Some problems with academic standards and comparability
HEPI has recently published an interesting brief report by Professor Roger Brown on the comparability of academic standards in higher education. Whilst there is a periodic and reasonably predictable media interest in university standards, similar to the annual panic over the alleged decline in A level standards every August, academic standards remain one of the most misunderstood concepts in higher education. This absence of clarity of definition means that debates about standards are characterised by misconceptions and muddled thinking.
The HEPI report represents an attempt to address this problem. It is also a response to the 2009 IUSS Select Committee report which offered some staggeringly unhelpful and misinformed observations on universities but was also memorable for the challenge to the Vice–Chancellors of Oxford and Oxford Brookes Universities to compare the standards of degrees at their institutions.
When we took oral evidence, we asked the Vice- Chancellors of Oxford Brookes University and the University of Oxford whether upper seconds in history from their respective universities were equivalent. Professor Beer, Vice- Chancellor of Oxford Brookes, replied:
It depends what you mean by equivalent. I am sorry to quibble around the word but is it worth the same is a question that is weighted with too many social complexities. In terms of the way in which quality and standards are managed in the university I have every confidence that a 2:1 in history from Oxford Brookes is of a nationally recognised standard.
When asked the same question Dr Hood, Vice-Chancellor of the University of Oxford, responded:
We teach in very different ways between the two institutions and I think our curricula are different between the two institutions, so the question really is are we applying a consistent standard in assessing our students as to firsts, 2:1s, 2:2s et cetera? What I want to say in that respect is simply this, that we use external examiners to moderate our examination processes in all of our disciplinary areas at Oxford, and we take that external examination assessment very, very seriously. The external examiners’ reports after each round are submitted through our faculty boards, they are assessed and considered by the faculty boards, they are then assessed at the divisional board level and by the educational committee of the university. This is a process that goes on round the clock annually, so we would be comfortable that our degree classifications are satisfying an expectation of national norms.(1)
This attempt to sustain the really rather extraordinary proposition that all degrees represent the same standard of achievement by students regardless of the context or inputs did higher education no favours. The Vice-Chancellors and Roger Brown argue that the issue is not about comparability and, despite the contortions at the Committee, it is difficult not to agree with that proposition.
But where do we go from there? Is it simply a free for all? Do we just let market forces rule (if they don’t already – it is an employer’s market)? Brown suggests a number of steps intended to ensure a minimum level of achievement of all graduates. These graduate threshold standards would be intended to offer reassurance to all stakeholders that anyone with a degree had achieved to at least a minimum level. Whilst performance above the minimum would vary among students and across institutions this would be fine because at least minimum standards were assured. This approach is very reminiscent of the recommendations made in the 1990s by the Higher Education Quality Council’s Graduate Standards Programme (GSP)(2). The GSP sought to establish just such a set of minimum threshold standards and to codify a set of attributes which would encapsulate ‘graduateness’. Interesting, thorough and academic, the GSP proposals didn’t take off.
Perhaps they are back on the agenda though. As part of its approach Brown proposes a number of steps:
• Publish learning outcomes
• Refine benchmark standards
• Establish external examiner networks
• Improve assessment practice
• Replace honours degree classification
• Clarify definitional problems, eg with ‘comparability’
It is difficult not to feel a certain amount of sympathy for this approach which rightly recognises the fundamental futility of seeking to establish comparability of academic standards. Sustaining what has been described as the ‘polite myth’ of standards comparability, ie that a 2.1 in English from Cambridge is of the same standard as 2.1 in the same subject from a newly constituted institution, given the differences in every input measure is simply not credible. Yet this is what the sector traditionally argues and it is rightly criticised both in Brown’s report and, despite all of its other errors, the IUSS Select Committee.
Many of the problems in dealing with standards arise from difficulties with definition and Brown rightly identifies the need to address this. However, at the heart of the current QAA quality architecture is the notion that greater explicitness is required about standards in order to give all stakeholders confidence in the security of standards. Brown seems to accept this in arguing the need for learning outcomes and benchmark statements. But there is really no alternative to accepting the need to trust the judgement of professionals and the range of proxies devised over many years to assure the legitimacy of their collective decisions. National Vocational Qualifications (or NVQs, of which Alison Wolf has acerbically commented that they are ‘a great idea for other people’s children’(3)) and the extreme developments of the US learning by objectives movement sought to impose maximum explicitness and thereby to minimise the need for judgement. But attempts such as these to provide comprehensive explanations to students in advance both mislead and misrepresent reality and may, ultimately, endanger the standards they purport to uphold – the nature of learning is just not amenable to such detailed pre-specification. Moreover, explicitness about standards, cannot, in itself, convince that those standards are being achieved. There is no necessary correlation between description and understanding; this is simply a variant of a naming fallacy. Standards are not, and cannot be conceived of in an academic context as pure, absolute, Platonic forms but are relative, context-dependent and contingent.
Martin Wolf, although referring to the challenges of HE expansion, highlights a related problem about comparability:
‘if 50 per cent of the generation are to go to university and degree standards are to be the same everywhere, either everybody at Oxford or Cambridge gets a first or vast numbers of students must fail to get a degree altogether’. (4)
Whilst Brown suggests we should seek to sustain the notion of comparability of standards, at least at the threshold level, it is not clear that there is value in this, even if it is feasible. So, where do we go from here? There is huge difficulty in comparing standards, over time, between subjects, between institutions. They are different. There is no point in pretending otherwise. Establishing a threshold is not impossible and may well be helpful but it is questionable whether it is worth it in a system where over 60% of students receive first class or upper second class degrees.
(1) Innovation, Universities, Science and Skills Committee, Students and Universities, Eleventh Report of Session 2008–09, Volume I, HC 170-I, 2009
(2) Higher Education Quality Council (1997), Graduate Standards Programme Final Report, London: HEQC
(3) Wolf, A (2002), Does Education Matter?, London: Penguin.
(4) Wolf, M (September 26 2002), ‘How to save the British Universities’, Singer and Friedlander Lecture, delivered at Magdalen College, Oxford.