COVID-19 has put education systems under enormous strain. With sweeping lockdowns imposed across the world, 90% of the world’s learners have missed out on face-to-face lessons at some point during the last few months due to nursery, school and university closures.
In the UK, 90% of teachers report that pupils are doing less or much less work than usual. On average UK pupils have spent just 2.5 hours a day on school work, but a fifth (or 2 million) did no schoolwork or less than an hour a day. School closures are projected to widen attainment gaps between better-off and less advantaged learners by about a third.
We know good attendance – in person, where safe – will be critical to efforts to close these attainment gaps and we believe that a project we have been working on with Bristol City Council this year to improve school attendance could show how behaviourally informed approaches can help.
We began working with Bristol City Council on this project last year. A crucial part of our intervention development involved interviews with parents at local schools to better understand some of the barriers to good attendance. Whilst many of these barriers are structural in nature, other elements seemed amenable to a behavioural approach. For example, schools tend to express attendance as a percentage. This can be confusing and may not clearly signal a problem when one exists, since in a school context ‘90% attendance’ sounds positive, but actually reflects 15 days of school missed.
As well as drawing on what we heard from parents, we were inspired by research conducted by Todd Rogers and Avi Feller in the United States. In their study, Rogers and Feller found that sending letters to parents telling them how many days of school their child had missed, whilst stating the importance of school attendance and their ability to influence it, reduced chronic absenteeism (missing 18 or more days of school) by over 10%.
Drawing on this evidence, we developed a series of messages for parents, for example:
We then set up a randomised controlled trial (RCT) involving over 9,000 students across 22 participating schools in Bristol. Parents in the treatment group received a message (like the one above) if their child’s attendance was below 95% over the course of a half-term. This allowed the intervention to be dynamic and personalised, as well as giving students a fresh start each half-term. In practice, only a third of parents in our treatment group received the intervention.
Things rarely go exactly to plan in life or in an RCT and we had to end the trial early after COVID-19 led to school closures on March 20th in England (the trial began on the 6th January 2020, and we had expected it to conclude on the 20th July 2020). We think it is likely that this contributed to us not observing a statistically significant result on our primary outcome measure – the student attendance rate during the project. This was 93.1% for the control group and 93.4% for the treatment group, which is not statistically significant.
The effect of attendance updates on students’ cumulative attendance rates during the project
However, exploratory analysis revealed that:
- This approach boosted the proportion of students keeping good attendance records (95%+) by 4 percentage points (59.5% to 63.3%), and this result was statistically significant.
- The intervention was effective during the early stages of the COVID-19 outbreak (compared to the control group, students in the treatment group attended school more in the run-up to closures). This result was also statistically significant. This suggests that this intervention could be effective at boosting attendance when schools return, which is something we would like to test.
The effect of attendance updates on students keeping good attendance records
Attendance before and during the early stages of the COVID pandemic
Overall, these findings suggest that simple text messages letting parents know how many days of school their child has missed, delivered at the right time, can make a difference to student attendance – even during extraordinary times. We estimate that if everyone had received the treatment, an extra 350 students would have achieved good attendance records (95%+) compared to if everyone had been in the control condition.
We would love to run further studies to try to replicate this result. If you are interested in participating or partnering with us on a student attendance project, please contact email@example.com.
We would like to thank Bristol City Council for funding and collaborating with us on this research project. We would also like to thank the hardworking school staff across the 22 schools that participated – without their commitment and enthusiasm this work would not have been possible.
 We defined this as the proportion of days that a pupil attended school from the point at which the intervention was implemented to the end of the academic year 2019/20, out of all the days that the pupil was on roll in that same time period.