Is it blasphemous to doubt the efficacy of AfL? While purists might argue that it’s ‘just good teaching’, we teach in a world where formative assessment has become dogma and where feedback is king. (Don’t worry, I’m not about to start upsetting the feedback applecart although there are occasions when pupils can benefit from it being reduced.) But AfL as a ‘thing’? I’m not just talking about some of the nonsense that gets spouted about lolly sticks and traffic lights, I’m questioning the entire edifice. Is assessment for learning really all it’s cracked up to be, or is it just me?
You see, here’s my problem: if we accept that learning is not the same thing as performance, then how can we possibly think that any AfL strategies are measuring anything except performance? For those that can’t be bothered to follow the link I’ll summarise thusly: learning is much too complex to be directly observed, it can only be inferred from pupils’ performance. And by learning here I mean the ability to retain information in long term memory and transfer to new domains.
Just in case all that’s a little opaque, let’s break down what happens in lessons which follow the principles of AfL:

  1. Teacher sets learning objective & shares success criteria
  2. Teacher introduces task designed to test whether pupils have understood and are able to meet the learning objective and success criteria
  3. Pupils attempt task
  4. Teacher assesses pupils (maybe using self or peer assessment) against the success criteria
  5. Teacher uses data from assessment to adjust teaching next lesson.

Have I got that about right? I must have taught about a billion lessons which have pretty much followed this pattern. And in every single one of them I have laboured under several misapprehensions. Maybe you have too? These are the pits into which we often fall:

  • The Input/Output myth: what a teacher teachers, pupils learn. This leads to…
  • Mistaking performance for learning – just because pupils can do something this lesson, does not mean they will be able to do it next lesson
  • Prioritising understanding above remembering – of course understanding is of a higher order than remembering, but it’s a devilishly slippery concept. It doesn’t matter what pupils understand if they can’t remember it. Ever had the experience of patting yourself on the back after a thrillingly brilliant lesson in which everyone understood everything and yet no one seems able to recall any of it next lesson? Yeah, me too.
  • Lack of practice. We assume, wrongly, that if our pupils can do a thing once they have learned it. Well, maybe. But they certainly haven’t mastered it. AfL is predicated on the idea that if pupils seems to got it, we can move on to something else. This is a deeply unhelpful way to approach teaching.

So this is the real problem: if we adjust our teaching based on the outcome of students performance in the lesson rather than on what they’ve learned, where does that get us? Where I think it gets us is into a hopelessly confused mess where we’re constantly adjusting our lesson plans based on information which is at best inaccurate and at worst, utterly wrong. Doesn’t this just result in us chasing the tail of pupil progress like a sackful of demented Alsatian puppies?
This post is really about people ‘doing’ AfL badly. Of course, none of this is to say that we shouldn’t use formative assessment, but let’s open the debate and be a little more thoughtful about why we’re using it and what we hope to accomplish. Let’s drain off the filthy bathwater and see if there’s a pedagogical baby worth salvaging. If we accept that AfL is just making sure kids know how to improve, then fine; that’s just good teaching. But if we’re seduced by the AfL industry pedalled by sharply suited Ofsted whisperers, then hopefully this will help pluck the scales from your eyes.
And if you’d like to take issue with any of this and feel compelled to point our the error of my ways, please do, but I’d rather you didn’t base your objections on that fact that I obviously haven’t understood what AfL really is. I have: click here for a sackful of posts proving this fact. That said,  I’d welcome some feedback on this – haha!