Michael Gove’s reforms were a damp squib
22 August 2017
SUGGESTED
Brexit
21 August 2017
Government and Institutions
23 August 2017
Healthcare
1 November 2024
The publication of A-level results always occurs at a time when Parliament is out and not much is happening, consequently, it gets a big chunk of news coverage. And in most years there is controversy, with angry words from teachers’ unions, threats from some poor duty member of the opposition and a parade of anguished students and parents.
This year, though, complaints seem to have been relatively muted. The 2017 A-level results are the first fruit of Michael Gove’s much-hyped reforms. These were intended to give A-levels a kick up the backside. The awards had long been subject to grade inflation which raised the proportion of students getting As from 10% thirty years ago to over a quarter today. Many members of the famous ‘blob’ kicked up a fuss at Gove’s apparent iconoclasm. But they shouldn’t have worried: the reforms have turned out in practice to be a damp squib.
Despite some of the favourable comments on the reforms last week, in reality nothing much has changed. Sure, new assessments in 13 key subjects may no longer be based on modular work, may be decoupled from As, and may rely on final examinations alone: coursework has gone the way of the dodo. We could reasonably have expected that this would have had some impact on the results, with students less able to use retakes of modules to boost their performance and being required to retain knowledge over a longer period. But, even though Gove’s reforms were designed specifically to differentiate between students of varying ability, Ofqual regulator Sally Collier emphasised that ‘students have done fantastically well’ again and that ‘all our kids are brilliant’.
There has been only been the most marginal decline in the proportion of those getting grades A and A*. And even this has been offset by an increase in the proportion of high grades in those subjects which have not yet gone over to the new exam specifications. So universities are still faced with very large numbers of applicants with high grades.
It could not be otherwise. For, despite Collier’s insistence that academic judgments play a role in setting grade boundaries, with the implication that such judgment matters, the reality is that grades are determined almost entirely by statistical considerations. A-level examiners are only asked to set the grade boundaries between A and B, and between E and fail. The other boundaries, including the important one between A and A*, are set by a computer algorithm.
Even the limited exercises of examiner judgment at the A and E boundaries are tightly constrained. Results are not allowed to vary outside ‘tolerances’ based on previous years’ results, the past records of candidates’ schools’, and candidate performance at GCSE. Examiners who insist on saying that a higher proportion of students should fail, or a higher proportion should get an A grade based on this year’s performance, are just, in effect, told ‘computer says no’.
It may be that students sitting the new papers will have been challenged more than previously and have had a deeper and more thorough learning experience. But we have no way of knowing this other than what interested parties tell us. Perhaps first year tutors at universities this autumn will be impressed by the standard of new entrants, and some years down the road perhaps employers will be similarly pleasantly surprised. Well, maybe.
But it seems unlikely. As ministers come and go, our exam machine ploughs on ahead, adhering to state-determined syllabuses, giving out awards in state-approved proportions, and guaranteeing incomes to a limited number of state-approved examination boards. Schools looking to offer real choices to their pupils and wishing to get off the treadmill may opt in increasing numbers for the International Baccalaureate. Selective universities and employers may be further driven to seek alternative ways of choosing between a plethora of ostensibly highly-qualified candidates – such as setting their own entrance assessments, which several have already started to do.
Examinations have two functions. The first (and by far the most important) is to set students genuinely challenging targets to direct their learning, and to encourage them to take a real interest in the world about them. The second is to enable university and employer recruiters to choose between applicants.
Our current system, even post-Gove, does neither of these well. We need more competition amongst providers offering exciting new syllabuses where students and their teachers can take risks and exercise choices rather than conform to narrow and predictable guidelines. Gove, like most of his predecessors and successors, didn’t get this – indeed at one stage he wanted just one monopoly examination body.
And rather than attempting to give candidates grades – nominally linked to ‘grade descriptors’ which all award specifications spell out in mind-numbing and meaningless clichés – why not just set a pass floor (preferably set rather higher than today’s E grade, which passes around 98% of candidates) and then simply rank students in terms of performance, as they used to do in civil service examinations? Universities and employers could then glean far more information about the relative standing of candidates than they can from last week’s results.
This year, though, complaints seem to have been relatively muted. The 2017 A-level results are the first fruit of Michael Gove’s much-hyped reforms. These were intended to give A-levels a kick up the backside. The awards had long been subject to grade inflation which raised the proportion of students getting As from 10% thirty years ago to over a quarter today. Many members of the famous ‘blob’ kicked up a fuss at Gove’s apparent iconoclasm. But they shouldn’t have worried: the reforms have turned out in practice to be a damp squib.
Despite some of the favourable comments on the reforms last week, in reality nothing much has changed. Sure, new assessments in 13 key subjects may no longer be based on modular work, may be decoupled from As, and may rely on final examinations alone: coursework has gone the way of the dodo. We could reasonably have expected that this would have had some impact on the results, with students less able to use retakes of modules to boost their performance and being required to retain knowledge over a longer period. But, even though Gove’s reforms were designed specifically to differentiate between students of varying ability, Ofqual regulator Sally Collier emphasised that ‘students have done fantastically well’ again and that ‘all our kids are brilliant’.
There has been only been the most marginal decline in the proportion of those getting grades A and A*. And even this has been offset by an increase in the proportion of high grades in those subjects which have not yet gone over to the new exam specifications. So universities are still faced with very large numbers of applicants with high grades.
It could not be otherwise. For, despite Collier’s insistence that academic judgments play a role in setting grade boundaries, with the implication that such judgment matters, the reality is that grades are determined almost entirely by statistical considerations. A-level examiners are only asked to set the grade boundaries between A and B, and between E and fail. The other boundaries, including the important one between A and A*, are set by a computer algorithm.
Even the limited exercises of examiner judgment at the A and E boundaries are tightly constrained. Results are not allowed to vary outside ‘tolerances’ based on previous years’ results, the past records of candidates’ schools’, and candidate performance at GCSE. Examiners who insist on saying that a higher proportion of students should fail, or a higher proportion should get an A grade based on this year’s performance, are just, in effect, told ‘computer says no’.
It may be that students sitting the new papers will have been challenged more than previously and have had a deeper and more thorough learning experience. But we have no way of knowing this other than what interested parties tell us. Perhaps first year tutors at universities this autumn will be impressed by the standard of new entrants, and some years down the road perhaps employers will be similarly pleasantly surprised. Well, maybe.
But it seems unlikely. As ministers come and go, our exam machine ploughs on ahead, adhering to state-determined syllabuses, giving out awards in state-approved proportions, and guaranteeing incomes to a limited number of state-approved examination boards. Schools looking to offer real choices to their pupils and wishing to get off the treadmill may opt in increasing numbers for the International Baccalaureate. Selective universities and employers may be further driven to seek alternative ways of choosing between a plethora of ostensibly highly-qualified candidates – such as setting their own entrance assessments, which several have already started to do.
Examinations have two functions. The first (and by far the most important) is to set students genuinely challenging targets to direct their learning, and to encourage them to take a real interest in the world about them. The second is to enable university and employer recruiters to choose between applicants.
Our current system, even post-Gove, does neither of these well. We need more competition amongst providers offering exciting new syllabuses where students and their teachers can take risks and exercise choices rather than conform to narrow and predictable guidelines. Gove, like most of his predecessors and successors, didn’t get this – indeed at one stage he wanted just one monopoly examination body.
And rather than attempting to give candidates grades – nominally linked to ‘grade descriptors’ which all award specifications spell out in mind-numbing and meaningless clichés – why not just set a pass floor (preferably set rather higher than today’s E grade, which passes around 98% of candidates) and then simply rank students in terms of performance, as they used to do in civil service examinations? Universities and employers could then glean far more information about the relative standing of candidates than they can from last week’s results.
SHARE
