A level inequality is baked in, not a 2020 phenomenon

The detailed equality analysis of 2020 A levels is here!

David Kernohan is Acting Editor of Wonkhe

And there are a few surprises in store, especially in the light of admissions reform conversations that are ongoing.

The analysis looks at the CAGs which were the Centre Assessed Grades that most of the class of 2020 ended up being awarded. Often improperly seen as teacher predicted grades, these also went through moderation at centre level. And there is a section on the “calculated” grades – the PM’s “mutant algorithm” – which were based on the CAGs, the centre level “rank” of students in each subject, centre results in previous years, and the standard A level “grade to a curve” mechanisms. There’s also colums for the final grade – the highest out of the CAGs and the calcuated grades.

We get to look at sex, ethnicity, major language, specific educational needs (SEN) status, free school meals (FSM) status, and socioeconomic status (high, medium, and low) – focusing on the proportion of candidates achieving grade A or above, grade C or above, and actual mean grade..

Step back

Before I tell you what it actually says, I’d like you to guess what the findings may be. Check the box you think is most likely to be the correct story.

How did different forms of grading affect the grades awarded to different groups of students:

Don’t worry about pressing enter, this is just so you can make a note. Your answer will have elicited your position on the role of exams in education – if you think exams are a great equaliser, removing the impact of prejudice among teachers and allowing for a fair assessment of a student’s potential, you would have ticked the first box. If you believe that economic status has a constant impact on educational attainment, however it is assessed, you would have ticked the third box. And if you believe that exams as an assessment method are unfair on students from disadvantaged backgrounds, you would tick the second box.

Step forward

The report tells us that:

By comparing attainment gaps in 2020 with those from exam results in 2018 and 2019, this demonstrated that the calculated grades originally issued to candidates on A level results day neither introduced new, nor exacerbated any existing, attainment gaps based on protected characteristics or socio-economic status.”

So the public revulsion over calculated grades was not because of anything about the grades themselves, it was more the way the mirrored closely the results in any other year without including the potential for any individual student to outperform expectations. The interim technical report (remember that!) set out the ways in which CAGs had been moderated to fit a typical annual grade profile, and demonstrated how less well off students were more likely to be downgraded from their CAGs.

Outrage, and a Gavin Williamson u-turn followed.

And the report concludes:

There is no evidence that either the calculated grades or the final grades awarded this year were systematically biased against candidates with protected characteristics or from disadvantaged backgrounds”

In other words, this year did not introduce any new bias against students from disadvantaged backgrounds. All the problems that we saw are a part of the system as it has been run in recent years, not new for 2020.

Points mean prizes

So if you ticked the third box, have yourself a biscuit.

We have, however, learned something important. Disadvantage lurks beneath the award of A level grades every year… if you are from a disadvantaged background you will do less well in your A levels, and we know this even before we take account of how hard you have worked or how clever you are.

The other parts of the university applications process – the statement, the references, the predicted grades – go some way to addressing this. If we want to put more emphasis on actual grades in admissions, we need to do some more work on why A levels favour the better off.

Along comes SMC

The day has also been enlivened by the Social Mobility Commission, who has released a set of recommendations about the future shape of 2021 exams. It suggests that:

  • School performance tables should be suspended in 2021.
  • The Department for Education, and Ofqual, should produce a clear, simple system for collecting centre assessed grades that can be used as a contingency.
  • Students should be allowed the option of an exam series in the autumn, as they were in 2020, without this being labelled as a ‘resit’, and the results should be released in time for the UCAS deadline for 2022 entry.
  • Additional places for full resit years should be funded at the same rate as current 16-18 funding.
  • The use of a back up exam is a sensible option, but it should be after the main series, not before.
  • Mitigations in content and structure of exams benefit all candidates, and so do not address gaps between those who have struggled with remote learning due to home circumstance and those who have not. As such, while some adjustment (like the reduction in content of English Literature) may be practically necessary and useful, it should not be regarded as a solution.
  • Generosity in grading for 2021 should aim for a midpoint between 2019 and 2020, but following a normal mathematical distribution, rather than replicating the anomalies of 2020.
  • Schools must have access to extra venues and extra staff to invigilate those venues to enable them provide Covid-secure environments during examinations.
  • Arrangements for students isolating at the time of exams have to take into account the vast difference in personal and socio-economic circumstances. Home invigilation should be avoided.
  • Arrangements providing grants and opportunities for gap years for those with fewer familial resources should be retained.
  • At the moment, some courses prejudice those who have done an extra year, and some institutions struggle to accommodate retakes of years because of funding reductions for older students – this could easily be addressed.

Leave a Reply