On Thursday we had the first A Level exam results in England without any adjustments since the COVID-19 pandemic. No matter what the outcome, it was to be a nightmare. Results were to fall back in line with 2019, doomsday articles appeared a few weeks back and then a very transparent “getting ahead of bad stories” push happened in the Times last weekend. We had Op-Ed’s from the Secretary of State for Education, Gillian Keegan, which followed an article with plenty of extrapolated narratives from university dropout data, blaming inflated grades running with the headline “Tougher A-level grades ‘vital’ as unprepared students quit university”. No direct evidence as such was provided in the article.

It’s not just this year that the nightmare started. Realistically it’s been ongoing for the last few years – and it’s a very different nightmare for different people. For all the nightmare is the disrupted education for young people, for some the emphasis is on the widening of gaps – the effects of technological poverty and deprivation, for others it’s the mental health issues that have spiralled since the return to school and the associated attendance problems, for some it’s the missed exams and ‘inflated grades’ for cohorts, but on the flip side for others it was a realisation that there is a different way of assessment that could be made to work, which is at odds with ongoing educational reforms that add greater emphasis to academic style examinations.

Below is a post adapted from Threads I made off the cuff this morning during an episode of insomnia. It focuses on A Level exams, the 2020 and 2021 cohorts, grade inflation and social deprivation.

Should A Levels really be based on single sitting high stakes exams?

I personally like exams. I like the ‘game’. Some say that is counter-productive to the learning experience, but I believe that is also a learning experience of itself – medium of presentation, knowing your audience, pressure situations and just the ‘randomness’ are all learning experiences students can take forward, even if the ‘exam format’ is not something they meet again.

Talking of ‘learning’, in theory I do much prefer the linear qualifications as opposed to modular due to the freedom you have for sequencing and interleaving or topics, and long-term retention of student knowledge – speaking as someone who also teaches on modular courses still.

Back to the main points. I do also acknowledge that there are students who struggle in exam scenarios for a variety of reasons. Fundamentally, there is not ever going to be a clear and accepted measure of learning – everything we can do is simply a measure of performance. External set examinations for the masses are a high control way of measuring performance, with lower risks than other methods.

The biggest concern I have is the removal of pathways that are not exam heavy, for those student who struggle with traditional exams – this is going to block progression and mobility. We are at a position where in a few years time finding a L3 qualification without an academic style exam as part of its pass criteria is going to be very difficult. That is my major worry. We are setting up for a system where students who struggle with exams are to be less qualified, by design.

Grade Inflation – A nefarious plot?

Let’s talk about the ‘algorithm’ first of all, as this ties some points together. In 2020, teachers were asked to grade and rank students. Only 1 piece of information was actually then used. The rank. The rank was mapped to the percentages of students who achieved certain grades over previous years at the institution.

This also meant smaller cohorts had the potential to be hurt by simple mathematical differences having a bigger impact. “10% of students get an A* here do they? Unfortunately this year you only have a group of 9…they’re 11% of the cohort…no A* for them”. It also did occasionally go the other way. I had a colleague where a student she had allocated as a grade D was moved to a B. The group was not as strong as previous years, but the ‘algorithm’ (or the people who set it up) didn’t care.

Fundamentally the DFE attempted to use a macro-solution (standards stay roughly the same nationally) at a micro-level. It was also painfully simplistic – and took in no information of the individual. People hear algorithm and think “oooh fancy techy-techy” – this was not sophisticated at all. Furthermore, by ignoring the individual, it was anti-social-mobility. It perpetuated “You are unfortunate enough to be in an area with a ‘sh*t school’, well that means YOU must be ‘sh*t’ as well”.

I felt the need to illustrate the algorithm as it explains the difficulty in standardising a national cohort of results without work to do it and compare directly with previous cohorts. In fact, the people who could do that were teachers! Who in my experience…did, when allocating grades which were subsequently ignored.

2021 was different. We had internal mini-exams we could choose to set with old exam material made available by exam boards. But we needed to make use of numerous evidence points we used. It had to have a breadth of curriculum, make sure we had the right weighting of assessment objectives, there were subject specific grading criteria shared (as no grade boundaries, it had to be holistic judgement), we had to monitor the level of control and later assessments took precedence over earlier.

Evidence had to be stored for all of this as well, the work needed to be available for auditing. This is important. As if they truly wanted to do something about ‘inflated grades’ then, they could have attempted a modified ‘awarding” and done it. If there were highly suspicious results at centres they could drill down…if the DFE didn’t, it was a choice!

Every year you make predictions about your students. I’m not talking the ‘official’ ones, but the unofficial narratives in your head. You might say ‘grade B’ for Timmy, evidence is there, numerous assessments at this level…but you know if there’s a lot of integration on the paper, or if paper 3 has a projectile where you form a tan(x) equation and a moments question with a hinge point, Timmy will struggle. However, Lilly who you might say ‘grade B’ for will really enjoy those questions.

They both have a bank of internal evidence that matches B grade standard students of previous years – but given the randomness of the exam, there are scenarios where they may both fail to get a grade B…in the absence of an exam what do you do? Ethically, they both get B’s – the evidence is there.

Contrary to belief and how it benefits media and government to frame, the ‘inflation’ is not a nefarious act at a macro-level, but simply the ability of ‘performance’ not to be a snapshot from a small timeframe on a given couple of days.

This might seem to contradict my earlier statement about exams being the best way. But as I said, they are the best method of ‘high control’ we have in the most cost effective way for large number, standardised qualifications.

A Level Exams 2023 – A reflection

From a personal point of view I have zero complaints with the results of my A Level cohort. A range from A* to E, and there was genuinely not a single surprise in there.

From a wider perspective, being from the North East, one of the key pieces of information that caught my eye was relating to A*/A grades, and how the North East was the only region to fall from 2019. Whilst there are some growth industries in the North East, unemployment is high and significant opportunities are low pay or in sectors of unsustainable employment long-term.

Major parts of the region consistently appear in the indices of deprivation index, in 2019 it had the highest level of digital poverty/inaccessibility in the UK mainland – is this a COVID legacy? Anecdotally the cost of living is having a huge effect on retention and attendance at institutions in the area, I personally see students working part-time more this year, and for more hours, than any year previously.

I will be interested in how the A*-B figures compare, the ‘high grades’ in the data. Does the gap widen even more? We’ve had years of being told that social deprivation shouldn’t be a factor in education. It’s a bastardisation of the idea we shouldn’t reduce expectations for students from challenging backgrounds. I agree, we shouldn’t, but we also can’t just ignore the issues that come with deprivation and pretend they aren’t barriers that influence student progress.

There is a lot of analysis that can be done. Headlines might have been made about simple reductions in grades, and I’m sure we’ll be firing up for the same for the GCSE results day at any point now…