Does online learning reinforce disadvantage?

There will probably be some headlines about online learning and teaching following a new study from TASO, which claims at the top level to have found evidence for a widening attainment gap for disadvantaged students.

But we need to caution right from the beginning that this is a single study based on data from one, anonymous provider. And perhaps more importantly, rather than being about “online learning” in a vacuum, it’s about what happened to attainment and progression (at that one provider) over three years – the latter two of these, 2019–20 and 2020–21, being pretty damn unprecedented.

The institutional assessment data collected was collected from those courses with more than 50 enrolments – for a grand total of two courses – meaning that we have 1,011 students across 58 modules in the study.

The team identifies a 5.3 percentage point fall in attainment from students from POLAR4 quintiles 1 and 2 in 2020–21 compared to the previous year of study, contrasting with a fall of only 2.6 percentage points for students from non-disadvantaged backgrounds. The report suggests that this difference may be as a result of students studying online. However, this gap in attainment didn’t widen in 2019–20, and it’s hypothesised there that no detriment policies may have been the primary driver.

The report’s authors are careful to repeatedly caution that untangling the effects of online teaching and assessment from the wider pandemic context is not possible. And as contexts go, that was a pretty big one – the fact that the pandemic as a whole had a highly uneven impact across society and students is now pretty widely accepted.

This untangling is a challenge for another day, and one which hopefully will be undertaken:

With a sufficient number of modules with different strands of teaching and assessment it might have been possible to tease apart the effects of the move to online teaching from the effects of the pandemic. Future work should look at coding of teaching and assessment of modules to enable finer grain analysis of online/blended provision following the pandemic.

The report does have a whole host of interesting findings in other areas – there are no statistically significant evidence for differential outcomes in these particular cohorts (in terms of attainment or progression) for students with disabilities, mature students, by accommodation type, or between those eligible for bursaries and those not. But there was evidence for differential outcomes dependent on highest entry qualification – students with A levels (as usual) had higher attainment than those who did not.

There’s also a rapid review of existing studies conducted since the start of the pandemic (as well as earlier meta-analyses), which also suggested that online learning negatively impacted student outcomes, with low-income and academically at-risk students being most impacted.

For now, what’s been done is presented as a model for other providers to follow, to conduct rigorous analysis of their own online and blended options – as well as the progress of different student groups throughout the pandemic, we might add. And you’d hope this work is being done across the sector, as this issue has become too emotive and contested a topic to rely on anecdote and conjecture alone.