IMA Workshop in Cardiff: Effective Feedback in Mathematics
Yesterday I went to a workshop organised by Rob Wilson and Matthew Pugh at Cardiff University. It was a very interesting day, although it’s disappointing to still see all-male speaker lists in 2019. I’m going to try to write up a couple of key points from each of the three scheduled talks - please don’t take these as summaries of the talks which had a great deal more content.
Mike Robinson - Feedback ideas that work…sometimes
Mike gave the first talk of the day which was an overview of feedback in maths HE. He made a point which was picked up by later speakers, that the maths higher ed literature doesn’t have a lot of work on effectiveness of feedback practises.
Mike talked about what feedback is for, and made the point that our traditional feedback practises are better at helping students identify their mistakes than at helping them improve their work in future. Ticks-and-crosses, and model solutions, are especially bad at this - they usually don’t show the difference between a good and a bad solution.
Here are some of the “ideas that work…sometimes”:
- making feedback motivating and encouraging. We rarely tell people how well they’re doing.
- marking work with students in small/medium size tutorial classes. An attendee asked if students were bothered by the privacy aspects of this, MR said he hadn’t had any negative responses.
- changing summative into formative assessments wherever possible
- being explicit about what students should do in future to improve: “this would be better if….” “try exercise 3 again…”
- getting students to assess their own work using model answers.
- error codes and checklists of standard comments. There are standard errors you see all the time marking undergraduate work, and writing out the same comments each time isn’t necessarily a good use of marker time (especially as many students won’t even collect the work).
- talking about handling and learning from criticism. Mike gave the example of having papers rejected and suggested using this as a way to discuss this with students.
I asked a question about helping TAs to give better feedback when marking, since often in larger courses I have very little contact with student coursework. Some people suggested that the idea of pre-prepared comments and error codes would work well here. I wonder if our TAs know how little work is actually collected - it would be pretty discouraging to those who really work hard on their marking.
Barrie Cooper - Fast effective feedback: search for the holy grail
The ‘holy grail’ refers to giving feedback which is both timely and effective and doesn’t use a disproportionate amount of staff time. Barrie gave some key principles:
- Map and understand your feedback practises: what are you giving students and why?
- Cut your cloth accordingly - be realistic.
- Distribute the workload, through multiple markers, peer marking, and by spreading the assessment out over a longer amount of time.
- Discuss your approach to feedback with the students, so they understand when, how, and why they receive feedback and so they can be part of the solution.
Barrie had a list of ‘quick wins’, with the single most important thing being automated assessment and feedback. It was interesting to see that around 50% of people in the room said that their department was using some kind of computer-based assessment and feedback. A specific example he used was automated competency testing: repeatable randomized basic quizzes that students had to master before progressing. The idea of generic feedback - categorising typical errors and using a feedback key - came up again.
Michael Grove - Approaches to feedback in the mathematical sciences: what do students really think?
One of the most interesting things in Michael’s talk was about Birmingham’s maths support drop-in sessions which took place in their library if I remember correctly. I don’t have the exact numbers, but the percentage of people attending who were maths undergraduates went from around 30% in 2011 to over 85% this year - I wouldn’t have guessed anything like these numbers.
Michael talked about a project where they asked postgraduates who were marking undergraduate projects to provide feedback by recording videos in which they showed relevant parts of the project and talked about its strengths and weaknesses. The response from the undergraduates was mixed, but the postgrads seemed to find it a mostly negative experience for the all the reasons you’d expect. They didn’t repeat the experiment. This wasn’t at all the main part of the talk, but I think including approaches that didn’t work is something really valuable - it was great to have Michael share this (didn’t Tim Gowers say something similar related to research at a keynote speech years ago?).
There were some interesting statistics on surveys at Birmingham asking maths undergrads if they had received enough feedback on submitted work - I didn’t manage to get the numbers, but it was clear a significant percentage didn’t think so. Michael talked about the importance of dialogue in giving feedback, and about the small group (~6) tutorial system they have for first and second years. In the first year especially they have problem sheets which set up tutorial activities, and also peer marking and peer feedback. They even give credit for engaging with feedback on report writing.
Williams Wynn organ
The maths department in Cardiff is just round the corner from the National Museum of Wales, so I managed to get a look at the Williams Wynn Organ which my grandfather played on VE Day 1945 and which is now kept in the art section of the museum. During the second world war it was installed at Wynnstay Hall in north Wales where my grandfather was stationed, I think with the School of Military Survey. He worked on the Ladybird books but the love of his life was music - he played the organ for over 80 years.
If you’ve got access to six hundred and twenty five thousand pounds you have a shot at buying either a one bed flat in Camden or Wynnstay Hall. The flat won’t have Capability Brown landscaping though.