Education researchers spend a lot of time worrying about implementation issues. In an ideal world, a promising approach would first be tested by its developers. It would then be tested on a relatively small scale by researchers who are independent of the developers before being rolled out for a much larger test ‘at scale’. Many promising approaches stumble along this path.
Indeed, on Twitter, Dylan Wiliam has been pointing to the fact that Reading Recovery has seemingly passed through all of these stages and has challenged advocates of systematic synthetic phonics to point to a programme they prefer that has done the same.
However, I can’t help wondering if we are thinking about implementation in the wrong way. We may be viewing it as the purely technical exercise of getting teachers to do and say certain things and neglecting the values and beliefs of those teachers. Nobody operates in an ideological vacuum and there are as many ways to subvert, consciously or otherwise, the intentions of the developers as there are teachers and classrooms.
In the diagram above, I have tried to summarise the technical and ideological problems of implementation as two axes, giving us four quadrants. The technical axis is based upon just how contained the intervention is, with the idea that smaller, more neatly packaged ‘interventions’ are technically easier to implement than wholesale changes to teaching practice.
In the top right quadrant there are discrete, packaged interventions that align with the dominant ideology. These are essentially the kinds of things that the UK’s Education Endowment Foundation spends most of its time testing.
In the bottom right quadrant are packaged interventions that do not ideologically align. The current Education for Learning appraisal of MiniLit, a phonics-based intervention for struggling readers, might fit in here. It will be interesting to see how the ideology gap affected this study when the results are published.
The top left quadrant represents entire approaches to teaching and learning that align with the dominant ideology. Wiliam’s own work on formative assessment might fit here. Formative assessment takes no ideological stance at all, fitting equally well with competing philosophies. However, it does require a wholesale change to teaching practice. It is interesting that one strategy Wiliam has developed to aid implementation is to ask teachers to focus on implementing only one technique at a time, shifting it more towards a packaged intervention.
In the bottom left corner sits the kinds of approaches that I promote on this blog such as using systematic synthetic phonics for initial reading instruction, explicit teaching in general and the adoption of a knowledge rich curriculum. These are clearly going to be the hardest of all to implement.
So what to do?
There are whole schools out there using these approaches, usually facilitated by school autonomy. I hope that, with time, these examples will prove successful and will leak out into the wider system as teachers move jobs and schools look for new ways to improve.
I also think MiniLit and formative assessment are instructive. Perhaps proponents need to package some of their wholesale approaches into smaller units. That way, schools can adopt them as an ‘intervention’ and they can be tested by bodies such as the Education Endowment Foundation.
If you find this last suggestion distasteful, I would make two points. Firstly, more students will be exposed to these approaches, even if they are a little compromised. Secondly, this will happen whether proponents do the job or not. We have an example of this with the Education Endowment Foundation’s entirely useless trial of Core Knowledge; a trial that could never have worked.
Part of the reason for the failure of this trial is that Core Knowledge is opposed to the dominant ideology. What can be done about this?
I think there are several ways to approach conflicting ideologies. Firstly, we must be out there, explaining the ideas. That way, an alternative narrative is available to that of the education establishment. It is not about somehow converting dug-in advocates of whole language or problem based learning, it is about providing options for those who are less ideologically committed.
Secondly, teachers want to do a good job. If they try something and find it produces results, I think that for most teachers, this will trump ideological commitments. The trick is getting them to try it without subverting it and for long enough to start seeing the results. That’s why discrete, packaged, well-defined interventions might be a key part of the solution.