Does DIRT work in maths?

I adore the idea of DIRT. Writing feedback that never gets looked at is absurd, and I love the “ethic of excellence” mentality. I’m struggling a bit with making it work for me right now, though.

This is my standard procedure.

I set and mark an exercise with a selection of problem types from the last 4 lessons.

I hand it back to the students. Their DIRT task is either to retry the question type they got wrong, or try my extension question in green pen.

I now have students doing individualised tasks right at the cusp of their understanding. Great.

Except.. how am I expecting them to successfully answer the questions? There’s been no additional explanation of the work to help them suddenly be able to answer this question that they couldn’t answer yesterday.

I’ve noticed a few students answering in green pen studiously and carefully; the only problem being what they’re writing is mathematically incorrect.

You can argue the comments I write are enough guidance for them, but I’ve found these fall into one of two traps

  • For my top sets, the work is complex and the misconceptions they have are hard to explain without a full paragraph of writing. Even writing a paragraph isn’t ideal. What I really want is to talk to them, read their body language, give them a chance to ask questions before I set them off to have another go.
  • For my lower sets, the average reading age is under 9. Teaching through the medium of writing comments in red pen is the most alienating thing I can do.

Plus the more I have to write, the more time-consuming doing DIRT is for me, and the less likely I am to do it frequently.

No one would argue the best way to teach new content is for kids to sit in silence reading your hastily scribbled comments. This seems even more crucial when you’re giving them work which you know they’re currently getting wrong, or alternately extension work which they’ve never seen before.

So, how to proceed? I’m going to experiment with prefacing DIRT work with a teacher-led explanation of the most common pitfalls, but I’d love to hear any other suggestions.

14 thoughts on “Does DIRT work in maths?

  1. Michael Tidd

    On reflection, I wonder if maybe it’s about what you target. We started focussing mainly on the Ma1 strands for target-setting as these were more relevant across the various units of work.
    Perhaps there is a difference between how we mark day-to-day maths, and how we provide feedback on aspects of application?

  2. Harry Fletcher-Wood

    This is part of an email the head of maths at my school (@dannytybrown) sent me a couple of weeks back of an email he’d sent his department based on his reading of some of the DIRT stuff that’s been published recently:

    I would like us to set a short ‘exit’ activity at some convenient point each week – it could be a small set of questions, or a hinge question, or a written task where they explain something, or… whatever. But it should be carefully selected to find out if they have learnt your basic objectives for that week (with perhaps a bit of challenge in there as extension).

    They will do this on paper, we collect it in and use it to formatively assess what they know. Our feedback could be as simple as a tick or cross, or it could be a couple of words about which bits they need to work on.

    So then the next lesson should contain DIRT, where pupils work on differentiated activities, based on what they did on the exit activity. It could be as simple as you saying ‘if you got Q1 wrong, you need to do these questions, and if you got Q2 wrong you need to do these…’, or… ‘If I wrote that you need to learn how to expand brackets, then you need to come and sit over here while I explain to you how to do it…’.

    While this is happening, all pupils who achieved all the objectives should be given something more challenging to develop their skills.

    Seems to offer at least a partial solution – cutting down on the amount of time spent writing for you and ensuring they actually get some guidance to improve?

  3. huntingenglish

    Our maths dept said their DIRT focused on an analytical analysis of test/exam feedback, with the DIRT then differentiated by their feedback. It sounds very much like the model Harry proposes above.

    I do think your point with a teacher led explanation to model the DIRT focus. In my last DIRT post I had this as the first two steps: Focus in the DIRT and then model and scaffold. The big problem with DIRT, and when it becomes patently ineffective, is when students are simply given feedback and expected to ‘fix stuff’. We need to reduce the cognitive load of the task by modelling more closely, based on the feedback, and providing scaffolds and issues. Doing a worked example/s based on common misconceptions, then giving students new problems to work on, would be the way to go for me. DIRT may be getting some attention on Twitter, but the principles are really fundamental to what we should be doing. Maybe the acronym has just given some clarity to something we intuitively know is sound pedagogy. It just needs closer guidance – with some more explicit instruction.

  4. srcav

    It is an interesting question you pose. I’ve written about our feedback tactics here:

    Your point about them being expected to do something they couldn’t without extra teaching is important. My classes approach it in groups, helping each other where they can with me circulating then we discuss it after, but I think the preface idea may be better. I wil. Try this on my return after half term. Thanks for sharing.

  5. missdcox

    Could homework be differentiated based on what they do/don’t understand? And homework could be to watch a video on how to do the bits they couldn’t and re-try the questions?

    Does DIRT have to be in lesson?

    I think the point that is coming out from this, that I’m beginning to find out, is that DIRT may mean differentiating 30 times in a class as they may not all be doing the same thing and so your input will be wide and varied!

    Also, could DIRT for some be a reflection time? Rather than ‘improving’ could they write a paragraph about how they feel about this work? What don’t they understand? Did they understand some parts and not others?

    Final thought, once you’ve marked you could group students on tables with similar DIRT work. They can then work together and you can intervene on a table rather than individual basis. I personally avoid putting someone that ‘can’ with someone that ‘can’t’ unless I feel they will both benefit.

  6. Matthew Fairbairn

    I find that when I do marking of a short quiz, there are maybe 4 or 5 common mistakes made by almost every student. So I then just go through those mistakes with the whole class (many students will have made 2 or more) and then set them differentiated questions to work on. The students who didn’t make the common mistakes I can quickly talk to. The issue is that then students might be listening to me talking about something they already know and didn’t get wrong, but I’ll accept that in order to get some quality DIRT going.

    The other option that works is to split the class into groups depending on what they got wrong. Some will have me talking to them, some will watch a video to explain, and if available some will be with a TA. Last year with a year 11 group that were seriously underachieving at the beginning of the year this worked brilliantly. Every student felt that they were working on their weak areas and they made great progress.

  7. mrbenney

    I am very interested to read on how dirt can be effective in maths. One simple suggestion would be (and this wouldn’t work on all topics) to mark work paying particular attention to the workings out stages. Errors can be highlighted with a green (or red pen). Depending on the pupil, feedback could be written on where they are going wrong. Other pupils would just need the error highlighted and instruction do redo the question(s). Pupils that made no errors could be “maths leaders” and circulate the room offering assistance to other pupils. Alternatively, these pupils could be asked to reflect on the procedure they followed to gt right answers. They could write it out in bullet point or use connectives to write a short paragraph. This would be a good use of literacy in maths.
    All of this is very simple and obvious but I suppose that is the idea.

  8. @mrnickhart

    One option is to plan for some modelling for different groups and some follow up questions for them to practise. The difficulty is that some groups will have to wait for ther input from the teacher, or complete their work before other groups have been seen. Combining DIRT with practising the most valuable stuff could work. While you talk to one group, the others will work on some times tables recall practice, or a paired calculation challenge that is sufficiently open that they can continue to work while you intervene with all groups. Mine have times tables flash cards that they can practise with alone. Also alone, they can use digit cards to make an addition calculation, subtract another number from the answer, trying to get as close as possible to a target number.

    DIRT + deliberate practise of the most valuable knowledge / skills could work well together.

  9. Max Ray (@maxmathforum)

    Maybe, a la, you could share some of the work you’re giving feedback on for your tops and your lowest class, and we could brainstorm short comments that would help the kids re-engage and think further, without writing epic paragraphs? In theory, DIRT time can be used for kids to figure out what went wrong on their own, depending on the topic and questions. Or to introduce a new approach through a new way of visualizing or approaching the topic, but it’s hard to give examples out of a specific math context.

  10. David Thomas

    We’ve been doing DIRT (under a different name) as a school focus for about a year now. To begin with I had exactly your problem – how will they improve without additional teaching? The solution I’ve used most is to categorise incorrect responses to a question and use those to set appropriate work at the start of the next lesson. Since reading Harry’s marking post last week though I’ve been thinking and about RAGing student answers: red takes them to a support sheet that’s scaffolded, amber is an instruction to find and correct an error they are capable of locating, and green is a question that stretches by mixing in an additional concept. Will see if this works!

  11. jonpatrick

    I’m not sure what the etiquette of citing from blog articles is, but I thought that you might be interested that this article appears in my reference list for RJA2. Great stuff.

  12. Pingback: 8. October 2013 #blogsync: Marking with Impact! | EDUTRONIC | #blogsync

Go forth and opine