Back in 2015, I wrote a series of three blog posts arguing that we need to change the way we train teachers. In Part II, I wrote about the research often carried out in faculties of education and I suggested that, “affectations around complicated-sounding theories such as post-structuralism provide something of a memetic peacock’s tail.” In Part III, I responded to criticisms of the previous two posts. Doubling-down on my comments about education research, I jokingly wondered whether it might be worth, “producing a catalogue of some of the silliest, most abstruse and most nakedly political education research,” in a manner similar to the ig noble prize for science research. I illustrated this post with a cutting from the programme for an upcoming Australian Association for Research in Education (AARE) conference which listed a presentation by Lucinda McKnight of Deakin University titled, “Meet the Phallic Teacher: Performing Curriculum Design And Identity In A Neoliberal Imaginary.” The title had struck me as a little absurd.
Nearly a year later, I was surprised to be pointed to a post by McKnight on the site of the Gender and Education Association that referred to my blog. Following an introduction by Jessica Ringrose, Professor of Sociology of Gender and Education at the UCL Institute of Education in London, McKnight dubs me a ‘phallic blogger’ in a quite extraordinary piece that I encourage you to read if you have the time. Given the level of over-analysis that my original use of the cutting had attracted, I decided it might be worth avoiding commenting on McKnight’s output in future.
However, I have now broken my silence because McKnight and Andy Morgan, a medic and lecturer at Monash University, have written a piece for the AARE’s blog site that illustrates a fundamental problem we encounter when we try to make teaching more of a profession.
The article starts reasonably well by pointing out some of the well-known problems with conducting randomised controlled trials (RCTs) in education. These are the types of studies where participants are randomly assigned to one of two (or more) conditions, with the results compared at the end. Often, the two conditions used are business-as-usual versus some innovative teaching technique. In such studies, you cannot ‘blind’ participants and researchers in the same way that you can in medical randomised controlled trials – students will know if they are getting business-as-usual or something different whereas patients cannot tell whether the pill they have swallowed contains the active ingredient or is just a sugar pill placebo. This can lead to expectation effects where students perform better in the innovative condition because it is new and interesting or because they expect to do better and so try harder.
Such problems are not wholly insurmountable. I run small randomised controlled trials as part of my PhD research and they are designed so that both conditions are novel to the participants and so they cannot tell which one is predicted to be superior. Such small-scale studies are a motif of my field of research – Cognitive Load Theory. However, if we want to run large studies that compare whole programmes then another solution is to run two innovative approaches against business-as-usual in a three-armed study. I have encouraged the UK’s Education Endowment Foundation to do this.
McKnight and Morgan, however, use their criticisms of randomised controlled trials to launch an attack on the entire concept of evidence-based education. They end by calling for, “an urgent halt to the imposition of ‘evidence-based’ education on Australian teachers, until there a fuller understanding of the benefits and costs of narrow, statistical evidence-based practice.”
To a layperson, this may seem like a bizarre thing to call for. Who is against using evidence? However, once we understand the context, we are afforded a glimpse into the way that some education academics view the world.
McKnight and Morgan overreach from a claim about randomised controlled trials to a generalisation about all evidence-based education. Randomised controlled trials are not the only source of scientific evidence used either in medicine or education. In the 1960s, a whole series of process-product education studies were run in which researchers sat in the classrooms of different teachers, making observations. These observations were then correlated to the gains students made in these classes and the practices of more effective teachers were identified. Rosenshine’s Principles of Instruction, is a good summary of what the researchers found. These studies were not randomised controlled trials. It is possible that some of the behaviours the more effective teachers demonstrated were not linked to their students’ performance – correlation does not equal causation. However, the fact that the findings were replicated by many researchers across diverse settings is highly suggestive and there is some evidence that attempts to teach these behaviours to other teachers makes them more effective too. In essence, this is equivalent to the medical field of epidemiology. Nobody would run a randomised controlled trial to show that smoking causes lung cancer in humans, but the correlational evidence is overwhelming.
McKnight and Morgan unwittingly acknowledge their overreach in a section on ‘gurus’. Although they have outlined concerns about randomised controlled trials, they conflate that with the effect sizes used by John Hattie. Hattie generates effect sizes from a whole range of studies with varying designs, which is one reason I have been critical of Hattie’s approach. I am not sure of the exact proportion, but the number of randomised controlled trials sitting behind Hattie’s figures is likely to be relatively small. Is Hattie’s data ‘scientific’? I’m not entirely convinced.
Another argument McKnight and Morgan use is that, “Evidence based medicine is about populations, not people.” I think McKnight and Morgan’s point is that, although an approach might work on average, there is no guarantee that it will work in any individual case. This is true, but it represents a pretty good starting point. Would you rather not know what tends to work in the majority of cases? This is an appeal to individualism typical of the progressivist philosophy of education.
The oddest argument in McKnight and Morgan’s piece is perhaps when they arrive at the ‘phallic teacher’ concept and it is worth quoting in full:
“Teaching is a feminised profession, with a much lower status than medicine. It is easy for science to exert a masculinist authority over teachers, who are required to be ever more scientific to seem professional. They are called on to be phallic teachers, using data, tools, tests, rubrics, standards, benchmarks, probes and scientific trials, rather than “soft” skills of listening, empathising, reflecting and sharing.”
This seems to be an argument that science, scientific trials and quantitative evidence are somehow masculine concepts that do not sit well within a ‘feminised profession’. Really? I don’t think science belongs to any gender and I can imagine a lot of people being offended by what appears to be the gratuitous use of an anachronistic gender stereotype.
Possibly the most vocal proponents of evidence-based education are those who argue for the central role of phonics in the teaching of early reading and many of these advocates are strong, intelligent women with a deep commitment to applying the best available evidence. Although short on evidence from randomised controlled trials, the totality of the evidence for phonics is overwhelming and supported by three government inquiries in the US, the UK and Australia. And yet a systematic approach to teaching phonics is clearly lacking in many classrooms. Why is this? Because many of those who are involved in educating a new generation of teachers buy into the idea that evidence-based education is a bad idea and that science cannot and should not be applied to teaching. The tragic consequence of this is those instructional casualties who could have been taught how to read but who have been failed by the system.
Aided by social media, teachers are now talking directly to each other and have started the process of taking control of their profession. It is no longer acceptable to present us with a false choice between being for or against evidence-based teaching. Clearly, not all evidence is equal. Some randomised controlled trials are better than others. Some correlational studies are better than others. As teachers, it is time to build the expertise to evaluate these claims ourselves.
We don’t need education academics acting as gatekeepers and telling us that the use of scientific evidence is ‘phallic’ and therefore bad.