1. This study investigates how disability, age, sex, ethnicity, nationality and early career researcher status are related to the selection of staff for inclusion in the 2014 Research Excellence Framework (REF 2014).
2. REF 2014 assessed the quality of research submitted by higher education institutions (HEIs) in the UK through a process of expert review. It replaced the Research Assessment Exercise (RAE), last conducted in 2008.
3. In 2009, HEFCE published a UK-wide equality and diversity assessment of the RAE 2008, ‘Selection of staff for inclusion in RAE 2008’ (HEFCE 2009/34). This investigated how disability, age, sex, ethnicity and nationality related to selection of staff for inclusion in the 2008 RAE.
4. REF 2014 was completed in December 2014. This study assesses the staff selected for the 2014 exercise in terms of disability, age, sex, ethnicity, nationality and early career researcher status. As with HEFCE/2009/34, the scope of our quantitative analysis is limited to assessing whether the process of selecting staff resulted in an unbiased outcome from an equality and diversity perspective, or whether some staff were disadvantaged.
5. This report forms part of a number of projects that together are intended to provide a comprehensive evaluation of REF 2014, and inform policy development for future exercises.
6. This report bases its methodology on the two previous reports on the equality and diversity of selected staff. Following the same principles as the previous reports, we considered the selection rates for different cross-sections of potentially eligible staff in the REF 2014. By using statistical models, we compared staff on a ‘like-for-like’ basis by taking into account other characteristics that may affect whether or not a member of staff is selected.
7. The scope of this quantitative analysis is therefore limited to addressing whether there were specific differences between certain groups of academics in the process of being selected for inclusion in the REF. It does not attempt to comment on the research process as a whole, the process of accepting or rejecting individual articles, or whether the REF 2014 panels assessed the work of different groups of academics consistently.
Results and discussion
8. The selection rate for staff with declared disabilities was lower than for those without. Although this discrepancy can be partially explained when other factors are taken into account, the modelling still suggests that the proportion selected for inclusion in the REF 2014 is lower for staff with a declared disability.
9. As in HEFCE 2009/34, the data shows a marked difference between the rate of selection for men and women in REF 2014; 67 per cent of men were selected, compared with 51 per cent of women. However, the proportion of women submitted has increased (from 48 per cent in RAE 2008). When age is considered in combination with sex, the model output shows that the gap between selection proportions for men and women has decreased for the most populous age group, that of staff between 30 and 60 years old.
10. While the continued under-selection of female staff probably indicates deeply entrenched supply issues, it could also be caused by maternity leave and childcare responsibilities, which in academia most often affect women between the ages of 30 and 50.
11. Analysis of male and female selection rates at HEI level shows that the majority of institutions do not have equal selection rates by sex. However, there is little evidence of a relationship between these rates and the overall percentage of staff submitted to the REF by an HEI. This suggests that a more selective submission policy does not necessarily indicate a greater bias in sex selection rates.
12. Differences in selection rates across Units of Assessment (UOAs) were also considered for male and female staff. A range of selection indices were observed, from 0.9 (for Communication, Cultural and Media Studies, Library and Information Management) to 2.4 (for Allied Health Professions, Dentistry, Nursing and Pharmacy). No UOA was found to be statistically significantly different from the sector average selection index of 1.3.
13. The effect of nationality on selection rates was considered, with selection rates being highest for European Union (EU) staff and lowest for UK nationals. Although the differences are partially explained by other factors, EU and non-EU staff still experienced a higher selection rate than UK staff.
14. The selection rates were similar for all ethnicity groups, with the exception of Black and Asian UK and non-EU nationals who had statistically significant lower selection rates, even with modelling for other factors taken into account.
15. Early career researchers (ECRs) had a selection rate of 80 per cent, significantly higher than that for non-ECRs (58 per cent). Analysis by sex shows a larger selection difference between ECRs and non-ECRs for female than for male staff, suggesting that the sex disparity in selection rates is less for individuals at the start of their research careers.
16. Analysis of staff by full-time equivalence (FTE) shows that staff with contracts at less than 1 FTE were significantly less likely to be selected for the REF 2014. This difference increases when other factors are taken into account, suggesting that these other factors have less influence on selection rates than FTE.
17. Although it is clear that there are still equality and diversity issues to be addressed in developing future REF exercises, the progress which has been made since RAE 2008 should not be overlooked.
18. Although many in the sector have noted the considerable volume of work, and the often challenging internal processes involved In addition, the strengthened measures to promote equality and diversity in the REF 2014 were widely welcomed by the sector, allowing greater consistency and a fairer approach to staff selection.
19. The work undertaken to ensure that equality considerations were taken into account in the selection of staff for the REF has been recognised as having an overwhelmingly positive impact. Moreover, this impact extends beyond the results of staff selection, to establishing equality and diversity as important considerations in universities’ everyday activities. This has given the sector a strong platform for further progress.
20. Despite the progress in some areas, the remaining equality challenges that remain have been thrown into sharp relief by this analysis. These include the continued under-selection of many black and minority ethnic staff (particularly black staff) and staff with disabilities, and the increase with age in the selection gap between men and women. The detailed analysis contained in this report will inform wider equality and diversity work in the sector, as well as being taken into account in preparations for any future REF.
21. No action is required in response to this document.