The former TES journalist writes for NAHT on current education issues. The views expressed do not necessarily reflect those of NAHT
Ofqual should continue to put pupil fairness first
I made an interesting discovery last week. Ofqual, the exams regulator, no longer seems to say explicitly that one of its key priorities – if not the key priority – is to treat pupils fairly.
I find this surprising, and I think it has important implications, of which anyone who has been following the GCSE English marking dispute should need no reminding.
But before going on to that, a bit of background is needed about this year’s marking saga, which is currently subject to an investigation by Ofqual following prompting by the NAHT. Ofqual is due to issue a report following that speedy investigation today (Friday 31st).
Many teachers are enraged because of a change during the course in the marks needed for a C grade in a new modular English GCSE. Of course, this has significance not just for pupils, but for their schools.
A foundation paper set for pupils in January had a C grade boundary set by the board, AQA (by far the most popular English board at GCSE) as 43 marks out of 80. Yet a new version of the same paper, set in June, saw candidates needing 53 marks out of 80 for a C.
The higher tier paper– taken by pupils aiming at top grades – had a C grade boundary of 41 in January but 44 in June.
Similarly, schools submitting their pupils’ coursework (which is now called controlled assessment) in January saw their candidates needing 43 out of 80 to gain a C. Those who decided to submit later in the year, in June, saw the grade boundary rise to 46 out of 80, even though pupils were working to an identical list of assignments throughout the year.
Finally, schools could also choose when to submit the third section of the GCSE – covering pupils’ speaking and listening – in January or June. Those choosing to submit in January would have needed 25 out of 45 for a C, but 28 out of 45 by the time June arrived.
The obvious explanation for all of this, as I wrote last week ( http://bit.ly/PETesS ) and has been widely reported, is relatively simple. AQA, uncertain as to how many grades to award in this, the first year of a very complicated new specification, may have awarded what in hindsight looked too many C grades in order for it to comply with an Ofqual policy which is seeing the overall number of any particular grade – over the entire qualification – kept at roughly the same as it was last year.
But, having already given out the grades it had part-way through the course, the only way the board could react was to raise grade boundaries for the only section of the course over which, as of this summer, it still had control: the June elements of the assessment. So it pushed up those boundaries, and the national proportions of A*-C grades were slightly lower than last year for English GCSE as a whole. It seems very reminiscent of the 2002 crisis which followed the first award of end-of-course grades under a new modular set of A-levels.
I’ve concentrated on AQA here, but it has been suggested that other boards were affected, too.
As I wrote last week, this may have produced the outcome which Ofqual, seemingly with agreement from the boards, wanted this year: that years of grade inflation come to a halt and that the proportion of grades, on average, does not rise. But was it fair to individual pupils, and to their schools?
If, as seems likely from the above, pupils – and their schools – submitting work earlier in the year had it “easier” than their counterparts who were given all their grades in June, how can that be right? How can it be just for one pupil to be given an easier route to a GCSE C grade than another, based simply on when they were assessed? And, given the huge consequences for schools which can follow based on success or failure in raising GCSE A*-C rates in English, if schools entering pupils early for an assessment have had an advantage over those leaving it until later, how can this be right?
To put it another way, if it was faced with already having set grade boundaries for earlier parts of the GCSE lower than they should have been, given Ofqual’s drive to curb grade inflation, was it right for AQA then to react by making it harder for pupils to obtain a C grade on June papers, in order to ensure that the overall A*-C grade percentage was kept in line? Or should it have it effectively reacted differently and said: “we’ve made a mistake with the earlier grading. Now, to be fair to later entrants, we need to ensure they are not disadvantaged. If that means overall A*-C grades rise, so be it: the principle of fairness to individual pupils, and to their schools, takes precedence.”?
To sum up by asking more rhetorical questions, has the need to contain grade inflation at a national level been put ahead of what could in this case be a competing need for equity to pupils and schools in the awarding process? Does the principle that a C grade means the same for one pupil as for another each year been sacrificed in the drive to contain grade inflation?
Ok, so I next wondered what principles Ofqual worked to in seeking to answer these questions. What are its guiding objectives? And my mind drifted back to the regulator’s early days from 2008, when it was led by Kathleen Tattersall, coincidently the former director general of AQA, who was Ofqual’s first chair. She used to stress in speeches, I think, that fairness to pupils was Ofqual’s first priority.
You can see this, for example, in a press release ( http://bit.ly/RqlSkq )setting out changes to A-level in June 2010. Under the headline: “New A-levels: fairness is our priority”, Ms Tattersall says: “We must have an A-level system that is fair and consistent for all learners, past and present.”
Indeed, looking back to other aspects of what Ofqual was saying back in 2010, it seems, unsurprisingly I think, that fairness was still being stressed in what it said about how it regulated then. Here is a page from Ofqual’s website, taken as a snapshot by the National Archives from February that year http://bit.ly/NudTDJ . Crucially, I think, this says “we want our decisions to be clear and understandable and will intervene when necessary when there is a risk to the fairness of the system or to an individual learner.”
If Ofqual were still talking explicitly about fairness to learners now, I think - so long as the widely-offered explanation set out above as to what has happened this year with GCSE English stands up to scrutiny - then there would be an unanswerable case for changing some of the grades awarded this year.
If pupils have been advantaged or disadvantaged according to the time of year they obtained their grades, then, if the priority is to be fair to the individual pupils, either grade boundaries from the January session have to be increased or those from the June ones should fall.
The interesting thing, though, is that in the last couple of years Ofqual no longer seems to have been talking nearly so explicitly about treating fairness to learners as a priority. I do find this surprising, and I guess it will be frustrating for schools who are furious about the decision and want the regulator to take action. But I could find no direct reference to treating learners fairly in the regulator’s current “how we regulate” page, nor in its corporate plan.
In fact, under “how we regulate”, Ofqual today lists three priorities:
“- Maintaining standards in qualifications and assessments.
- Raising awareness of any issues while maintaining public confidence in the qualifications system.
- Ensuring the qualifications industry is as efficient as possible.”
Once again, I am surprised. I put that sense of surprise to Ofqual’s press office, and had it confirmed that: “While our website may not specifically refer to fairness to learners, it is in everyone’s best interests to have an exam system in which standards are secure, grades are right and which commands the confidence of the public.”
Well yes, that macro level perspective sounds very sensible to me. But what about the micro level: the learner and school perspective? It seems to me that the notion of treating different pupils taking the same exam fairly and equally should be such a fundamental part of the grading process, it should come before anything else I can think of. All pupils do ultimately have a stake in the reputation of a qualification. Yes, grade inflation can devalue the currency in the sense at least of meaning that, as I’ve argued here before, it’s not clear to me that there is necessarily value to individual pupils when national results as a whole rise, because employers and universities are likely to raise their grade demands accordingly. But an individual pupil is affected far more directly when, as is alleged here, he or she has not been treated equally with peers. Equality in treatment under the grading process has, surely, to come before considerations of whether the national figures nudge up slightly or down slightly. If we get equality right, then it may be okay to try to take action against grade inflation. But only if that is the case: the latter aim, of course, should not be prioritised over the former.
Ofqual’s objectives stem from the law that set it up, the Apprenticeships, Schools, Children and Learning Act 2009, which was passed under Labour and only amended slightly in 2011 by the coalition. Although the notion of Ofqual’s political independence is not simple and is something I want to write about further, any suggestion that a downgrading of emphasis on pupil fairness since 2010 has happened because of the change of government may therefore be a simplistic reading.
In fact, I have heard it argued – and not by Ofqual – that the fact that learner fairness does not feature in the list of explicitly stated priorities does not mean, of course, that it is not an important consideration at all times, for the boards, and for the regulator. Despite the events of the past week, this still seems plausible to me. Let us hope that this is the case, and that the principle of fairness prevails today.