Building interleaving and spaced practice into our pedagogy
This post builds on the conclusions I reached in the post Forgetting is necessary for learning, desirable difficulties and the need to dissociate learning from performance. Make sure you’re up to speed with that post before reading on.
In my department we have historically done blocked practice from year 7 to mid-year 11 and then switched to interleaving from Christmas in year 11 onwards in the form of weekly past paper homeworks. According to Bjork and Rohrer’s research we should see greater sustainability in learning outcomes from the interleaving part of our practice. I thought I’d dig into our own assessment tracking data to see if it agrees with their claims- it does!
The following graph shows the mean assessment level (in terms of progress to go to meet Minimum Expected Progress) of our 2013-14 cohort as they progressed from starting GCSE maths in year 9 to their final GCSE grade:
It would appear that the story in our assessment data agrees well with Bjork’s and Rohrer’s findings on interleaving and spacing. This cohort went on to achieve excellent progress and attainment when compared with national averages (80% A*-C), but it was only when they started the interleaved practice sets of questions that the retention and transferability of their learning began to build substantially. Interleaving is effective because it gives students experience in selecting a strategy to solve a problem as well as executing the strategy. In blocked practice we give students the strategy and they don’t gain experience in selecting it.
I plan to bring interleaving and spacing forward in the department’s practice right to day one in year 7 in our new SOWs. There will still be blocked practice during the early stages of learning, but students will get interleaved questions at regular intervals right through the five years. Currently, I’m thinking of this being in the format of open-book end-of-unit activities that recap the content of the whole topic, then another to recap topics taught cumulatively to that point. Bjork says low-stakes/high-frequency is important for these to be successful, hence these being open-book activities. Spacing is inherent in interleaved practice, but I am also planning for starters to be used for retrieval events of previous material.
There still remains the issue, as discussed in the previous post, that if we are doing this correctly, in-lesson performance of students will be lower. As Rohrer showed, an interleaved-taught unit showed significantly less in-lesson progress, but ultimately three times the long-term retention and transfer of student learning. Managing staff/student expectations isn’t going to be easy.
Food for thought.
Detail behind the data:
Students sat full GCSE exam papers during the assessments. The grade boundaries on all assessments up to the exams were higher than the final GCSE exam turned out to be (35 for a C) by approx 0.3 levels. This would have the effect of reducing the interleaving gains to 1.0 levels over the ⅔ of a year- still considerable and a step change from the rate of progress during blocked practice.