“Pupils are responding to teacher feedback to improve their
work with evidence of pupil progress.” I was the stereotypical new teacher
spending evenings marking stacks of books and class time was then spent rewriting
answers in a different coloured pen. These comments made it seem worthwhile
though, pupils were improving thanks to my marking and DIRT.
There was only one problem. Pupils were not getting any better.
Economist Charles Godhart created Godhart’s law: ‘when a measure becomes a target, it ceases to be a good measure’ and this was my issue with DIRT. It seems entirely sensible. Pupils will ignore feedback unless they actively act on it and so time to ‘reflect’ on feedback will help and I am not saying DIRT has no place in marking policies.
Yet when it becomes a target with evidence needed of
feedback, I found I was focussing on how to show pupils were responding to
feedback to an external observer rather than what made pupils improve.
Mark Enser talks a lot about the idea of teaching like nobody
is watching, focussing on what others expect rather than what benefits our
pupils and it helped me realise the 3 ways I was screwing up feedback.
1. Feedback was given too late
I see most of my classes twice a week so typically pupils
would write an essay on a topic and then the next week I would have marked it
and given feedback. One of the earliest examples was when I noticed pupils were
getting confused between a movement along and a shift in a demand curve and
were using the terms interchangeably.
The problem was that by this point they had completed enough
practice on this before errors had been corrected. Paul Kirschner and Carl Hendrick
talk about practice not making perfect but making permanent and this is what
was happening, the errors were becoming embedded. Rosenshine explains the
importance of ensuring a high success rate before practice to avoid this
problem.
Now what I would do is ensure these misconceptions have been
identified and corrected immediately. This is done through questioning and
utilising all class responses (what Doug Lemov refers to as ‘show me.’) By giving
feedback and responding to errors before practice, pupils can complete
practice with misconceptions tackled at source.
None of this appears in their books and there is no DIRT, but
it has made my feedback infinitely more effective and led to much faster pupil’s
progress.
2. Feedback is a poor way of improving knowledge
I was sat on a Saturday constantly writing the same comment
again and again ‘explain the impact of monopoly power on consumer surplus.’
Then the next lesson pupils rewrote their paragraph with this helpful nugget of
information to ‘explain the impact of your points.’
The problem was that they did not know the impact a monopoly
has on consumer surplus and so being told to explain it wasn’t helpful and their
redrafts weren’t much better.
A lot of errors I spot are down to a lack of knowledge. This
is partly ameliorated by point 1 but there will always be some misconceptions. My
favourite feedback I gave was after reviewing pupils’ answers on the impact of
an indirect tax and discovering they did not fully understand why it was less effective
for inelastic goods.
Kirschner and Hendrick talk about feedback being ineffective
to teach knowledge gaps as what is needed is instruction. I retaught this
material, checked for understanding extensively and a few weeks later we did a
retrieval quiz on the topic. There was no need to redraft the essay as I was
trying to ‘improve the pupil, not the work’ and essays in the future were
improved due to a stronger knowledge of the topic.
There was nothing in their books to indicate DIRT but it was
the best and simplest feedback I’ve ever given.
3. Focussing feedback on pupils and not on me
A common comment I would write is ‘you have not used the
extract enough’ when marking pupils’ essays. The reason they were not was that
I had not modelled it carefully enough, so they didn’t really understand how.
Dylan Wiliam speaks about AfL as ‘responsive teaching’ and
part of that for me is adapting teaching practice rather than pupils necessarily
improving their work. Initially I would ask pupils to redraft but use the
extract more but as Daisy Christodoulou points out in her book, this is like giving
a comedian feedback to ‘be funnier.’ True, but not helpful.
What I then started to do was make sure I gave a lot of worked
examples of what application to context looks like. After they had seen full
examples this was faded with a partial worked example
where pupils had to
improve it to use the extract more before moving on to independent work.
I never bothered to redraft the original answer as the issue
was not that answer but how to apply to context in the future. Feedback should
be a medical and not a post-mortem.
This also eventually leads to the point where pupils fully understand
success criteria and lead to what Mark Enser speaks of as self-regulation, where
thy can assess and improve their own work as they will need to do in their exams.