I think the reason people may not have considered or voiced this question before is that the approach of Realist Synthesis does not presume that rich insights and findings about how programmes or policies work will reside in any particular types of document. It's all situation-specific.
In fact, in my field of health care and public health, research articles (which are also simply documents of course) about a programme are often very poor at revealing the underlying rationale and causal assumptions that underlie a proposed or evaluated health service change. In contrast policy documents, or (for example) the minutes of policy or commissioning meetings, in some cases will provide the only or a more rich source of the underlying rationale of programmes and explicit statements about way people believe service changes will make things better for patients/service users. However, in my experience policy documents will often frame and express programme theory in higher level terms of policy rhetoric, without clearly specifying how the specific combinations of programme elements will work together to produce the expected outcomes.
As for the other problem you allude to - how to find the particular types of policy/service documents which will provide rich data and insights about programme/policy theory and explaining programme/policy success or failure - there must be others on RAMESES who have some experience of doing this (I do not).
Hope my reflections are useful. Good luck!
ESMI, University of Exeter Medical School
From: Realist and Meta-narrative Evidence Synthesis: Evolving Standards [mailto:[log in to unmask]] On Behalf Of Arun Verma
Sent: 18 April 2018 10:19
To: [log in to unmask]
Subject: Using realist synthesis to evaluate documents?
Good morning everyone,
I'm starting to get to grips on the complexity of literature around realist methodologies. However, I have a question (which I'm not sure if it has come up before).
I would like to evaluate policy/service documents for a project I'm doing using a realist framework. Is there a way in which one can conduct an evaluation of documents using a 'Realist synthesis' protocol? Are there any papers on this - I'm struggling to be creative with keywords in my search at the moment.