Evidence for Policy Prediction: Intervention-centred or Context-centred?

There are two approaches to thinking about evidence for predicting policy effects in situ – here and now. One is Intervention-centred; the other, Context-centred. This presentation will lay out the central differences between the two, describe situations where one would likely be far better than the other, and outline their relative strengths and weaknesses.

Not surprisingly, the intervention-centred is more ‘manualisable’ since it depends primarily on business as usual in the sciences, whereas context centring constitutes a kind of social technology and requires amalgamating evidence in a case-by-case manner with no well-rehearsed methodology to fall back on.

That makes troubles for us. Because when it comes to actual success in situ, context matters. So we must beware of putting all our eggs in the easier-to-manage intervention-centred basket. Recent emphasis on ‘what works for whom where’ moves the two approaches a little closer together, but not enough, I worry, to do the job well.


Part of the Trinity Term MethodsHub seminar series.

Arrive early to avoid disappointment.