The Problem with AfL

September 1, 2013

I was interested to read some recent blogs debating the merits of Assessment for Learning (AfL), particularly here , here and here. While I’m sure I risk repeating some of the points already made, I can’t resist weighing in with a quick summary of my own views on AfL.

The basic idea of AfL, or formative assessment, is sound. We should use assessment to find out what students don’t know and then teach them it. We should provide useful feedback. But what has happened in practice, is probably the worst example of a sensible idea becoming a checklist to be mindlessly followed since the three-part lesson. Here are the problems with AfL:

1) AfL techniques are used regardless of whether there will be any opportunity to use the result. Formative assessment hinges on finding out useful information and then taking action. It is not useful for its own sake. There is no point checking understanding or learning when you don’t have time to do anything about those who don’t understand or haven’t learnt. There is no point finding out what students can’t do if you have no plan to do anything about it. There is no point providing feedback that nobody will ever act on. If you can’t alter your teaching because of what you have found out, there is no point finding it out. When planning to use AfL, you should always start with what you will do with the information you gain from it, otherwise you are simply ticking boxes with no benefit to the students.

2) AfL often relies on self-assessment and peer assessment. We are the worst judges of how good we are at something. We are also very poor at judging the performance of others at carrying out activities that we are not good at ourselves. These inherent biases in human beings are, in psychology, known as aspects of the Dunning-Kruger effect. Please take time to look into this before using either self-assessment or peer assessment. Currently, these seem to be emphasised heavily in AfL.

3) AfL initiatives often ignores teacher experience and expertise. If you have taught the same topic dozens of times to dozens of classes (and assessed the learning) then you start to get rather good at teaching it. You start to be able to tell which classes or students will grasp it in a second and which will need constant checking and extra explanation. If you are an experienced teacher and you know a class well you can often tell if students have learnt something from the expressions on their faces, how quickly they start the work or just their level of attentiveness during the explanation. You may also have a good feel, particularly with students you know well, for which students would tell you if they didn’t understand and would ask questions if they needed to. Additionally, experienced teachers will have a variety of techniques for unobtrusively monitoring students and “assessing” them without anybody noticing. However, too much AfL is about making a big show of the process of assessing. I have been criticised in observations for not using AfL in classes with less than ten students in where I looked at every student’s  work and spoke to every one of them them about how they were getting on and, where required, how to improve. It is possible for teachers to use the principle of AfL, i.e. knowing how students are getting on and responding to it effectively, without any sign of this being clear to the average observer.

4)  AfL has become about gimmicks. This follows on from the last point. AfL, particularly when looked for in observations, has ceased to be about the general principle of whether teachers know how classes are doing and act on it and has become about a narrow range of techniques. Some of these are good techniques (marking or questioning); some are bad (“no hands up” policies or picking students at random to answer questions). However, the techniques of AfL have become more important than the principles. The most obvious example of this is the cult of the mini-whiteboard. Now, there really is nothing wrong with getting students to hold up answers on mini-whiteboards. However, it has now reached the point where teachers are encouraged to feel that they have used AfL if, and only if, they have used mini-whiteboards. Techniques where all the students in a class answer verbal questions and share their answers are a good thing, and I don’t want to attack that, but mini-whiteboards are not always the most efficient way to do this. Some classes cannot be trusted with them; some classes can communicate the relevant information much faster and more accurately in other ways. At their worst, mini-whiteboard cultists will have students using mini-whiteboards for doing long and detailed work that would be done to a much higher quality and be easier to mark if done in books.

Update 24/9/2013: A fifth point has now been added here.

Now, given that so many fads in education are complete rubbish from start to finish, AfL is far from being the worst idea promoted in schools. The basic concept is sound, although sometimes it can be stating the obvious, and teachers should think about whether they are using formative assessment. However, the techniques promoted by AfL enthusiasts are often ineffective and used only to tick boxes. Any set of techniques, no matter how important the principles behind them, will become toxic once we have an army of idiots promoting and enforcing them. I will use formative assessment in almost every lesson I teach (and when I don’t it will only because there is no time to act on it) but I will still dread having “AfL” as a target in an observation, because it will inevitably mean spending half an hour finding a set of mini-whiteboards with a set of working pens.


  1. Reblogged this on The Echo Chamber.

  2. Now if I could just ask you to colour in this traffic light to let me know if you have understood what we did today …..

    Fully agree with this post!!

  3. One of the main problems, I think (and forgive me if you’ve made this point already) is the way in which techniques are generalised. A particular ‘gimmick’ (to use your word’) is perceived to be useful in a particular situation, and then rolled out as The Answer To Everything.

    For instance, I’ve been using mini-whiteboards for years, and didn’t even realise that they were part of AfL (I go to sleep in the sort of INSET sessions where these things are initiated in our school, and still haven’t really worked out what AfL is all about), and find them invaluable for the sort of questions which are (a) simple – I don’t mean ‘easy’ but ‘not complex’ – and (b) have a clear right/wrong answer. In my own subject, A-level English Language, I’ll use them for checking that the students have mastered a bit of grammar or phonology I’ve been teaching. But for 90% of the stuff they have to understand, whiteboards would be pointless.

    I’ve spent half an hour clicking between your post and others’ on AfL, to try and compensate for all those hours of gentle snoring, and it looks as though the whole thing is like that: techniques which work well in some contexts being taken as magic bullets to destroy bad teaching. As you point out, experienced teachers will assess performance and give appropriate feedback anyway. So the big question is – why have the last 30 years or so been so devoted to all these ‘initiatives’ and new techniques (at least, they’re marketed as ‘new’), while the knowledge our students have acquired has actually decreased? (I’ve run a national general knowledge competition for schools since 1985, and, apart from science, where pupils do seem to know more than they did in the 1980s, have had to lower the ‘difficulty level’ of questions quite substantially, particularly in questions on history, geography and literature.)

  4. I know Hattie endorses peer assessment from research. However, I have only found it useful in a few contexts, such as to familiarise students with mark scheme requirements or to help them spot typical mistakes I have already explained from the front. I have always wondered why it comes out so well in research. Is it because otherwise inexperienced teachers might not get pupils to reflect on how they could do better?

  5. I concur entirely…. my school is on an AfL mission at the mo.

    I particularly merit your point about not doing it, if it cannot be acted upon.

    This is especially true in subjects that tackle their teaching in a topic by topic basis.

    I think peer/self assessments are really important although its not a reliable a ‘measurement’ as teacher assessment I do think kids benefit from having to make judgments about the work of themselves and others.

  6. A great post and as always great comments.

    The only thing I would add is that I think it is wrong to assume that the “peer” or the “self” in the process has to be any sort of expert.

    For instance if I am trying to test recollection of an item, I can provide the answer to the peer or even the self by means of flashcards for instance. The student will very quickly realise that they didn’t know the answer.

  7. Totally agree with the point about gimmicks. I put it down to the fact that the importance of subject pedagogy is hardly recognised so every method becomes a generic magic pill.

    However I think self assessment is a very important life long skill both for further study and work. As you say we are not naturally very good at it so it needs to be taught. ‘Inside the black box’ asserts that peer assessment is an important stepping stone to self assessment as we see others’ mistakes more easily than our own.
    I would agree that peer assessment does not seem to be regarded in that light but just another ‘activity’. If exams are to become more the norm then we must ensure that children are able to perform the checking process in a rigorous manner which after all is just self assessment by another name.

  8. I agree with most of this but I think you need to combine two ideas here. One of your points is that AfL techniques used out of context just don’t work (agreed). Another point is that peer and self-assessment are prone to bias (agreed) and therefore shouldn’t be used (disagree).

    Peer and Self-assessment shouldn’t be used to actually assess anything! You are right that the results aren’t accurate enough to be relied upon. Peer and Self-assessment should be used because they provide a model for students to understand what constitutes a good answer from a weak one. It’s a bit like in Educating Rita when she gives the answer, “Do it on the radio”. Her answer is correct but it is not a good answer.

    It’s a great technique to use to help students give exam-ready answers. Though I know this concept is, in itself, contentious for many.

    As with any technique, understanding WHY it is used and WHEN is the most appropriate means you can use it wisely. Using a technique inappropriately will usually be a waste of time.

  9. I’m slightly worried that AfL is still subject to these misconceptions and problems – we were going all-out to promote AfL in our schools 10 years ago. How can there still be teachers (and even whole schools) who don’t understand these fairly simple principles?

    It’s not all that often I agree with Old Andrew, but this is spot on, and sums up most of the problems with faddy approaches. It’s a very simple concept, AfL – it’s about assessing what the kids have learned so far and using this to determine what and how you are going to teach them next. Have they got it and are they ready to move on? Do they need more time to embed it? Or do they need to start over with a different approach?

    It’s what good teachers do instinctively, but when you’re chasing the league tables it’s all too easy to forget why you’re marking pupils’ work. But as they say, weighing the pig never made it fat, and unless you’re going to do something useful with the assessments then there isn’t a lot of point in carrying them out.

  10. […] I was interested to read some recent blogs debating the merits of Assessment for Learning (AfL), particularly here , here and here. While I’m sure I risk repeating some of the points already made, …  […]

  11. […] this month I wrote a blogpost The Problem with AfL which explained that while I am in perfect agreement with the principle of formative assessment, […]

  12. […] and was not proper formative assessment (ie assessment that informs practice). He recommended Old Andrew’s blog on AfL which summarised how a sensible idea became “a checklist to be mindlessly […]

  13. I share these sentiments entirely. The basic principle is sound I agree but somewhere an Animal Farm clause has crept in.. we do afl ‘in order to improve’. This has further become translated into ‘in order to reach the next level’. Consequently we get shallow execution of performance indicators – which are now circus tricks and no longer serious indicators of in depth learning.
    In Drama the whole process is an afl process and you hear and see students engage in this automatically. (‘No its better if come in later and say nothing – just look…like this. Yes then the audience can fill in the words…can we have a light change at that moment because that will connect back to the moment when… which means that… ‘ They are assessing the work and the role they play within it – not themselves. However we see this good individual pursuit of connoisseurship through practical engagement (which is what we want isn’t it?) being discarded in favour of traffic lights and lolly sticks, thumbs up and down and a whole range of fads and fetishes. By the time we have executed the ‘regime’ there is not much time left for actual learning.

  14. […] or whatever you want to call it, has its problems. Andrew Old has a great big list of ‘em right here. It seems it’s usually a problem of schools steamrolling out the policy as the Next Big Thing […]

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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

%d bloggers like this: