The topic of grade inflation has become a cumbersome companion to assessment.
Together they lumber awkwardly through the education system in the UK and beyond, threatening to undermine compulsory and higher education alike. True, the average trend is towards increasing grades. This is particularly the case in the upper echelons, where three quarters of students are now reportedly awarded first or upper second class degrees, compared to half this amount 20 years ago.
What’s really going on?
Responses to this trend are typically characterised by a deficit discourse in which blame is appended to a decline in standards, and is labelled as grade inflation. In other words, an increase in grades without a corresponding increase in learning or performance. However, grade inflation is not a foregone conclusion. The need to find out who or what is to blame is a diversion from the development of a better understanding of what the key influences actually are. Whilst assessment results are subjected to a constant barrage of critique, grade inflation is the lazy rhetoric that interferes with the development of this greater understanding of the most influential determinants.
The obvious conclusion to draw is that standards are decreasing and degrees are being devalued. However, the obvious is often built on the unstable foundations of generalisation and a superficial understanding of individual contexts. The obvious clings on to oversimplification and kneejerk reactions. The reality, however, is a far less obvious affair.
Evidence for why grades are trending upwards, or the less loaded phrase of grade improvement, reveal a complex landscape. According to our recent research, the most influential determinants of grade improvement were shown to be the geographic location of an institution, research output quality and the increasing quality of student cohorts – although even this variable was determined on grade entry points, which the recent A Level debacle in the UK has pulled into question.
Structural changes and smarter students
Of course, these are not the only influential factors, with greater expenditure on academic services; the professionalization of teaching in higher education; the move to a modularised system with multiple points of assessment, and; the shift from norm-referenced (i.e. relative) to criterion-referenced (i.e. absolute) measures of performance being evidenced as significant determinants. What this evidence reveals is that a combination of student aptitude, and changes to the structure and quality of UK higher education, appear to be largely accountable for graduates attaining higher grades. It also, importantly, points to the problems associated with our criterion-referenced approaches to assessment being critiqued using a norm-referenced rationale. Evidence, on the other hand, that grade inflation in UK higher education is a significant issue, is far more difficult to ascertain.
Making comparisons even within a particular higher education system, such as the UK, is a complex affair. Different cohort sizes, approaches to assessment, and – indeed – global pandemics will all influence grade outcomes. However, improvements in grades being unexplained by a particular statistical model should not automatically be labelled grade inflation. Indeed, our analysis points to the importance of the local contextual features of each institution, which are rarely considered within existing research.
A question of discipline
The disciplinary diet of institutions, for instance, is likely to have an impact on the overall distribution of grades. History, Languages and Philosophy tend to award consistently higher grades than other disciplines. As a result, those institutions with more of these disciplines will likely have a higher proportion of upper overall degree classifications. What this points to, is the need for individual institution-based case studies. These would enable greater insight into the local contextual variables that are more and less influential in the grade attainment of students. They would allow for the exploration of similarities and differences, and the development of a more nuanced understanding of the factors informing the student experience and grade outcome.
Unfortunately, in the absence of this more nuanced understanding of the influence of local contexts to illuminate the determinants of grade improvements, the rhetoric of grade inflation is being left to wreak havoc. For example, further erosion of institutional autonomy with increasing governmental control is being threatened. Suggested penalties for grade inflation include the levying of fines up to 2 per cent of an institution’s income, the imposition of conditions on institutional registration, and the removal of degree awarding powers.
These externally enforced potential “fixes” are proposed by government ministers who have little understanding of how the higher education system works beyond obtaining an undergraduate degree and/or sitting on a select committee. They are unhelpful and damaging to a sector-owned and led response, and ignore the context bound nature of assessment that is enacted within and by the communities of practice in which it takes place.
A more meaningful perspective is that the variables consolidated within a variety of statistical models highlight a much more positive and proactive picture of a higher education system that is engaged in a process of continuous enhancement. Externally imposed ‘fixes’ are not the answer and will lead to the artificial distortion of grade distributions that will do more harm than good. As the foundation of our higher education qualifications we have a professional responsibility to engage with a sector developed and implemented means of explaining the unexplained variables associated with grade improvement. It is time that we subject the deficit discourse of grade inflation to the same level of critique that it is subjecting our academic standards to.
I’m not particularly impressed with much of this. It seems to outline the conditions in which changes are taking place rather than explaining them.
Here are some relevant terms that don’t appear in this piece: market; value added; attainment gap. There are other such structural matters that translate into pressure that can be put on staff.
In addition, several of the factors listed as “structural changes”, while they do contribute to the conditions in which grade inflation can occur, have been present for some time without rampant grade inflation taking place.
Finally, as students become more credentialed, shouldn’t we also be talking about grade inflation at pre-university level (especially A-level)?
Paul, you need to read the publication it is based on. The assertion about location is a real one. There are other pressures no doubt – as you say – it is a complex thing. There should be change and contextual admissions provides a way forward.
There is also a mathematical issue – in the recent years there has been pressure to use the full range of marks whereas previously it was uncommon to see marks above 80%.
If marks in the First Class range are higher, this can lead to a higher overall result, potentially taking students into the First Class band when previously they would have received a2(I).
I am equally underwhelmed by this article as Paul for a host of reasons – but to highlight two:
It fails to ask the fundamental question what do the grades mean/represent? What should a first represent? Compared with a 2.1? What are the fundamental differences between the degree classes?
And secondly it fails recognise that whatever statistical analysis you do you have no actual way of being able to compare whether grades given by one tutor in one institution are the same standard as the same grade given in another. Neither do we have any mechanisms currently to check the comparability of standards between different subject disciplines even in the same institution. Only through calibrating academics and their standards through social moderation activity can we hope to satisfactorily address the national standards issue
My disappointment with this piece is that it suggests that the problem should be viewed though institution-based case studies. This is to counter ‘potential externally enforced potential “fixes”’ but my concern is that this presumes and accepts that autonomous institutions can and should determine standards that essentially ‘belong’ to academics, more specifically to the subject communities to which we belong.
There is a wealth of research evidence showing variation in the standards applied in assessment, not only between subject areas but also within subject areas. Of course variation does not necessarily lead to grade inflation. And, true there is complexity around the claims of grade inflation and multiple contributory factors. However only looking at this from an institutional viewpoint is far too limited. We academics need to be confident that we are using standards that are also used by our colleagues across our subject community not just those in our institution. Gaining this confidence is easier said than done because standards used for the assessment of complex, largely open ended tasks that characterise higher education assessment are multifaceted, impossible to fully articulate and socially constructed. Despite this we need to find ways share and understand in meaningful ways the standards that we use and are used by others within our subject communities (way beyond criteria and level descriptors etc.) in order to seek agreement. Only then can we move towards a consensus of what standards should be applied in our subject and be confident that if grades are improving (or decreasing) it is because of the quality of the work being graded. And when we have dealt with that we can think about differences between subjects!