The great education and public health experiment

We are in the middle of a rather odd educational experiment. I said odd, because normally we run experiments in the hope of learning something new. But so far in this educational experiment, where children have received a wide variety of different distance learning ‘treatments’, we’ve done little to systematically find out how they are doing so that we might make better decisions in the future.

Sure, individual teachers have hunches that they’ve identified an effective approach, whether it is ‘live’ instruction, setting and collecting work on google classroom or emailing worksheets once a week. But the impossibility of directly observing learning has been amplified by not being able to see students, so hunches have to remain as just that.

Schools are now being asked to enter an extremely high stakes experiment, where we risk long-term health and economic outcomes as well as cognitive development.

I am a pessimist and so I think this risky experiment where we try to get children into schools in different sorts of arrangements is going to last well into the next academic year. The combination of social distancing requirements yielding teacher and space shortages, shielding and sick families, shielding and sick teachers and the need to contain the movement of children around schools means we need to learn as much as we can about how to manage this disrupted education system over the long-term. And fast.

Many politicians, lobbyists and journalists have started playing the ‘evidence card’ to justify their position on what should happen next. But I’d contend that we cannot currently use scientific evidence to make moral judgements about optimal school re-openings because the evidence we’d ideally like simply doesn’t exist.

For some this means that we should therefore wait and watch what happens in other countries. I have sympathy with this position, which employs the precautionary principle with respect to teachers health, but it isn’t cost-less to families who will lose their jobs as a consequence. Moreover, I’m sceptical that we will learn much from other countries as they take their tentative first steps simply because we will not be able to separate the causal effects of their school re-opening decisions from everything else that is going on in their country.

Others (including some SAGE committee members) say we should get on with tentatively re-opening, even if the move is reversed, because it will help us learn how to re-start school openings in the future. However, what exactly will we learn from the universal re-opening of schools following the DfE’s approach? If infection rates rise in late June, what does this tell us about how to adjust the way we try to re-open schools again in September?

My very controversial suggestion is that, if we genuinely want to proceed on the basis of evidence, we should treat this as a proper experiment and ask schools/towns to be randomised into different re-opening arrangements. It feels more ethical to do this than to implement a reasonably arbitrary DfE proposal, given that the things we will learn early on can potentially save lives, save jobs and improve student learning later in the academic year.

We should, of course, only do this once we have the suitable scientific tools in place to facilitate appropriate data collection. We need to learn fast, so we should be collecting as close to ‘live’ data as we can from schools on:

  • Incidence of COVID-19 amongst staff, pupils, their families and their communities, using the mythical test, track and trace system that may one day materialise.
  • Surveys of teacher and parent economic, psychological and social wellbeing, to learn how school arrangements facilitate or impede home and work life. (As someone who runs surveys I would say this!)
  • Regular learning input measures (e.g. time on task) and attainment data on the most critical areas of the curriculum, especially  reading, writing and maths.

I cannot tell you what the treatment groups should look like, but perhaps someone who isn’t trying to hold down a job whilst looking after their children will have time to work that out! For me, some of the areas we need to learn about are:

  • Rotas: If they have to be used (and I still don’t see how we can avoid them), which arrangement minimises COVID-19 infections and maximises learning and home life? 1 week on, 1 week off? Half days for all? 4 days on, 10 days off shifts that run through weekends? What are the gains to keeping sibling groups intact in terms of home life, versus the costs to school organisation?
  • Class sizes: How much can we reduce the COVID-19 infection rate by running classes of 15 rather than 30, and does this trade-off depend on the age of the children involved?
  • Replacement teachers: If we have smaller class sizes and we shield teachers who live in high-risk households, which untrained teachers should instruct classes? Are recent graduates with minimal training more or less effective than teaching assistants?
  • Enhanced homework arrangements: If children are spending less time at school, what is the most effective homework to set them? Centralised online learning platforms such as Oak or Hegarty? CGP-style workbooks? Project-based tasks?
  • Distance learning for shielding families: What are the costs and benefits of providing some live instruction, or one-to-one tutoring, or feedback on work, or motivational phone calls, versus a basic model of posting weekly resources from the large scale, centralised providers?
  • Student movements: What are the health gains to forcing students in Years 7 to 9 to take all subjects with their form class in the same classroom (whether streamed or mixed ability), relative to the educational disruption? Should more than one class be allowed out on the playground at the same time? Should children be able to eat a packed lunch or a cooked dinner at school?

I wouldn’t like to pretend that running large scale, high stakes and dynamic experiments is easy, particularly because even ‘live’ data is retrospective. It tells us about infections last week. It tells us about learning that happened (or failed to happen) weeks or months ago. But I cannot see another way to gather high quality evidence at speed about how we trade risks around health, learning loss and jobs.

I hope everything I have written here is irrelevant because all pupils and all teachers are able to return to full-time schooling on the 1st September. But if we are to take the possibility that education is disrupted beyond September seriously, then we owe it to teachers and children to learn as quickly as possible how to make it work.

One thought on “The great education and public health experiment

  1. Pingback: Let's talk about feelings (but still with data).... - Teacher Tapp

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s