A big gap between the number of high honours in different faculties affects any analysis of 'dumbing down', writes GARRET FitzGERALD
LAST MONTH, reports alleging grade inflation, or “dumbing down”, in second- and third-level education, prepared by a group called Network for Irish Educational Standards, secured considerable media attention.
The issues raised are important and, in principle, this group’s engagement is therefore to be welcomed. However, I have two major problems with the section of their report dealing with the universities, of which I have some knowledge, due to my former role as chancellor of the National University of Ireland. I shall confine my comments to the NUI.
First, whereas in the case of the institutes of technology, the authors distinguish between four distinct areas of study, in the case of the universities they based all their conclusions on an overall figure for each university. This is surprising, as it is widely known that overall averages of this kind can – and in this case certainly are – extremely misleading.
Second, they seem to have chosen not to contact the universities to establish the true position in relation to this matter, preferring instead to speculate as to what might have led to the outcome they are criticising.
These two failures in methodology undermine the validity of the conclusions in respect of the NUI. Had the authors inquired, they would have learnt that in 2002, on my proposal, the NUI Senate made adjustments to the universities’ grading system.
My initiative derived from my awareness of a large discrepancy between the substantial number of high grades traditionally awarded in science and engineering, and the very small number in other faculties.
This disparity reflected two factors: first, a striking unwillingness of the medical faculty, despite the exceptional quality of their students, to award more than a tiny handful of first honours; and, second, the fact that examiners in the humanities traditionally made 75 per cent rather than 100 per cent their maximum mark. With 70 per cent as the normal mark for the award of a first, the proportion of humanities students securing a first was between one-seventh and one-half the proportion of such honours in the sciences.
Such a disparity could be justified only if the quality of humanities students was vastly lower than in the case of students of the sciences – and it was possible to demonstrate that this was not in fact the case.
An analysis of the relationship between Leaving Cert results and the results of university primary degree exams demonstrated a close correlation between these two sets of figures. Of the students who had secured 500-plus points in the Leaving Certificate, 90 per cent were subsequently awarded first-class honours in their degree examination – whereas only 3 per cent of those with less than 400 points achieved such an outcome.
Having thus validated the Leaving Cert 500-plus results as reasonable predictors of high performance in subsequent primary degree examinations, there was a basis for checking whether there were significant differences between students of the sciences and the humanities that could justify many more firsts being awarded in the sciences. The table above shows there was in fact no significant difference between the proportion of high performers in these two areas – with, however, medicine recording a hugely greater proportion of high points students than the other scientific faculties. It was clear, therefore, that students’ chances of a first in their degree depended primarily on the faculty they chose –- with humanities and medical students clearly at a major disadvantage, Common justice required that this issue be addressed.
However the senate was determined to avoid the grade inflation that was damaging the reputation of the British university system. Accordingly, very conservatively, the steps taken were limited to measures designed to ensure that the humanities would in future mark students out of 100 per cent – and not, as had been traditional in most cases, out of 75 per cent.
Towards this end, descriptors were drawn up for each decile of the marking system. These required that a student who answered a question perfectly be given 90-100 per cent, that a near-perfect answer be awarded 80-90 per cent – and so on. By restricting our reform because of concern to avoid grade inflation, its consequences were rather limited – some will no doubt argue, too limited. Readers can make up their own minds by referring to the last two columns of the table below.
Data published in Britain show that 75 per cent of students there are now awarded a “good degree”(lst or 2.1), whereas because of NUI determination to avoid grade inflation, this percentage remained very much lower, at about 60 per cent.
It is of course a matter of debate as to whether the NUI has got the right balance between the Scylla of grade inflation and the Charybdis of undervaluing our best students.
A serious debate on this issue would, be very welcome – but it must be on the real facts, not grounded on misleading overall averages that fail to distinguish between the very divergent exam results in different faculties.