The Office for Students Access and Participation data, released on 29 March, exists to support registered providers in developing an access and participation plan for submission in May.
But for such a potentially explosive release, response has been muted. We are offered a tantalising glimpse of the student experience for a range of disadvantaged groups, and the chance to identify the providers where admirable progress is being made along with those where the progress is yet to come. If higher education is indeed, as Justine Greening suggests, all about social mobility we need this data more than ever. But the design of the data presentation means that few have been able to see the full scope of what we now know about the sector.
Taking a closer look at the data, which spans the entire student life cycle from access, to continuation, attainment, and progression onto further study, we found the expected outcomes trends based on students’ characteristics and backgrounds, but that the key predictive factors in access and participation are providers themselves, and the mission groups to which those providers belong (with the caveat that not all providers are members of a mission group). Looking across each stage of the lifecycle, we spotted trends among Russell Group, MillionPlus and University Alliance providers but also the broader categories of pre- and post-1992 institutions.
If you want to understand more about the data or to use our visualisations yourself to explore what is available, please see our data article.
Little deprivation in the Russell Group
Incredibly, if you exclude Queen Mary, every single Russell Group provider has admitted less than the sector average from index of multiple deprivation (IMD) quintiles one through four (that’s the 80 per cent with the most deprived backgrounds) for each of the five years in question. Inverting this – looking at access for the 20 percent of most advantaged entrants – shows that many of these institutions are disproportionately admitting better off applicants. Only Queen Mary is below the sector average here, and the only institutions within five percentage points of the average are Kings College and UCL, illustrating the London effect. By contrast, nearly 40 per cent of students admitted to the University of Exeter are from IMD quintile 5.
Increase in declared mental health issues
There’s been a great deal of recent attention paid to offering appropriate support for students reporting mental health issues. Just about every provider in England has seen an increase in entrants who declare mental health issues at the time of entry over the past five years. In the most recent year of data, Imperial College admitted the lowest percentage of students with declared mental health issues (0.9 per cent), whereas for Plymouth College of Art, the figure was 11 per cent. With measures like this, differences are likely due to more consistent reporting, but it is still striking that – overall – specialist arts providers see higher admissions rates than others.
Access for Black students
In terms of access for Black students, the sector ranges from 38 per cent of institutional intake (London Metropolitan) to one per cent (University of St Mark and St John). Marjon also has one per cent of its most recent intake from an Asian background, compared to 58 per cent at the University of Bradford – an issue it may struggle to address without some major demographic shifts over the next five years.
Looking across five years, the percentage of Black students at each Russell Group university has remained below the sector average. The sector average has risen from 8.6 per cent to 10.9 per cent over this time period – the only two Russell Group providers to pass the year one sector average in year five are Queen Mary and King’s College.
Continuation: the mission group factor
Continuation rates (the proportion of students who proceed from a first to a second year) for most ethnic groups of students tend to vary by mission group, though differences are most apparent among white students and least apparent among Black students for whom there is more variation by institution (possibly attributable to demographic spread as well as institutional factors).
Continuation for all ethnic groups over the five-year period has remained broadly stable across the groups, but is at a higher level among Russell Group institutions than in other membership groups, such as MillionPlus. There is a similar mission group trend for students who report having a disability – generally higher in Russell Group institutions, and broadly stable over the past five years.
It is noticeable that the sector’s average retention of both students aged over 21 and those aged 21 and under has remained quite flat over the period measured, albeit at a higher level for younger students.
Attainment: sector-wide rise
Though there is significant overlap, post-92 institutions tended to record lower average attainment rates (the number of students graduating with a first class or 2:1 degree) than pre-92 institutions – rates across the sector have improved at a steady rate since 2012, with little in terms of regional trends in attainment over the five year period.
Attainment across all undergraduates appears to be more influenced by institution than it does by ethnicity. We can see this by isolating the attainment rates for universities of different mission groups – over the five years, students of all ethnicities studying at Russell Group universities had similar rates of attainment, as did students of all ethnicities studying at MillionPlus member institutions. These trends hold true when we isolate Asian students, Black students and white students.
The sector average gap is steady over the five-year period at a little over 10 percentage points.
Progression of disabled students
It’s harder to tie progression data to a particular institution – course choice, background, and other characteristics have a well-documented impact on the likelihood of skilled employment or further study after graduation. This is an argument we’ve largely seen raging around Longitudinal Educational Outcome (LEO) data, but it applies here for the responses to the old Destination of Leavers from Higher Education survey, taken six months after graduation.
So for example Russell Group institutions, in the main, do better than Million Plus institutions at seeing disabled students progress to employment or further study (kudos to London South Bank and Staffordshire for bucking this trend) but this is very likely to correlate closely to POLAR or IMD for the overall institutional intake, which itself is linked to prior attainment. Some combination of all of these means that students from IMD quintile one, two, three or four graduating from Russell Group providers are more likely to see this kind of progression.
In all of these cases, there is less difference between the Russell Group and the better-performing end of the pre-92 providers – and the general trend for improvement over the years tends to hold across all groups.
That said, some smaller HEIs – particularly arts colleges and other specialist institutions – have seen a drop in the progression rate between year one and year five for disabled students. Though the sample size is likely to be low (a low number of disabled students attending) it does point more to wider issues around employment opportunities in the arts.
What we now know
Most of the values we talk about above are now not deemed to be statistically significant by the Office for Students – the data went few numerous iterations through the delayed release cycle which included a change to the significance calculation method. So in a strict, statistical, sense we know very little.
But this is a new process. The documented concerns with the usability and quality of the data will eventually be addressed – we will become more skilled in reading and understanding the findings, and in manipulating the data to reach useful conclusions. This release has been a crash course in data analysis for many access and participation staff – particularly in smaller providers. It’s been a developmental journey for those interested in sector trends too.
However, especially when we look at groups of institutions, we can see some fascinating trends. The diversity of the sector shines through, and in many cases the diversity is unexpected in scope and direction. It remains to be seen how these findings influence the development of institutional plans – and whether the simple act of declaring targets for higher education will really be enough to address multi-generational, socially-rooted, inequality.