'A Theory of Evidence for Use',
Sponsored by British Academy, BARDA scheme; October 2009-2011
Investigator: N Cartwright
Short description:
This project aims to provide a systematic account of evidence for policy effectiveness that can be of practical use in evidence-based policy. Its fundamental assumption is that whether an intervention will be effective in situ depends on the causal structure of the situation in which it is implemented and on its method of implementation. This assumption generates criteria for relevance and evaluation. Modelling the causal structure can seem a tall order. But, it is argued, it is an enterprise we do often, albeit informally, and do reasonably well. And there are a number of heuristics that can help.
This project aims to translate this assumption into an accessible account of how to marshal the right kinds of evidence for predicting outcomes. In particular the project will focus on what philosophers call 'INUS' conditions (which epidemiologists call 'sufficient causal complexes') – the auxiliary conditions necessary for a cause to produce its canonical effects, and on answers to the question, how would the proposed intervention as implemented achieve the targeted effect? Reasonable, evidence-based assessments of these can often generate a good enough grip on the causal structure to provide relatively reliable predictions about policy outcomes.)