Most of my teacher-training took place in an independent school. This is not uncommon for Classicists, because our subject is not taught in enough state schools to make placement in the state sector possible for everyone. While I spent my first short placement in an absolutely delightful comprehensive called the Herts & Essex High School, the bulk of my training took place at Brentwood School, a large independent school in Essex. So, by the time I was at the stage of leading full classes like a proper teacher, I was in a school in which the largest class size was, from memory, around 20. This, as any state school teacher (or indeed pupil) will tell you, is not normal.
My earliest lessons in my first job, at a state grammar school in North London, felt like walking into Wembley stadium. I simply could not believe how enormous a class of 31 felt after the classes I had experienced at Brentwood. It took a considerable amount of getting used to for me. But what about for my students? Would they have been better off at Brentwood?
Class size has long been a topic of debate in education systems worldwide, particularly in the UK and the US. Historically, the assumption has always been that smaller class sizes are associated with better student outcomes. Many people both claim and assume that reducing class sizes leads to significant improvements in educational attainment, particularly in the early years of schooling. On instinct, I find it highly likely that small classes could be hugely impactful in the early years, when personalised learning can make such a massive difference to a very young child’s development. But do small(er) class sizes really lead to better outcomes at secondary level? In all honesty, the evidence is surprisingly unconvincing.
A meta-analysis conducted by John Hattie, while it did find a very small positive effect of reducing class size, the impact was so small that it was written off as negligible. The effect size for reducing a class from 25 to 15 was typically around 0.1-0.2, well below the standard for statistical significance. The problem, according to Hattie, is partly that teachers do not significantly adapt their teaching to reap the benefits of smaller class sizes. Hattie is Director of the Melbourne Education Research Institute at the University of Melbourne and is an expert on performance indicators. He has dedicated a lifetime of research to analysing the impact of differing measures on student attainment, including class size, and even he can’t prove that it matters. Hattie’s work, it is fair to say, has met with a good deal of criticism from other scholars, but then it wouldn’t be a significant piece of work if it had not done so. It seems to me that lots of people are so invested in the belief that size does matter, they simply cannot accept the results when they indicate otherwise.
The Education Endowment Foundation, which is dedicated to improving educational outcomes for low-income students, found that reducing class sizes has low impact for very high cost; they argue that the case for reducing class sizes in state schools is based on very little evidence. The same was found by The Sutton Trust, the UK’s leading social mobility charity. They found that reducing class size is not an effective way to improve school results and while smaller classes can have a small positive impact (around +2 months on average), the benefits are not significant until class-sizes are reduced to below 15-20; this kind of reduction is never going to happen in the state sector, as the cost would be simply astronomical.
So, reduced class size does not seem to be convincingly linked to improved outcomes for students. But what about improved wellbeing for the staff? There is a reason why teachers get so understandably upset when someone says that size doesn’t matter, as they did on Twitter this week when behaviour advisor and co-founder of ResearchEd Tom Bennet shared this blog by a US educationalist (who has since deleted his Twitter account!) saying “This upsets a lot of people but it’s pretty clear: class sizes don’t make the difference people think. They have to be massively reduced in order to have a real impact, and to cut class sizes even by half you’d (obviously) need to employ twice the number of teachers. That’s simply not going to happen. So let’s focus on better, achievable improvements, like AFL, direct instruction, behaviour curriculums.” The thing is, he’s absolutely right! But the reason teachers get so upset about it is that they struggle to separate the issue of their own workload from the research into student outcomes. In reality, these are two separate issues, but I would argue that they are of equal importance. I don’t think we should write off the idea of small (and thus potentially affordable) reductions in class-size if we can argue convincingly that this would make a tangible difference to teachers.
With the UK facing an escalating recruitment and retention crisis and teacher vacancy rates at a record high, teacher workload is perhaps the most important problem we face in education today, yet the government does not want to talk about it. Reduced class sizes without question lead to a reduction in workload for teaching staff. Just off the top of my head, each reduced class means fewer books to mark, fewer assessments to grade, fewer appointments at Parents’ Evenings, fewer EHCPs to cope with, fewer individual relationships to build, fewer parents sending emails. This is not insignificant and we do need to examine whether even a small reduction in class size would make enough of a difference to teacher workload (and thus potentially improved retention) to balance the cost.
There is no question that making dramatic reductions to class sizes in the UK – reductions to the extent that the marginal gains for students uncovered by Hattie’s research would be take place – that’s never going to happen. There’s no point in even talking about halving the size of classes and hence doubling the number of teaching staff. Yet how about we agree that capping classes at – for example – 24 might be worth exploring? Is it completely beyond our wit to do a cost analysis on that? Some people seem to believe that, unless the reductions are hugely significant, they won’t make a difference to teachers. I suspect that these are people who have never had a few extra bodies in their class. Have they never been in a situation where a timetabling glitch means that their class has swelled to 33 or 34? I have. The difference is tangible and undeniable. It is also hugely significant when the numbers go the other way and your class peaks at 24 or 25 instead of the usual expected standard of 31. Small reductions could have tangible gains for teacher wellbeing and I for one think that matters enough to at least explore the possibility.
