based cluster analysis to analyse the trajectories over time of male and female prisoners aged 18 and above who were at risk of self-harm.