The Unusually Good Cohort

 It was never going to be the perfect process. When SARS-COV-2 landed on our shores back in February, the exam season was essentially cancelled immediately (although we didn't know that at the time).

No exams for pupils meant of course that the assessment process was going to consist of teacher-estimates (backed by professional judgement and whatever evidence we could remember for individual pupils based on how they had performed over the course of the year).

That raw process was always going to show a statistical aberration from what formal, invigilated examinations produced by way of grades produces, year-on-year. The usual process provides a reasonably good level playing field for pupils across the country and can be relied upon to ensure that pupils from every school in the country are graded on how good they are at sitting in front of a formal question paper (yes, indeed, we don't all agree on the suitability of that model for setting kids' life chances).

When the teacher-predicted results were submitted, gathered and analysed it became clear that there was a problem. Curiously the class of 2020 had performed significantly better than other years - a situation unlikely to have been the result of unusually good teaching and learning taking place between August 2019 and March 2020.

So. A problem. What to do?

Of course, dead easy. Look at the performance of pupils in each school in each particular subject across a three year period and determine where a cohort has clearly been over-estimated. Then take a statisical mallet to those results and hammer them down. Brilliant!

This will, of course, address the cases where for example new teachers have been too optimistic in estimating a pupil's performance (it happens; we all did it in the first few years of our careers because we believe in our awesomeness as teachers and we believe that pupils will ultimately perform well on the day).

So the "statistical mallet" will tease out and rectify those cases. No doubt about that.

But at this stage we need to consider the Unusually Good Cohort. The UGC comes along every few years. We recognise them quickly. They are usually a joy to teach; they work hard every lesson; they do their homework; they do well in early-term class tests. They are what we all imagine every class is like when we decide to become teachers.

And they stick out like a sore thumb, come exam time.

They exist. When I met my old French teacher six years ago for her ninetieth birthday celebration, she waxed lyrical about my year group (who left school in 1981). She described how my year group had been "noticed" in primary school in 1974. For the avoidance of doubt, although the UGC were my friends. I wasnt one of them.

I had a UGC a couple of years ago. Higher Physics. With a "normal" nationwide average of about 30% of pupils achieving an "A" in the exam, this cohort delivered just short of 50% A grades come exam-time.

But UGCs are statistically unusual - and this year they have been hit with the mallet.

Schools teaching pupils with fewer socio-economic challenges (please don't call them "good schools") don't have much headroom for unusually good cohorts to shine. Their results are usually high. But schools who don't enjoy such freedom from "challenges" have lots of headroom. UGCs will appear like a flashing blue light alongside (and indistinguishable from) over-optimistic estimates.

So, this is where the headlines about deprived pupils being disadvantaged have come from. Pupils with significant challenges, who have listened to teachers and parents, who have buckled down and worked unusually hard and who have done their school proud and who have shown that "good" schools and "bad" schools is nothing to do with postcodes have been slaughtered.

Part of an Usually Good Cohort? Tough.

This needs fixed. And quickly.