Sunday, October 12, 2014

The Art of Making Science Work

Between any scientific discovery relevant to a problem facing society, and the delivery of an effective policy solution, lies the no-man’s land of implementation uncertainty.

On 28th July 2014, I gave a talk at the University of Cambridge to a group of educational psychologists gathered at the ‘Implementing Implementation Science’ conference. The challenge, I explained, was about navigating through three sets of tricky issues:
* Being Scientific about Science
* Being Sensible about Public Policy
* Being Effective about Implementation

Here are my notes for the talk, addressing each of these key issues:

Being Scientific about Science
Rethinking our epistemological assumptions:

What do we really know about science? Does it give us a picture of reality or a succession of hypotheses?
o Science as a system of standard procedures & indisputable proof
o Science as continuous with intuition & spiritual understanding
o Science as common sense, but whose common sense?
o Science as a revisable probabilistic & empirically checkable set of tools

• Degrees of predictability in different spheres of investigation. What is expected in anticipating the movement of planets or particles may not be an appropriate model for anticipating the changing behaviour or attitudes of people.
o Policy makers and the naïve model of scientific certainty
o Acceptable variations of predictability v. Unacceptable fluctuations of reliability
o Cause & effect of behavioural changes

What are the criteria for replicability? The conditions for technological replication need to be adapted to the setting in question.
o Replicability in impact of physical clearance of harmful substance
o Replicability in impact of prescription of medication
o Replicability in impact of caring support

Being Sensible about Public Policy
Preparing for delivery obstacles:

The challenge of managing political expectations: how implementation failures can simply be a case of ‘over-hyping’ or ‘under-selling’.
o The twin nightmare of setting criteria for evaluation forms AND agreeing what to put into the press release.
• Over-hyping what may be achieved is a recipe for disaster.
• Under-selling the actual value of a policy is self-defeating (for it will lead to under-investment).

• The problem of guiding delivery: there is no such thing as ‘pressing a button’ in initiating a social policy solution.
o Even the paradigm of rolling out an inoculation programme has proven to be problematic.
o Recipients of treatment/initiatives are not mindless subjects.
o And people who ‘deliver’ the solutions are not mere mechanisms – e.g., of sending out disempowered staff to tell the public about a public body’s commitment to empowerment.

The need for patience and active learning: how lack of curiosity kills the policy mouse.
o Science’s key strength is its empirically-based revisability: e.g., when there are many variables, you have to learn from what actually happens – alley gate example and community safety.
o Community health projects: the need for relationships to build up in neighbourhoods.
o The financial regime stamps out cautious and responsible management of public funds (spend it or lose it syndrome – support for asset transfer).

Being Effective about Implementation
Promoting cooperative problem-solving in schools:

Research commissioned by the Carnegie UK Trust (& ALT) into the impact of student participation in schools and colleges found that :
o students were happier and felt more in control of their learning; while disruptive behaviour in class was reduced.
o it had the twin effect of teachers’ practice improving and students gaining in awareness of the learning process;
o enhanced skills of communication and competence as a learner;

Communication is an integral part of policy delivery: if you’re too vague, delivery is lost in translation – E.g., more participation -> formal school council with no major decision, and no wider engagement with most pupils

Training: People are partners, not tablets: you can’t just hand them out – E.g., restorative justice in schools: some heads claim they do it because they sit perpetrators and victims down to talk about it (BUT HOW is the key)

Evolving Procedures (Observation changes quantum outcome, interaction generates social impact): E.g., implementing solution development rather than delivering solutions in participatory budgeting. You need wider engagement, proper sampling & explaining, effective facilitating, before the exercise can deliver priorities young people take ownership in.