Yesterday we heard the news that we had finally won our case with Pearson/Edexcel about an A* grade in Art GCSE. Despite several negative responses to appeals and complaints, spurred on by the candidate’s parents, we persevered and ultimately won the case. It was just one grade, but it mattered to the student and her parents a great deal. As it should. Also, this is the first A* in Art GCSE any of our students has achieved in recent memory so it mattered a great to the school too. The fact that the student was in Year 9 at the time makes it all the more special. Beyond the relief and satisfaction that justice has finally been served, this case highlights a number of underlying principles about exam moderation, the crude nature of some statistical processes and the David vs Goliath nature of challenging examination outcomes.
The situation was that one of our Year 9 students (let’s call her Jane) had a special interest and talent for Art and was entered alongside the Year 11 cohort. All along her work stood out from the rest of the cohort; this made internal moderation quite complicated and, for numerous reasons, the marks submitted for the Y11s were on the generous side. As is right and proper, the moderator identified this over-marking and duly submitted a report that led to a regression of the marks. They were down-graded. This would have been fair enough had it not been that the same process applied to Jane. The moderator noted that her work was marked within the normal tolerance; informally it was acknowledged that her work warranted an A*. However, to our great disappointment, the final result was an A; a consequence of the global regression. The idea that her work might be judged against some absolute standard for an A* was blown out of the water; the grade was just the outcome of a statistical manipulation; them’s the rules!
We appealed using the standard EARS processes. We received a long explanation of how the process worked and how Edexcel had done nothing wrong. The marks stood. I’ve encountered this before: the stock response seems to be to avoid looking at the case but to defend the procedure. It’s infuriating. To appeal further we were told we’d need to pay for a full cohort re-moderation. However, even if we went to the trouble of setting up the exhibition all over again, this wasn’t going to solve the issue. We accepted the moderation of the Y11 cohort; we were appealing against the application of a global regression to Jane – a clear outlier. Further informal enquiries suggested that Edexcel appeared to have acted properly and that we would probably have to accept the outcome. Essentially, we were told that the rules themselves would have to be changed if we were to succeed and, even if that happened, it would only affect future cohorts and would not apply retrospectively.
Deflated, I rather gave up. It felt like a David vs Goliath scenario – but without the slingshot. I was inclined to persuade the family to move on; to accept defeat, albeit reluctantly. But the family persisted. Surely, something could be done. Having been given a bit of a push I stepped up a gear and wrote directly to Glenys Stacey – I went to the top! She could immediately see that we had a case but did also confirm that Edexcel hadn’t done anything wrong per se; they’d applied their published procedures. However, she offered a glimmer of hope suggesting that, instead of an appeal, we could make a complaint. I’d never considered this before but, sure enough, Pearson has a complaints procedure on its website.
The first response from Pearson was another negative:
To underline the response, the final line states:
It all sounds very reasonable and that seemed to be the end of it. Hopes dashed again. We could have left it there. But I had some questions that remained unanswered and I wanted absolute certainty that everything had been considered:
The complaints officer agreed to examine this question and, whilst waiting for his response I sent this rather emotive message:
A week later, to my delight and astonishment, we received this message:
And then this:
Although I think the subject team’s response was acceptable in relation to 5AD01, I do think that the effect that the moderation caused across both units was unfairly severe and I do believe there is a stronger case for considering her to be an outlier in terms of the 5AD02 internal marking and moderation. I do not, therefore, believe that the decision to include this candidate in either regression was the correct one.
I have asked that this particular candidate be removed from the regression for both units – 5AD01 and 5AD02 and that she be awarded the marks that were in line with the moderator judgments. This has now been amended on our system and updated results will be available to see on Edexcel online very soon.
‘Jane’ was delighted. Thrilled. Not just because she now has an A* but because she’d been robbed of it in the first place and the sense of injustice had been tough to take.
It’s a small victory in the grand scheme of things and we had to work far, far too hard for it – but it was worth it in the end. It highlights just how precarious some of these processes can be and how easy it is for one student’s grade to be swallowed by the machinery of a national exam board trying to maintain standards across the system. The idea that exam boards ‘cannot apply a stepped adjustment for just one candidate’ is totally unreasonable. As it turns, out, Yes They Can! If you don’t give up and keep making the case. Lesson learned: If you have clear case and the appeal doesn’t work, Complain!
David 1: Goliath 0.