Is evidence good for absolutely nothing?

By Ben Durbin

Wednesday 28 September 2016

“What is evidence good for? Absolutely nothing!” This was one of the memorable moments from Campbell Collaboration CEO Howard White’s opening speech at the What Works Global Summit this week (quickly followed by the qualification: “Unless it gets into policy and practice”)

This set the tone perfectly for the conference that followed, with a rich array of sessions covering topics ranging from evaluating catch-up interventions in India to the lack of evidence supporting many non-school literacy interventions in developing countries to the challenges of ensuring major development donors embed evidence into the funding decisions.

The conference had a particular resonance for me personally.  For the past three-and-a-half years, I have led NFER’s Impact Team. As an education research foundation, we came to the same conclusion as Howard White – unless we were as successful at engaging policymakers and practitioners as we were at producing high quality research, then we would be good for nothing.

That was certainly the theory, and what followed was an iterative process of learning from others and learning by doing. In 2014 we published a literature review examining the evidence on effective approaches to mobilising knowledge for the classroom. We have worked closely with the Education Endowment Foundation (who received lots of favourable references at this week’s conference) on approaches to conceptualising and measuring research engagement, and to developing and evaluating interventions to mobilise research knowledge in schools. Through discussion with a range of stakeholders we’ve developed guidance and practical support addressing both the school-level cultural and leadership practices needed and the systemic and institutional changes required to enable a truly research-engaged education system.

We’ve also learned lessons about how to successfully influence policy with research. Many of these seem obvious, and there are lots of good examples of where these principles are followed, but this is by no means universal:

  1. Ask the right questions – what’s interesting to an academic may not also be useful to a policymaker.
  1. Engage early – don’t wait until your research is ready to publish before reaching out to policymakers and other stakeholders. Involve them in defining the problem, shaping the research questions (see 1), and throughout the life of the research (one of the conference sessions included a fascinating discussion of how DfID and the EPPI centre do this with their systematic reviews)
  1. Engage widely – map all of the stakeholders at the start of the research, focussing particularly on primary stakeholders with a direct ability to effect change (who are unlikely to move in academic circles) and the wider group an influence over them (also unlikely to be academics). Develop a differentiated strategy for how to engage each group.
  1. Communicate clearly – work hard at drawing out clear messages for policymakers, even (or especially) where these are nuanced. Communicate your findings using media suited to each stakeholder group (research summaries, practitioner toolkits, through social or traditional media).

I have now taken up a new role at NFER, as Head of International Education. My team is looking to bring NFER’s 70 year track record in high quality educational assessment, analysis, evaluation and reviews – and successful policy and practice influence – to bear in meeting global education opportunities and challenges.

I don’t know whether to be downcast or encouraged this week to realise quite how widespread the evidence challenges are, with common issues faced in all countries. But it does give me confidence that NFER has an important role to play in undertaking research that successfully engages with international policy and practice.  We’re ready to meet Howard White’s challenge.

Ben Durbin is Head of International Education at NFER. He can be contacted at b.durbin@nfer.ac.uk or via Twitter @benpdurbin