Why the fuss about GCSE English marking?
Thought I'd post this because the monkey family has been close to the sharp end and reports in the media don't seem to make things clear. I have no personal axe to grind because Spider managed to pass even with all the problems this year, so he's alright Jack and off to do Science A-Levels. (In fact, he did great. Yay! Spider! \o/\o/\o/) But what happened to others is still unfair.
Yes, there has been grade inflation, widespread cheating and all kinds of problems. Other subjects have managed to adjust their marks to stop grade inflation without causing a storm. What happened in English Language was different. This is how it worked for Spider's board, OCR. Other boards will differ in the detail, but the general picture should be pretty similar.
60% of the total mark is for coursework, i.e. assessments under controlled conditions done at times chosen by the school. Usually, a hefty chunk of that 60% is already done and marked by the end of Year 10. It should certainly be finished by Easter of Year 11. This is marked by teachers using exam board criteria and marks are moderated by the boards. All the marks are in the form of 32 out of 46, 57 out of 78, etc. Total it up, percentage it, and you know for sure, not just as a prediction, how much each kid has got out of the 60% that's already been marked.
The remaining 40% comes from the exam, which can be sat either in January or June. Predicting marks for that is obviously harder since it depends on performance on the day.
So, when kids sit the exam, they and their teachers know exactly what their numerical mark is for the 60% already done. That mark cannot change. What can change is the grade boundaries. You expect these to wobble a bit from year to year, but what happened this year was that they shot up, apparently by as much as 10% between the cohort who sat the exam in January and those who sat it in June. This is even weirder if you remember that the 60% coursework for those two cohorts would be exactly the same--in fact much of it was done in Year 10 before most kids even knew whether they were being entered for the Jan or the June exam.
.Teachers know where the grade boundaries fell in previous years, which usually means they can predict English results very accurately since 60% of the total mark is fixed at an early stage. Understandably, they look at the January grade boundaries to predict the June ones, because exactly the same tasks are being assessed in both cases. Because predicted results have such a solid basis, it's not unreasonable for kids to decide that they've done enough for a safe C in English so they'd be better off concentrating revision time on other subjects. I bet many of the kids who've failed English and are therefore shut out of A-Level courses could have managed to push up to a C if they or their teachers had been warned of the coming change. Sure, it would be nice if everyone worked as hard as possible to achieve the best result they can, but education is now so focussed on targets and box ticking that cynicism is part of the process and you can't really blame kids or schools for taking a strategic approach.
You can argue that English has become too easy, that marking needs to get tougher, or that the wrong things are being assessed. But I can't imagine a valid argument why 42 out of 60 plus 31 out of 40 should get you a C in January but a D in June when the tasks and the marking criteria for the mark out of 60 are identical for the two sittings. (I've invented the numbers.)
Apparently, the only defence Ofqual can come up with is that they got it right in June but wrong in January. Not sure how that's supposed to make it OK.
Michael Gove meanwhile uses the whole fiasco as proof that the whole GCSE system is 'not fit for purpose' and needs to be replaced. He may well be right, but he is probably the last person on earth I would trust to set up a better system. Even if he can, how does this help the kids who've been let down this year?
Yes, there has been grade inflation, widespread cheating and all kinds of problems. Other subjects have managed to adjust their marks to stop grade inflation without causing a storm. What happened in English Language was different. This is how it worked for Spider's board, OCR. Other boards will differ in the detail, but the general picture should be pretty similar.
60% of the total mark is for coursework, i.e. assessments under controlled conditions done at times chosen by the school. Usually, a hefty chunk of that 60% is already done and marked by the end of Year 10. It should certainly be finished by Easter of Year 11. This is marked by teachers using exam board criteria and marks are moderated by the boards. All the marks are in the form of 32 out of 46, 57 out of 78, etc. Total it up, percentage it, and you know for sure, not just as a prediction, how much each kid has got out of the 60% that's already been marked.
The remaining 40% comes from the exam, which can be sat either in January or June. Predicting marks for that is obviously harder since it depends on performance on the day.
So, when kids sit the exam, they and their teachers know exactly what their numerical mark is for the 60% already done. That mark cannot change. What can change is the grade boundaries. You expect these to wobble a bit from year to year, but what happened this year was that they shot up, apparently by as much as 10% between the cohort who sat the exam in January and those who sat it in June. This is even weirder if you remember that the 60% coursework for those two cohorts would be exactly the same--in fact much of it was done in Year 10 before most kids even knew whether they were being entered for the Jan or the June exam.
.Teachers know where the grade boundaries fell in previous years, which usually means they can predict English results very accurately since 60% of the total mark is fixed at an early stage. Understandably, they look at the January grade boundaries to predict the June ones, because exactly the same tasks are being assessed in both cases. Because predicted results have such a solid basis, it's not unreasonable for kids to decide that they've done enough for a safe C in English so they'd be better off concentrating revision time on other subjects. I bet many of the kids who've failed English and are therefore shut out of A-Level courses could have managed to push up to a C if they or their teachers had been warned of the coming change. Sure, it would be nice if everyone worked as hard as possible to achieve the best result they can, but education is now so focussed on targets and box ticking that cynicism is part of the process and you can't really blame kids or schools for taking a strategic approach.
You can argue that English has become too easy, that marking needs to get tougher, or that the wrong things are being assessed. But I can't imagine a valid argument why 42 out of 60 plus 31 out of 40 should get you a C in January but a D in June when the tasks and the marking criteria for the mark out of 60 are identical for the two sittings. (I've invented the numbers.)
Apparently, the only defence Ofqual can come up with is that they got it right in June but wrong in January. Not sure how that's supposed to make it OK.
Michael Gove meanwhile uses the whole fiasco as proof that the whole GCSE system is 'not fit for purpose' and needs to be replaced. He may well be right, but he is probably the last person on earth I would trust to set up a better system. Even if he can, how does this help the kids who've been let down this year?