A couple of weeks ago I was speaking to someone who was complaining about the way medical researchers often overstep the mark when asked about the potential impact of their findings. He referred in particular to scientists on the BBC's Today programme. When asked when their specific breakthrough might be available, they would say, "'Oh, in about five years' time."
My companion's point was that no researcher could possibly know this. "You see," he explained, "many breakthroughs never make it out of the lab. For one reason or another, drug companies don't pick up the ideas, or the funding disappears, or there's a glitch which no one had anticipated at the time of the announcement."
This had never occurred to me, having never worked in a laboratory beyond O-level physics. I had always assumed - naively, it transpires - that, although there was still much work to be done, all such breakthroughs were translated into a useable product, an increase in knowledge, or an immediate change in public policy. Had I paused to think for a moment, I would have realised that this was not true. In fact, it is just another manifestation of what has become known in my trade as "physics envy".
Let me explain. Those of us who work in the social sciences have a tendency to look at our neighbours, who do not have the word "social" in their titles, and think how simple life must be for them to get their voices heard when they believe they have something to say. They can carry out experiments, control all those messy variables that interfere with a clear result, and then do it all over again in exactly the same way to show that they were right in the first place.
The difficulty for those involved in educational research is that classrooms are not labs. At any given point there are so many different factors at play that nothing done on one day can ever be precisely reproduced. Where physics appears to be able to say something quantifiable and certain, education can only say, "Last Tuesday it looked like this."
Of course, neither scenario is true. The advice of scientists struggles to get into the public domain partly because the results are often not as clear-cut as they might appear at first. And even when they are, what politicians should do about the findings must be weighed against any number of other considerations.
Similarly, educational research can amass so much evidence from a variety of studies, or from the scale of one study, that the cumulative effect of the findings becomes overwhelming. Even small-scale research projects can provide insights that are illuminating and speak powerfully about a particular issue.
The Government has invested a great deal in educational research via an extensive programme, carried out in three phases, called the Teaching and Learning Research Programme. Funded by the Economic and Social Research Council, its aim - as the title suggests - is to focus on teaching and learning, with a view to producing findings that can be applied in all sectors of education. The first phase is drawing to a close now.
But the question remains: how will the Government use the programme's findings? And how will those who have been awarded the money communicate what they believe to be salient in their work to the policy-makers?
The relationship between policy and research is always going to be fraught.
Any good research has a complexity that takes time to absorb and apply. It is a feature of this work which is not always appreciated by those who are working to an electoral timetable. Yet it has to be a better starting place for reform than relying on the whims of a politician or the next bright idea of a policy pundit.
Bethan Marshall is a lecturer in English education at King's College, London