Simple, Complicated or Complex? Not all educational interventions are created equal...

In this article of 2019, the great Dylan Wiliam sets out his reasons why 'Teaching will never be a Research-based Profession'. 

"As soon as humans are part of the picture, things get a lot more complicated. And that is why I do not think that teaching will ever be a research-based profession...Teachers need to know about research, to be sure, so that they can make smarter decisions about where to invest their time, but teachers, and school leaders need to become critical consumers of research – using research evidence where it is available and relevant, but also recognising that there are many things teachers need to make decisions about where there is no research evidence..."

There is a lot to unpick here, but in this particular post, I will focus on Wiliam's challenge to teachers and school leaders to become critical consumers of research.

I will argue in this post that it is all about considering the categorisation before the implementation.

Complexity Theory tells us that the successful implementation of research in schools depends on first recognising into which of the following three categories the intervention falls:

1. Simple

2. Complicated, or 

3. Complex

As teachers we should ask ourselves: is the intervention in this research Simple, Complicated or Complex? The implementation framework then follows.

1. Simple Interventions

I should begin by noting that Simple does not mean 'easy' in this context. Rather, a Simple intervention is where,

The intervention is well-defined and uniform across students; the mechanism for change is clear and obvious; the outcome is predictable and measurable.

When the intervention is Simple, the teacher or school leader knows exactly what the new input entails. The input would be exactly the same across different students and across different settings - this is what is meant by 'well-defined' here. In brief, we could pick up the intervention, drop it into another school and it would look exactly the same for every student. 

Furthermore, the mechanism for change is implicit in the intervention: it is well understood how and why the intervention will work.

Finally, it is clear from the outset what kind of change is expected and it is clear how it could be measured. 

A Simple intervention is a neat package. It is analogous to a Newtonian model of experimentation: input, process, output follow in a linear pattern; the process is easily replicable, demonstrably causal and the truth of the underlying theory is universal and epistemological in nature.

In Education, the Simple intervention is a rare beast. 

A Simple intervention can be implemented via Logic Modelling, which I will expand on in a future blog post. Moreover, such a linear process could be replicable across different contexts and settings, as such it is fertile ground for RCTs and Action Research. In my experience, very few interventions can be implemented this neatly. 

2. Complicated Interventions

A Complicated Intervention can described as one where,

There are multiple inputs, requiring significant planning and training over a period of time; however, the mechanism for change is clear and implicit in the process and the desired output is clear and measurable.

In brief, there can still be a causal path through a Complicated process. The whole process is exactly the sum of its parts. However, the process is not easily replicable due to the multiple and varied inputs: not all teachers will implement the research in the same way every time. Nevertheless, we should be mindful that teachers aren't researchers: there may still be merit in implementing the intervention, even if the process is not uniform. 

In a complicated intervention, if one part of the process fails, there exists a mechanism to fix the problem, as a kind of clock-maker, as the whole process is the sum of its parts. What often makes an intervention Complicated is the more involved need for tracking, monitoring, feedback loops and training. 

3. Complex Interventions

A Complex intervention is one in which,

The input may change over time and is highly sensitive to initial conditions, the change is emergent in nature and the output is not necessarily predictable or measurable in an objective sense. The effect of the whole process becomes more than the sum of its parts.

I contend that a great deal of school improvement fits a Complex categorisation. I believe that this is because schools are a type of Complex Adaptive System (CAS).

Due to the non-linear nature of processes within a CAS, Logic Modelling is not appropriate and a more reflexive, non-linear approach must be employed throughout the stages of planning, implementation and assessing impact. These three stages should feed back into each other and inform future decisions in the cycle. The expected change is emergent in nature.

Schools are messy, learning is messy and the vast network of human interactions and inputs are very messy indeed. However, that is not to say that Complex Interventions are a lost cause. Wherever improvement is possible, educators have a moral imperative to make progress. This progress may be self-reflexive, involve feedback loops and continuous self-reflection leading to dynamic changes in direction and emphasis based on experience. The desired change may be a long time coming and it may be slippery to measure or define. 

That is still no reason not to try. 

As William says at the top of this piece, teachers are not researchers but "critical consumers" of research.

In future posts I would like to expand on how we should treat Complex programmes in schools. The implementation must fit the nature of the beast - if it is Complex, we must accept a long, slow process that yields results over time and needs regular review, collaboration and refining. There may even be false starts and "re boots"! And of course, we should be prepared to ditch the scheme if it isn't working. This is where Theory of Change can be so powerful.

References

Cunningham, R. Complexity theory and school improvement: some possible connections (2003)

Gleick, J., Chaos: the amazing science of the unpredictable (1988)

Greenhalgh, T., Papoutsi, C. Studying complexity in health services research: desperately seeking an overdue paradigm shift. BMC Med 16, 95 (2018) doi:10.1186/s12916-018-1089-4

Morrison, K., School Leadership and Complexity Theory (2002)

Obolensky, N., More for Less: the Complex Adaptive Leader (2019)

Stenhouse, L.  Research as a basis for teaching: Readings from the work of Lawrence Stenhouse (J. Ruddock & D. Hopkins Eds.). London, UK: Heinemann. Page 5. (1985)