The nursing research assessment exercise 2001: An analysis

https://doi.org/10.1016/j.cein.2005.09.001Get rights and content

Summary

The 2001 UK research assessment exercise (RAE) was explored for units of assessment (UoA) 10 (Nursing) and 11 (Other Studies and Professions Allied to Medicine). In nursing funding was the best predictor of high RAE ratings. Other factors that appear relevant are numbers of doctorates awarded and publications in prestigious journals and in those with high impact factors (IF). UoA 11 showed similar but not identical trends. Specific funding sources were more important than overall funds in UoA 11, and the journals with high ratings in UoA 10 did not necessarily score highly in UoA 11. Further the diversity of journals was higher in UoA 11.

Section snippets

Background

The Higher Education Funding Council for England (HEFCE) distributes public money for teaching and research to universities and colleges. It employs the research assessment exercise (RAE) to allocate funds. The RAE is the main measure of UK research reputation and a major source of funding. In 2002–3 the total funding for research from HEFCE (which was distributed based on RAE rankings) was £910 million, and in 2005–6 is predicted to be £1,251 million (figures obtained from www.rae.ac.uk and //www.hefce.ac.uk/

Methodology

The RAE databases for UoAs 10 and 11 were downloaded from the HEFCE HERO website. The tables in each database were merged into an SPSS datafile.

For both UoA 10 and 11 data were checked and obvious errors corrected. Most data cleaning consisted of making journal titles consistent. Common problems were presence or absence of full stops (periods), use of definite article or not, double spaces, inconsistent use of “and” and ampersand, and occasional typographical errors, as seen in other

Results of UoA 10

In the 2001 RAE submissions were rated in seven bands: 1, 2, 3b, 3a, 4, 5 and 5. A rating of one indicates research work at less than national level, and a 5/5 were the top grades where all or almost all work was of international level.

Nursing was the lowest rated UoA in the 2001 RAE, as it was in 1992 and 1996 (Anthony, 1997, Robinson et al., 2002). Only four institutes gained a five, and none gained a five star, see Table 1.

Data submitted for evaluation consisted of research activity

Publication type

Mean (average) RAE ratings were computed within SPSS, on a seven point scale where 1 = 1, 2 = 2, 3b = 3, 3a = 4, 4 = 5, 5 = 6 and 5 = 7. Journal articles scored higher than authored books, chapters in books or conference contributions, though lower than edited books, see Table 3. While internet publications via subscription scored most highly, there were only nine of them, and they were all from the Cochrane Library, and seven were from one institute, the University of York.

Journal publications

A maximum of four papers were allowed to be entered in the 2001 RAE. In some cases an academic may not have four papers in top rated journals, and they may elect to enter fewer, or put in less prestigious or less obviously research based items. Most (n = 2206, 87.5%) submissions were in journals, and most of the rest (7.5%) were books or chapters in books. 542 journals were entered, but of these 316 (14.3% of submissions) were only entered once. 50% of submissions were from 39 journals. The most

Comparison with UoA 11

UoA 11 covered a wide range of professions, biomedical sciences; nutrition; optometry; radiography; occupational therapy; physiotherapy; speech and language therapy; art, music and drama therapy; health promotion; and other studies and professions allied to medicine.

UoA 11 showed different predictors using regression analysis, with specific funding sources rather than total funding being significant, see Table 8. UK charities and research students were positive and conference articles and EU

Limitations

After accounting for quantitative measures new universities scored lower than would have been expected in social policy (McKay, 2003). Measures of esteem may be the explanation, but were not analysed in UoA 10 or 11 for reasons given in the text. Membership on the panel may be a factor, indeed all the five rated institutes in UoA 10 and none of the 1,2 or 3b scores were represented in the panel (Robinson et al., 2002) in the 2001 RAE, though as Robinson et al state “Cause and effect are

Discussion

In UoA 10 funding appeared to be the most reliable predictor for high ratings. Doctorates awarded was correlated significantly with RAE rating. The fact that this variable does not remain in the regression equation may be due to the fact the funding produces more doctoral students. However in UoA 11 the type of funding and the number of doctoral students were the most predictive factors.

Journal publications had higher ratings than conference proceedings or chapters in books in both UoA 10 and

References (15)

  • Anonymous

    How the RAE works... research assessment exercise

    Nursing Standard

    (2002)
  • Anonymous

    Payout ends dispute over RAE ranking

    Nursing Standard

    (2002)
  • D.M. Anthony

    A review of statistical methods in the Journal of Advanced Nursing

    Journal of Advanced Nursing

    (1996)
  • D.M. Anthony

    Funding research in nursing: results of the RAE... Research Assessment Exercise

    Nursing Standard

    (1997)
  • D.M. Anthony

    Understanding advanced statistics: A guide for nurses and health care researchers

    (1999)
  • J. Geary et al.

    Journal rankings in Business and Management and the 2001 Research Assessment Exercise in the UK

    British journal of management

    (2004)
  • B. Lindsay et al.

    The ’invisible’ nursing RAE 2001 – an analysis

    Nurse Researcher

    (2003)
There are more references available in the full text version of this article.

Cited by (0)

View full text