Taking on grade inflation in UK higher education

My colleagues at Universities UK, alongside GuildHE and QAA, are today outlining their commitment to tackling the complex area of grade inflation. As they begin to take this work forward on behalf of the UK Standing Committee for Quality Assessment, I wanted to expand on the drivers for this work and what we hope that it will achieve.

Grade inflation in history

Grade inflation is a long-standing issue in the UK higher education sector and has recently become a point of media discussion when annual Higher Education Statistics Agency statistics are published. This year’s statistics show that over a quarter of graduates received first class degrees in 2016-17, and that there has been a three per cent increase in first or upper second class degree outcomes in the past three years.

Historical data demonstrate that the phenomenon is found across the UK, and across the entire spectrum of the sector. It is of long standing, reaching back to the early 1990s at least. It is erroneous to say, as various commentators have recently, that it is the product of an increase in fees for UK undergraduates in England in 2012.

It is difficult to disentangle genuine grade improvement from artificial grade inflation. There are a number of possible reasons for grade improvement, including better links between secondary and tertiary education, changes to assessment mechanisms such as improved clarity of learning outcomes, or improved academic and student support mechanisms – or where an institution has improved qualifications of its intake, or addressed inadvertent bias.

Such improvements are welcome, and should be recognised for the value they add to the student experience and the quality of UK higher education. So there is a clear need for a rigorous and robust examination of the distinction between grade improvement and artificial grade inflation.

UKSCQA and its members are well-placed to lead on this work and to address these issues. As a co-regulatory body that provides sector-led oversight of the higher education quality arrangements that continue to be shared across the UK, it draws together partners from the four funding bodies, sector body representatives, academics, and students.

UKSCQA is currently leading on a consultation to review the UK Quality Code, and already oversees a programme of work to improve comparability of degree standards across the UK. At the same time, an element of externality is necessary so that students and employers can have full trust in the outcomes, and the programme team will draw upon external expertise to ensure that this is achieved.

Exploring grade inflation

There are already controls in place to mitigate the risk of grade inflation. These include the external examining system when working as its best, and UKSCQA is overseeing an HEA project to strengthen this system. More consistency in degree classification algorithms will reduce the influence of one current variable in shaping outcomes. However, it is equally important that the sector is able to examine, challenge and address grade inflation where it may exist. Therefore, the committee has asked its sector members, and QAA, to develop a substantive programme of work to explore the complex issues surrounding grade improvement and inflation. The programme will:

  • clarify the grade classification boundaries in use across the sector and make recommendations for the inclusion of criteria in credit and qualification frameworks;
  • investigate the drivers for the historical increases in the proportion of good degree outcomes across the UK;
  • propose a framework for how autonomous and diverse institutions working with funders and regulators can collectively control the risks of sector-wide grade inflation.

This work, which will be coordinated by UKSCQA members UUK, GuildHE, and QAA, will draw upon the expertise of independent researchers and a UK and sector-wide reference group whose membership reflects the increasing diversity of higher education by including representatives from the further education and alternative provider sectors. The programme will also produce:

  • sector-agreed criteria for degree classifications, which will appear in qualification frameworks
  • information for students, employers and the public about degree classifications and how they are managed by institutions and the sector;
  • guidance for institutions in controlling grade inflation and recommendations on UK-wide and national measures to manage risks.

There are a number of measures which can be used to reduce the risk of grade inflation, and these need to be recognised. These include the publication of degree outcomes data, external examiners (currently being explored through the HEA external examining project) and convergence of degree algorithms (explored in a recent study undertaken by Universities UK and GuildHE). A recent follow-up study to the UUK/GuildHE degree classification algorithms report has suggested that greater convergence is essential in addressing grade inflation.

Yet there is also a need for the sector to take meaningful and timely action to respond to stakeholder concerns on grade inflation, as other contributions to Wonkhe and elsewhere have suggested in recent days. UKSCQA will lead the coordination of a sector response on this issue.

It is committed to a thorough, robust and independent investigation of the issues, so that students, employers, the sector and its wider stakeholders can have confidence in the comparability of standards across the UK.

5 responses to “Taking on grade inflation in UK higher education

  1. For confidence of comparability of standards across the UK, AND comparability of standards over time, would it not be best to allocate grades according to a standard distribution? Just give a 1st to the top 5%. When I was studying, some time ago, my friend got a 1st. They were one of the few; they were an outstanding graduate and deserved to be recognised as such. Todays 1st graduates are each just one of many. The fact that pretty much anyone can get a degree these days, in any subject they want to follow, and the fact that a 1/4 of students get a 1st, leads people to say that todays young people are better educated, even more intelligent – and they are clearly not all outstanding by any means! I think O levels used to be graded according to a bell curve (?) – so a grade A was something pretty special rather than an expected result.

  2. The long-standing problem with the standard distribution method is comparing like with like across years. Just because someone is in the top 5% in one year doesn’t guarantee that the same level of performance would get them into the top 5% in any other.

    SD-based grading is great at telling you how much better than everyone else in their year a student was but not how they compare to a graduate from any other. Imagine the chaos in sifting out applications for job vacancies or further study where you’re looking for a specific level of ability and not a relative performance.

  3. @Amanda Bowles:

    No, giving grades by top x% makes things completely incomparable across years, as you can end up with a bunch of smartypants one year and a bunch of dummies the next. Setting a First as top x% would result in some smartypants who are better than the dummies getting lower grades than them.

    You also have additional difficulties with the cohort: do you do the grades on a programme basis (i.e. top x% in BSc Fishing Studies get a First) or the entire graduating cohort at a university (top x% of the whole year group get a First). In the former scenario, you have a problem with cohort size: there will inevitably be programmes with only a handful of students where you can’t make a sensible distribution. With the latter, you get comparability problems. Is it really fair to compare performance in BSc Mathematics to performance in BA English?

  4. I am curious to know how much recent quality measures are affecting grade inflation. At my institution for example, we have been given ‘guidelines’ on how many firsts and 2.1’s we should be giving…..many more we are told (!). To what degree are TEF measures such as ‘value added’ contributing to grade inflation? Your thoughts on this issue would be appreciated

  5. Grade inflation was a very minor issue in the 1990s. It became an order of magnitude worse due to the uncapping of university places and general marketisation of HE. League table position is the number 1 driver of applications to an institution, and the grades churned out has a huge effect on this, either directly or through “Value Added”-type confections. Institutions need applicants, and lean very heavily on academics to increase grades, to increase numbers. Academics mark their own homework and like to avoid causing administrative trouble so they acquiesce. The normal QA process is ineffective because everybody is doing it slowly enough that they don’t get caught out. So little of this inflation is due to genuinely better results that it’s not worth mentioning. The answer is twofold. Amanda Bowles is correct, part of the solution is to apply a standardised distribution of marks, but on a per-module basis and with individual care and attention from the tutor concerned who will generally have a feel for several years’ worth of delivery and a sense of particularly good or bad cohorts, i.e. “aim to give 10% first-class marks unless you have amazing students, then give higher marks”. Second, grades, outcomes, whatever you want to call it, should be removed from league table calculations (if we don’t abolish them altogether-which we should.) All the rest of the crap in this article is just a waste of time. For me it’s an example of policymakers at the top panicking at having no way to control the emergent complexity.

Leave a Reply