Commissioner for Fair Access discussion paper: retention, outcomes and destinations

Discussion Paper: Retention, Outcomes and Destinations

This is one of a number of discussion papers that will be published on the Commissioner for Fair Access website on key issues relating to fair access. The aim is to bridge the gap between detailed research (where it exists), which is often only accessible to experts, and the wider public conversation, especially in political circles and the media. The hope is that these papers will contribute to, and stimulate, that conversation by presenting data and evidence as accessibly and objectively as possible. Each paper will also include a commentary section by the Commissioner.

This paper considers:

• Trends in retention, outcomes and destinations for full-time first degree entrants ^[1] to Scottish Higher Education Institutions ( HEIs) from deprived backgrounds ^[2] over recent years;
• Patterns in retention, outcomes and destinations by subject and institution type;
• The Commissioner's thoughts on the challenges presented by the data.

Background

The Commission on Widening Access ( CoWA) focussed on what could be done to remove barriers to entry but also acknowledged the need to look at what happens next:

"Fair access is not just about ensuring more people from deprived backgrounds enter higher education, it is just as important to ensure that they can maintain their studies and successfully graduate."

In recent months the Minister for Further Education, Higher Education and Science reinforced this message, highlighting at the first meeting of the Access Delivery Group that it is vital to deliver access to success, in terms of retention, qualification and outcomes.

The Commission's final report incorporated retention into a number of its recommendations (8, 19, 24, 26, 30) and made a specific recommendation (33) for the Commissioner for Fair Access to look at inequalities in outcomes and destinations:

The Commissioner for Fair Access should: ...consider what further work is required to support equal outcomes after study for those from disadvantaged backgrounds or with a care experience.

This paper is part of the initial work being undertaken to address this recommendation.

Distribution of entrants by subject and institution

The analysis in this paper focuses on students from SIMD20 areas attending Scottish HEIs. We know that learners from SIMD20 areas are under-represented among Scottish domiciled full-time first degree entrants ^[3] but before we look at retention, outcomes and destinations, we need to consider the distribution of SIMD20 entrants by subject and institution type and how this compares to the distribution of entrants from other areas.

Chart 1 shows the distribution of entrants by institution type ^[4] in the three academic years from 2013/14 to 2015/16, the latest year covered by the Higher Education Statistics Agency ( HESA) student dataset at the time of writing.

Chart 1: Full-time first degree entrants, by institu-tion type and SIMD, 2013/14 to 2015/16

Source: HESA student data

The majority of entrants from SIMD20 areas were at post-92 universities and a lower percentage were at ancient universities compared to entrants from other areas. The data for previous years shows a very similar pattern.

Chart 2 shows the distribution of entrants by subject area over the same three year period. The distribution is broadly the same for SIMD20 and non- SIMD20 entrants, although a slightly higher percentage of SIMD20 entrants were on Allied Medicine or IT courses and slightly lower percentage were on Engineering, Physics or Medicine courses.

Chart 2: Full-time first degree entrants, by subject and SIMD, 2013/14 to 2015/16

Source: HESA student data

Again, the data for previous years shows a similar pattern. Additionally, the distribution by subject is similar for SIMD20 and non- SIMD20 entrants within each institution type.

Retention

For universities, retention is traditionally measured by looking at the number and percentage of entrants progressing from their first year to their second year ^[5] . The general consensus is that this transition is particularly important and, according to HESA data, more full-time first degree students drop out between their first and second year than at any other point.

The Scottish Funding Council ( SFC) published data on retention rates of entrants from SIMD20 areas in their 2015/16 Report on Widening Access. Chart 3 shows how the SIMD20 retention rates published by SFC compare to retention rates for entrants from other areas. The rates for SIMD20 entrants have been consistently lower over recent years (by around 5 percentage points).

Chart 3: Retention rates, full-time first degree entrants, by SIMD, starting second year of study in 2013/14 to 2015/16

Source: HESA student data

Chart 4 shows how retention rates for the same two groups of entrants varied by institution type. In each institution type retention is lower for SIMD20 entrants than for entrants from other areas and the size of the gap is similar across all types. Another key point is that retention rates are slightly lower in general for post-92 institutions. Given that a higher percentage of SIMD20 entrants go to post-92 institutions, this will also have contributed to the overall retention rate difference between the two groups of entrants.

Chart 4: Retention rates, full-time first degree en-trants, by institution type and SIMD, starting sec-ond year of study in 2013/14 to 2015/16

Source: HESA student data

Chart 5 shows how retention rates for the two groups varied by subject. In almost every subject, retention rates are lower for SIMD20 entrants than for entrants from other areas and this difference is fairly consistent across subjects.

Chart 5: Retention rates, full-time first degree en-trants, by subject and SIMD, starting second year of study in 2013/14 to 2015/16

Source: HESA student data

A similar subject-level pattern is present for each institution type which suggests entrants from SIMD20 areas are less likely to progress to their second year of study at university, even after accounting for subject and university type.

Outcomes (degree classifications)

Defining a positive or negative outcome in terms of degree classification is somewhat subjective, and every student will have their own personal aims and expectations. For the purposes of this analysis we have focussed on two outcomes - firstly, the percentage of qualifiers obtaining a degree without a classification, which is a proxy for the number of students who chose not to study for an honours degree, or did not have the opportunity to do so. And secondly, the percentage of qualifiers gaining an upper second-class honours degree (2:1) or better. We have used this threshold as many employers and postgraduate programmes require applicants to have obtained at least a 2:1.

Unclassified degrees

Undergraduate degrees which are unclassified generally fall into one of two categories. They are either degrees in subject areas such as Medicine and Dentistry, which are generally not given an honours classification, or they are ordinary degrees. In Medicine and Dentistry, Subjects Allied to Medicine and Veterinary Science, the majority of first degrees are unclassified and these subject areas accounted for around 40% of all unclassified degrees. These subject areas have therefore been excluded from the outcomes analysis.

Chart 6 shows the percentage of qualifiers who obtained unclassified degrees over recent years. The percentage with unclassified degrees is consistently higher for SIMD20 qualifiers (by up to 15 percentage points) i.e. SIMD20 qualifiers were less likely to achieve an honours degree than those from other areas. Chart 7 shows that this is also true across each institution type.

Chart 6: Percentage of unclassified degrees, full-time first degree qualifiers, by SIMD, 2013/14 to 2015/16

Source: HESA student data

Chart 7: Percentage of unclassified degrees, full-time first degree qualifiers, by institution type and SIMD, 2013/14 to 2015/16

Source: HESA student data

Another thing to note is that the percentage with no classification was generally lower at ancient universities and higher at post-92 universities, where students were four times as likely to leave with an unclassified degree (34% of qualifiers vs. 8%). This will have contributed to the overall difference between qualifiers from SIMD20 areas and other areas, as most SIMD20 qualifiers attended post-92 institutions.

Chart 8 shows that the percentage of qualifiers with no classification varied considerably by subject. In all but one subject area, qualifiers from SIMD20 areas were more likely to obtain an unclassified degree than those from other areas.

Chart 8: Percentage of unclassified degrees, full-time first degree qualifiers, by subject and SIMD, 2013/14 to 2015/16

Source: HESA student data

A similar subject-level pattern is present for each institution type which suggests qualifiers from SIMD20 areas are less likely to obtain an honours degree even after taking account of subject and university type.

2:1+ degrees

Chart 9 shows the percentage of qualifiers who obtained an upper second-class honours classification or better (2:1+) over recent years. The overall percentage of qualifiers obtaining 2:1+ degrees was consistently lower for SIMD20 entrants compared to those from other areas (by around 15 percentage points). In part, however, this is a reflection of the fact that a higher percentage of SIMD20 qualifiers leave with unclassified degrees, as shown previously.

Chart 9: Percentage of 2:1+ degrees, full-time first degree qualifiers, by SIMD, 2013/14 to 2015/16

Source: HESA student data

By including qualifiers who did not obtain a classification, the percentages shown on Chart 6 reflect opportunity (and possibly aspiration) as well as academic attainment at honours level. We are including students who chose not to study for an honours degree or did not have the opportunity to do so. Excluding qualifiers who did not obtain a classification should provide a better proxy for academic attainment at honours.

For illustration, 47% of qualifiers from SIMD20 areas obtained a 2:1+ degree in 2015/16 overall, compared to 63% of qualifiers from other areas, a 15 percentage point difference. If we look only at qualifiers who obtained honours level classifications, however, 69% of honours level qualifiers from SIMD20 areas obtained a 2:1+ degree, compared to 78% of qualifiers from other areas, a 9 percentage point difference. This suggests that the gap in the overall percentage of 2:1+ qualifiers is a reflection of both lower attainment of SIMD20 qualifiers at honours level, and the lower percentage studying at that level in the first place.

Chart 10 shows that the percentage of honours level qualifiers obtaining 2:1+ degrees varied across the different university types and that the percentage was lower for SIMD20 qualifiers in each case. More generally, the percentage of qualifiers with 2:1+ degrees was lowest at post-92 universities, where the majority of SIMD20 students studied. As before, this will have contributed to the overall difference between qualifiers from SIMD20 areas and other areas, as relatively more SIMD20 qualifiers attended post-92 institutions.

Chart 10: Percentage of 2:1+ degrees (excl. un-classified), full-time first degree qualifiers, by in-stitution type and SIMD, 2013/14 to 2015/16

Source: HESA student data

Chart 11 shows that the percentage of honours level qualifiers obtaining 2:1+ degrees also varied across subjects and was lower for SIMD20 qualifiers in almost every subject area. There is a similar pattern across subjects at each of the different university types which suggests that SIMD20 students who do study for an honours degree were less likely to get a 2:1 or better even after taking account of subject and university type.

Chart 11: Percentage of 2:1+ degrees (excl. un-classified), full-time first degree qualifiers, by sub-ject and SIMD, 2013/14 to 2015/16

Source: HESA student data

Destinations

Defining a positive or negative destination for a university graduate is also somewhat subjective and, as with degree outcomes, every student will have their own aims and expectations. For the purposes of this analysis we have used the Destination of Leavers from Higher Education ( DLHE) survey, which graduates complete six months after leaving university. We have focussed on graduates who were active in the labour market at that stage (their primary activity was work or looking for work) and calculated the percentage in professional (or managerial) level occupations ^[6] .

This means we have not considered graduates who continued into further study. The representation of students from socioeconomically disadvantaged backgrounds in postgraduate study is a (distinct) access issue in its own right.

Chart 12 shows that the percentage of those looking for work who found professional level jobs was consistently lower for SIMD20 graduates (by around 5 percentage points).

Chart 12: Percentage in professional level jobs (excl. those not looking for work), full-time first degree graduates, by SIMD, 2013/14 to 2015/16

Source: HESA destinations data

Chart 13 shows that the percentage in professional jobs did not vary dramatically by institution type but that the percentage is lower for graduates from SIMD20 areas at each university type.

Chart 13: Percentage in professional level jobs (excl. those not looking for work), full-time first degree graduates, by institution type and SIMD, 2013/14 to 2015/16

Source: HESA destinations data

Chart 14 shows that the percentage in professional jobs varied considerably by subject but is lower for graduates from SIMD20 areas for almost all subjects.

Chart 14: Percentage in professional level jobs (excl. those not looking for work), full-time first degree graduates, by subject and SIMD, 2013/14 to 2015/16

Source: HESA student data

The degree outcomes data showed that qualifiers from SIMD20 were less likely to leave with a 2:1 or better. Chart 15 shows that graduates with a 2:1 or better are more likely to have found a professional level job which suggests this is contributing to the difference between SIMD20 graduates and graduates from other areas. However, even when comparing graduates who obtained a 2:1 or better, the percentage from SIMD20 areas in professional level jobs was lower. This suggests that graduates from SIMD20 areas were less likely than other students to find a professional level job even after accounting for subject, institution or degree outcome.

Chart 15: Percentage in professional level jobs (excl. those not looking for work), full-time first degree graduates, by degree classification and SIMD, 2013/14 to 2015/16

Source: HESA destinations data

Commissioner's Commentary

Fair access is not enough. Fair outcomes are just as important. Difficult as it is to increase the proportion of entrants from the most deprived social backgrounds, it is more difficult to make sure they succeed - which means making sure they don't drop out, and they get good degrees and graduate jobs. Maybe the First Minister should set higher education another challenge - to ensure that by 2030 students from SIMD20 areas not only have the same chance of securing places in higher education but also enjoy the same outcomes as students from the most privileged social backgrounds.

That would be a tough call, as this discussion paper shows. Currently SIMD20 students have consistently lower retention rates only partly explained by the fact that they are more likely to attend post-1992 universities (which have lower retentions rates overall). Fewer of them are awarded honours degrees. Of those who get honours degrees a lower proportion gains a 'good' degree (a First or a 2:1). And it gets worse. Even those SIMD20 students with 'good' degrees, the same as their more privileged peers, are less likely to get professional jobs.

There are three possible responses to this evidence of 'under-performance'.

1. Invest in further support?

Nearly everyone agrees about the first. Universities need to invest more in supporting students from less traditional backgrounds. They also need to invest in research to identify the most important obstacles these students face. My guess is that, once initial deficits in skills and knowledge have been overcome, the support they need is just as likely to be financial - and, in a subtler sense, social and cultural - as academic.

2. Cherry pick entrants?

The second possible response is more controversial and I feel, unwelcome. Universities would only admit more SIMD20 students if they could be sure they will perform as well as traditional students. But there are some aspects of 'under-performance' that universities do not control, in particular access to the professions and other high-status (and high-pay) jobs. So this would act as a brake on fair access unacceptable in terms of social justice (and political realities).

3. Make HE culture more class-inclusive?

The third response is perhaps more controversial still. But part of the reason for 'under-performance' by SIMD20 students may be that our current definitions of academic 'performance' are deeply interwoven with largely unacknowledged assumptions about behaviour linked to class and culture. That does not mean dumbing-down or double standards should ever be tolerated. Nor is it an argument that academic standards are merely contingent, in the sense that they reflect the prevailing social order - although they cannot be regarded as fixed either, uninfluenced by the tides of history (How else can the inflation in the number of first-class degrees be explained?).

Two analogies may help. First, it is now commonplace to recognise the hidden power of gender discrimination. Everyone accepts some forms of status, achievement and performance are 'gendered'. The same is true of class and culture. Second, it is also commonplace to talk about 'institutional racism' that is so deeply entrenched it may go unrecognised. But there is a reluctance to accept the same is true of class. Maybe it challenges our easy assumptions about social mobility that focus on up-skilling, empowerment and modifying behaviour while downplaying barriers of prejudice.

These thoughts are relevant to achieving fair outcomes for SIMD20 students. However much support they receive during their time at college or university, they still suffer discrimination. Not so many enjoy the positive reinforcement of families and peers that helps stop more socially privileged students dropping out. Faced with competing social, and maybe financial pressures, they need more resilience to stay the course. Far fewer have the 'middle-class' habits, and actual social connections, that smooth the paths into professional jobs.

So higher education too has to change. I am reminded of Brecht's ironic comment after the suppression of the revolt against the Communist regime in East Berlin in 1953: "Some party hack decreed that the people had lost the government's confidence and could only regain it with redoubled effort. If that is the case, would it not be simpler, if the government simply dissolved the people and elected another?"

Do assumptions lecturers make about how 'good' students behave reflect their own, perhaps more privileged, experience at a time when access to universities was more restricted? Do courses place too much emphasis on elite knowledge, and traditional models of professional practice? Are academic regulations about progression or detailed criteria for degree classifications sufficiently flexible? I do not know the answers to these questions. But I am convinced these are questions universities, and to a lesser extent, colleges, should ask themselves.

Professor Sir Peter Scott
Commissioner for Fair Access