Differences in Degree Outcomes

New from HEFCE this week, a report on “Differences in Degree Outcomes:the Effect of Subject and Student Characteristics“, which looks at the outcomes of students who graduated in 2013-14. Some of this data I have previously reported when looking at HESA data on the impact of ethnicity on degree outcomes for the previous year.

The results of the HEFCE survey are not startling – they almost reinforce things that we already know in terms of what factors have an impact on achievement: the challenge now is to learn how to address each of these, and with the recent comments by the new universities minister on widening participation, and our own commitment to supporting a diverse population of students then awareness of these trends and how we then tackle them will be crucial for success of individuals and of the institution.

HEFCE considered the following variables when looking at the differences in outcomes:

  • age
  • disability status
  • ethnicity
  • The Participation of Local Areas measure (important for high WP populations)
  • sex
  • subject of study
  • prior attainment (in terms of qualifications held on entry to higher education)
  • previous school type
  • institution attended

The interesting part of the analysis is not the differences in outcomes that can be seen, but how much these differences can or cannot be explained by the influence of other factors.

Subject

Certain subjects are more likely to award 1sts/2(i)s, and the table below represents those subject we offer at Staffordshire – it will be interesting to compare our recent results with those for the sector by subject.

Subject % first or upper second % first
Subjects allied to medicine 69% 24%
Biological sciences 70% 18%
Physical sciences 73% 25%
Mathematical sciences 73% 35%
Computer science 66% 28%
Engineering and technology 74% 30%
Social studies 73% 16%
Law 69% 12%
Business and administrative studies 71% 21%
Mass communication and documentation 75% 15%
Historical and philosophical studies 82% 19%
Creative arts and design 72% 21%
Education 68% 18%
Combined 60% 16%

I always thought it was apocryphal that law didn’t award firsts – across the sector it would appear to be true!

Entry Tariff

On entry tariff, there is a clear relationship – higher entry leads to higher numbers of good degrees, which can also be seen when looking at league table data. This is one of the reasons that the Guardian league table uses a “value added” measure which seeks to adjust for entry tariff..

hefce1st1

 

Mode of Study

In general, part time students have worse outcomes compared with full time. Even adjusting for variations on entry tariff, part time students have worse outcomes than full time.

Age

The raw data shows that young students are 11 percentage points more likely to gain a good degree compared with mature entrants.

Gender

Across all entry tariffs, women are more likely to gain good degrees than men.

Disability

Graduates with a disability are slightly less likely to gain a good degree than those without a declared disability.

Ethnicity

This is the area with the biggest gap. 76% of white students gain a good degree, compared to 60% of black and minority ethnic students.

Even allowing for other factors, the unexplained gap is still equivalent to 15%.

Previous School

In most cases students from state schools outperform those from independent schools.

Neighbourhood HE Participation

Students coming from neighbourhoods with the highest rates of HE participation also gain the highest numbers of good degrees.

Implications

The recent speech by Jo Johnson referred to the importance of universities in driving social mobility and the sector’s work in widening participation.

This data provides further information that could be used to justify the costs of supporting WP in universities, and for focusing on trying to close gaps in attainment.

Much focus is given to looking at the data provided by UCAS but to understand how well the sector and individual universities are performing in terms of closing these gaps, then much fuller datasets need to be considered, taking into account retention and progression and ultimately employment – even if all our students gain the degrees they deserve, but still fail to progress into appropriate graduate roles, then social mobility isn’t realisable for everyone.

As we move into a potential quality regime that could be metrics based, together with a Teaching Excellence Framework, which will certainly use a variety of metrics (possibly including learning gain), then there will be plenty of work to be done in generating data and analysing it..

However, the focus also has to go beyond analysing data. How can we use it to understand our students both as individuals and as cohorts? How can we use data to support our staff better in teaching and assessing their students? Finally, how can we learn to change practices and behaviours based on evidence?