Last year, the European Commission published a review of findings from research it had funded into adult continuing education. Its author’s comments on corruption as a cause of policy failure in some countries are interesting enough, but I was even more struck by what he had to say about how to exploit computer software to support the process of policy-making.
Paolo Federighi’s basic proposition is appealingly simple: policies based on evidence are likely both to command public support and produce the expected results. In Federighi’s words, ‘The adoption of smart and intelligent policies depends on data that can guarantee the pertinence of public support and demonstrate the adequacy of the impact produced’.
These are appealing goals, particularly in an area like adult learning, where policy makers are often reluctant to commit resources. If we can persuade policy makers that we know which types of adult learning intervention can produce positive benefits, then surely that can only be good for our hard-pressed field. But in our messy, fuzzy, anarchic field of practice, how can we produce neatly packaged bundles of evidence that might be useful to busy policymakers?
Federighi’s solution is to apply something called Intelligent Decision Support Systems (IDSS) to the field of adult education research. Basically, an IDSS uses artificial intelligence, machine learning, taught algorithms and data analytics to help support decision-making in real-time, by setting out possible courses of action and evaluating the likely results of these proposed actions.
The growth of IDSS development and use in recent years has been substantial, though I have to admit that it had passed me by until I read Federighi’s report. They’re used in healthcare to assist doctors and nurses in making efficient clinical decisions, in the financial industries to help identify portfolios of investment, and in traffic management to model flows of vehicles and pedestrians. And now the suggestion is that they can be used to support policy-making in education.
Some people will object to this very notion on principle. My Stirling colleague Ben Williamson (who helpfully explained to me what an IDSS is) has written more broadly about the ways in which our lives are shaped by algorythms and digital technologies. And you can see why some might react to this process with abhorrence.
I’m also entirely unpersuaded by the practicability of this idea. The best algorithm in the world can process the information available to it. Adult continuing education is a vast field, yet the resources devoted to rigorously studying it are laughably small.
And even if we had the best research in the world – which we don’t – policy making is itself a complex process, whose actors (contrary to the firm prejudices of many academics) are well-informed, anxious and extremely clever. One group of policy makers – politicians – will want the IDSS to take account of electoral responses to any given intervention. Other groups – civic servants, those who run quangos, and those who manage local implementation – will have quite different needs.
So I am extremely sceptical about the application of IDSS to policy making in adult learning. Nevertheless, the idea now has some momentum within the European Commission, and we can therefore expect to hear more of it. Still, comfort yourself with this thought: another feature of policy is that by the time someone has investigated and reported on the feasibility of IDSS in adult continuing education, the officials who promoted the idea in the first place will all have moved on.