Realism on the edge: where research and evaluation processes need their own CMOs  

This paper discusses three recent and challenging realist inquiries that required tweaking standard practice to maintain realist principles: an evaluation of a reintegration program for Aboriginal women leaving prison in the... [ view full abstract ]

This paper discusses three recent and challenging realist inquiries that required tweaking standard practice to maintain realist principles: an evaluation of a reintegration program for Aboriginal women leaving prison in the Australian central desert, where two of the co-authors were themselves program clients; a realist-informed assessment in East Sumba of how irrigation is impacting on potable water and sanitation access, with interviews conducted by local young people in a language for which no written form exists; and an evaluation headed by a Javanese doctor of the variable take-up of Frontline SMS by health workers in a number of eastern Indonesian districts, with culture emerging as an important factor. In each of these projects, realist concepts had to be translated not just linguistically but also culturally, from gaining acceptance of realist ontology and epistemology from Aboriginal women who spoke English as a third or fifth language and had their own distinctive cosmology and axiology, to tweaking methodology to deal with cultural situations that influenced who talked about what, and in what ways, in different Indonesian islands. In order to maintain rigour, it proved important to analyse the evaluation contexts, the mechanisms they triggered or prevented, and the demi-regularities in discovery outcomes that resulted. (For example, analysis of transcript pronouns in Bahasa Indonesia – the language in which the interviewers transcribed their Sumbanese interviews, as that is a purely oral language – was used to distinguish cases where interviewers were likely to be capturing their own hypotheses rather than interviewees’ perceptions.) Although the challenging circumstances in which these inquiries were conducted made this type of analysis imperative, the results indicate that – given realists' position on epistemology – even ‘mainstream’ inquiries could benefit from conducting this type of analysis and presenting it in reports.