We’re always interested to hear how teachers are applying #cogscisci principles to their day to day teaching. Adam Wray (@AdamWteach) has been trialing structured revision homework and retrieval practice at his school, and has written about the challenges and successes he has faced so far (including an October half term update on how things are going).
How we used structured revision homework and retrieval practice to boost our students’ performance in exams
In reviewing the performance of our latest year 11 and having now had the experience of the 9-1 syllabus we decided that we needed to try something different. We needed to help our pupils retain the immense amount of knowledge needed and recall it under pressure in the exams. The below details how we brought together existing practice with new ideas in order to meet the challenges of the new GCSE.
I’d been listening to Craig Barton’s podcast, and that led to more research and engagement on twitter around evidence based teaching, and in particular around retrieval, spacing and interleaving, with work by Adam Boxer, Mark McCourt , and the wider #cogscisci community being particularly helpful. In parallel, my colleague Jim was coming to similar conclusions and reviving some year 11 revision techniques he had used in the past.
The key issues we saw were:
- A significant number of our students do very little structured revision of their own at home.
- Many hold the belief that its best to leave revision to the last minute as they’ll just forget it anyway.
- Many (particularly in Combined Science) see little value in getting the GCSE (“I don’t need it for anything”), unlike in English and Maths. Nationally 45% (Ofqual 2018) of combined science students will get at least one fail grade ( i.e. less than 4-4) so self-motivation for revision and new learning can be low in this cohort.
- In previous years little revision work was set until just before the mocks at Christmas, and then starting again close to Easter when all topics teaching completed. (Homeworks were previously to support new learning, not an issue per-se, but not giving any teacher directed revision of past learning.)
- Prior to this year low stakes quizzing was mainly on current topic / last lesson, which means students do not space their practice.
What we have started:
We are trialling a new approach in the Chemistry and Biology parts of OCR 21st Century Combined Science. ( Not Physics at the present time as it has a different timetabling structure and in using an alternate existing methodology can act somewhat as a control)
- The timetable gives Biology and Chemistry 1 double (1hr 10 mins) and one single (40 mins) lesson each per week. The double lesson will solely be focussed on new content and practical investigations. The single lesson will be used as the homework review / retrieval lesson.
- The school has provided each student the CGP revision guide, workbook and answer book. From the start of the academic year the weekly homework for Y11 is to revise 2-3 pages in the CGP books.
- We give them a sheet showing the topics and page numbers for each week through the year, with key dates like half terms, mocks and exams on it. This gives them a potential revision timetable for the year. This structured timetable ensures all the mock exam content is covered in time, and then the rest of the content before the final exams.
For the homework we ask them to:
- Complete the assigned 2-3 pages of question sheets from the workbook as best they can. They can try from memory, but can also use their CGP revision guides, textbooks, notes, parents, online resources to best answer the questions.
- They self-mark the questions using the mark schemes provided
- They reflect on the marking to target any further revision in their own time.
The structure of the single lesson is this:
|Time||Teacher activity:||Student activity:|
|0-10||Retrieval Roulette (10 questions) on board as students enter. Teacher welcomes students in, walks around reviewing homework, collecting marks from students to record in markbook, asking each student for areas they would like further review on and noting common areas for later teaching, then stamping the students book with a “verbal feedback given” stamp. (CYA for inspect!)||Get out homework, Open homework at starting page placed on desk. Answer retrieval roulette at back of exercise books. If finished roulette do extension work / flashcard review|
|10-15||Put answers to retrieval roulette up. Ask for showing of hands / writing on whiteboard on scores If appropriate reteach/ discuss around answers to roulette.||Read answers for retrieval roulette and correct answers, add up score. Communicate scores(old/new) to teacher when requested.|
|15-20||Review last week’s homework topic, highlighting key areas. V short re-teach or review of question if common error. Commenting on common areas for further revision – signposting resources.||Listen to teacher – no notetaking|
|20-25||Review next week’s homework sub-topics… key areas to think about, links to other topics, intent to “warm-up” neurons in long term memory associated with that topic||Skim read questions in coming week’s homework. Ask any questions relating to next topic.|
|25-40||EITHER Teach NEW topic material OR Modelling answers OR Split class / small group teaching plus extension worksheets / OR Computer based revision|
Other notes / background:
- The school has class sizes of max 20 in year 11, influencing ability of teacher to live review sets of mark homework during the class.
- Each class is mixed ability.
This is a free spreadsheet based resource that has a bank of questions across all the topics on a particular course, links to them can be found here. The spreadsheet produces a quiz that has 5 questions randomly from all the topics learnt to date prior to the current one, and then 5 questions on the current topic. The content / question bank can be adapted for any topic / scheme of work and there are many examples on Adam Boxer’s site and TES.
Cognitive science thoughts:
When revising we ask the students to START with questions and not to pre-read notes etc. this is to allow them the struggle to remember from previous learning to strengthen those connections, rather that to just pull it from a recent “crammed” memory of the topic. Doing questions rather than re-reading, note highlighting is supported by research studies by Roediger III, H. L., & Karpicke, J. D.
Once they have gone through the “struggle” they THEN go and look at notes and review the topic, going back the question to improve it. This gives them some purpose / mission in reviewing the notes rather than skimming through.
They then mark the question to find the “mark scheme answer” – further correcting, giving any cognitive shocks of wrong answers, allowing further review. This also is giving them ownership of their own learning through retrieval here. Having the “responsibility” of having the answers is a key new change for these year 11 students, getting marks and being judged on the marks is no longer the game or aim, it’s their learning and improvement that is important.
Within the lesson the retrieval roulette provides the retrieval challenge, interleaved across topics studied so far. This gives spaced, interleaved retrieval practice building student’s ability to work out what topic and type of question as they come across as in their exams rather than be answering a set of questions of a known type or topic. Much research has been done on the effects of retrieval, spacing and interleaving the studies by Dunlosky et. Al and Smith et al. support this approach to maximise knowledge recall in a stressful exam a significant time (hundreds of days) away from the initial learning episode.
Communication to Students
At the start of this new program / approach we gave a presentation and engaged in a short debate with students on what we were going to do and the cognitive science behind the approach, sharing some of the research study information from Busch and Watson, as well as other supporting information. Many students were still adamantly holding onto the belief that there is no point in starting revision activity now as they will just forget for the exams, much better to leave it to late April/May! We will continue to push and share the theory and approach, praising those that build on the structure of the homework with their own revision, and celebrating effort put in, with little focus to the students on the actual “marks” of any recall apart from signposting where extra effort goes in.
Measuring the effect / success
We are not exactly sure how to do this, but we will look at Mock exam performance in January, and the relative improvement on the Y10 end of year exams compared with similar cohorts last year as they will be significantly similar content. We may do a baseline set of testing with a set of retrieval roulette questions now, and redo the same question set just prior to Christmas. We will also do some structured interviews with students on their experience to gain feedback.
October half term update:
After running for half a term we have now had a few lessons practicing this new format in anger and I have the following observations and changes:
- Students are liking the structure and use of CGP workbooks
- Lower attaining / low botheredness students found it challenging to answer standard retrieval roulette questions. Especially on topics that they had been taught in year 9 / early year 10. They tended to be “deer in the headlights” and not attempt any of the retrieval roulette questions and be prone to chatting/getting distracted.
- Getting the discipline established to have workbooks on desk, opened at the page with the overall marks available has been a challenge… though slowly getting there.
- It took a while for all students to really get to grips with self-marking, and then writing in correct answers.
- Its been invaluable having the conversations with students on what they found difficult and what they did well, often building a whole class picture quickly (balancing equations, doing bond energy questions for example).
- I am wondering why I am collecting in marks at all, as I am getting the feedback I need from the conversations, and the marks don’t really tell me much. I might drop the specific mark and change to a subjective effort grading, and a different format so I can record topics / comments for each student. I won’t have the “Jonny got 17/20 in his last homework” (meaningless but makes parents feel you are on it) available for discussions with parents, but I would have a good view on the effort each student is applying, and an idea of areas they need more support – which I think is much more powerful.
- Students just fed back this week that they would like more time in the lesson to go over the things they found difficult in the homework. This is great and what getting the discipline in the other parts of the lesson will allow.
Multiple choice on Roulette:
After seeing work presented on twitter (sorry I cant remember/find tweet!) on using multiple choice on retrieval questions I adapted the roulette spreadsheets to make a customisable mixed open and multiple choice 10 question retrieval quiz to both print out and put on the board. As one of the more “bothered” students pointed out as we get closer to the exam time I should fade out the support of the Multiple choice questions. As a thought as I write this I could easily adapt my spreadsheet to give 2-3 levels of the quiz for print out in decreasing levels of multiple choice.
I have trialled the multiple choice for 1 lesson and it was well received and had the desired effect of the lower attaining / less bothered students making some progress through the questions whilst keeping the more able ones engaged.A copy of the mixed starter spreadsheet template with OCR 21st chemistry content can be found here
Effort Marking not Attainment:
I am going to change to recording an effort grade 1-5 rather than an attainment grade for the homework, and only discuss effort and improvements with students rather than attainment level.
More time for feedback / help on difficult areas:
Based on student feedback I am going to allocate more time both in planning likely areas, and in class on feedback and support for areas of difficulty found on “last week’s” homework.
Thoughts on future direction
We are initiating this in the beginning of year 11… our
thinking is that we need to adjust our approach from the start of GCSEs, if not
realistically the start of KS3 to optimise the spaced learning effects. To do that will take more significant work in
thinking through schemes of work, teaching topic/ learning episode sizes and
timing and the associated homework / topic testing / low stakes quizzing timing
and content. These all discussed well in
for Mastery mainly relating to maths and addressed with respect to GCSE
sciences by many including blog posts Damian Benney (@Benneypenyrheol).
References / Acknowledgements
Retrieval and interleaving:
- Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students’ learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14(1), 4–58.
Retrieval practice and stress
- Smith, A. M., Floerke, V. A., & Thomas, A. K. (2016). Retrieval practice protects memory against acute stress. Science, 354(6315), 1046–1048.
Testing better than reading (retrieval practice)
- Roediger III, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255.
Other key sources of information:
- Ofqual (no date) Relationships between grade distributions for students taking different combinations of 9 to 1 GCSEs, Ofqual website. Available at: https://analytics.ofqual.gov.uk/apps/GCSE/9to1_2018/ (Accessed: 17 September 2019).
- Mr Brown for use of GCP worksheets in structured homework.
- Mr Barton Podcast: http://www.mrbartonmaths.com/podcast/
- Busch, B. and Watson, C. E. (2019) The science of learning : 77 studies that every teacher needs to know. 1st edn. Oxon: Routledge.
- McCourt, M. (2019) Teaching for mastery. 1st edn. Woodbridge: John Catt Educational Ltd.
- #cogscisci, on twitter
- Retrieval roulette as originally created by @Adamboxer, with OCR21st century science question set from: https://achemicalorthodosxy.wordpress.com/2018/08/18/retrieval-roulettes/
- Optimal Time For Spacing Gaps (?) | mrbenney (no date). Available at: https://mrbenney.wordpress.com/2016/11/03/optimal-time-for-spacing-gaps/ (Accessed: 17 September 2019).