How the tables work

The tables are not available on the database. II How your subject did. III-IV How your university did at a glance. VIII-XV How your department did. XVI THES League table of excellence

The 1996 Research Assessment Exercise is the fourth in series aimed at providing the funding councils with the data necessary to fund research selectively. The first two, in 1986 and 1989, were confined to the old university sector and were conducted by the University Grants Committee and the Universities Funding Council. The 1992 and 1996 exercises, conducted by the Higher Education Funding Council for England on behalf of all the UK funding bodies, have incorporated the former polytechnics and colleges of higher education which are now within the university sector.

The 1996 exercise was the largest yet, with 191 institutions making submissions covering 2,894 units of assessment. The work of around 55,000 researchers employed in British academic institutions has been assessed on a peer-review basis by 59 panels of experts, covering 69 subject areas.

The census date for the review was March 31 1996. It was left to the universities being assessed to decide which members of staff in post on the census day would be included in the assessment.

Submitting departments were asked to put forward:

* Up to four pieces of work published between January 1 1990 and the census date (for humanities and arts subjects) or between January 1 1992 and the census date (for other subjects) for each assessed member of staff * Details of research students and studentships * Amount and sources of external funding * Statement of research plans The outcomes are presented in four forms in this supplement.

Pages viii to xv show the individual scores for each of the 2,894 subject submissions.

These are arranged as follows: Column 1 : 1996 Rating

* A rating of 5* in this column shows that the research assessed is of international excellence in a majority of the sub-areas within the submitting department, and at least of national excellence in all others.

* 5 shows international excellence in some sub-areas and national excellence in virtually all others * 4 shows national excellence in virtually all sub-areas * 3A shows national excellence in a substantial majority of sub-areas * 3B shows national excellence in a majority of sub-areas * 2 shows national excellence in up to half of sub-areas * 1 shows national excellence in no, or virtually no, sub-areas Column 2: 1992 Rating.

These are shown according to the five-point scale used in the 1992 exercise.

These scores are equivalent to the 1996 outcomes, with the difference that the 3 and 5 categories were not divided in 1992. It has not been possible to provide 1992 scores in all cases. Where no submission was made in 1992, this is indicated by a symbol. Where a submission was made, but a direct comparison is not possible, this is indicated by a symbol. In 1992 some submissions received separate assessments for basic and applied work. These are indicated with a colon in the 1992 ranking column with the basic score to the left and the applied score to the right of the colon.

Column 3 : Proportion of staff selected

"A" indicates that between 95 and 100 per cent of the department's academic staff were included in the research submission "B" 80-94 per cent "C" 60-79 per cent "D" 40-59 per cent "E" 20-39 per cent "F" less than 20 per cent Column 4 : This shows the full-time equivalent number of staff assessed Pages iii to vii provide a graphic representation of the outcome for each institution. Explanatory notes are provided on page iii.

Pages ii and xvi carry the league tables of overall performance by subject and by each institution. They are ranked according to average weighted scores per assessed member of staff, worked out as follows : The score is a weighted average taking account not only of the ratings given in the exercise, but the size of departments. Each individual rating is multiplied by the number of research active staff in the department being rated.

Outcomes were weighted on a 1 to 7 scale as follows A 1 rating is worth one point; a 2 two points; a 3b three points; a 3a four points; a 4 five points; a 5 six points; a 5* seven points.

Thus a department with 12 research active staff and a 5 rating totals 72 points. All the totals for the institution or subject are added up, then divided by the total number of research staff to arrive at the average. Thus an institution with three departments, one with 12 research active staff and a rating of 5, one with nine research staff assessed at 5* and one with seven research staff assessed at 2 would have a weighted average of 5.32 arrived at as follows: (12x6) + (9x7) + (7x2) = 149 / (12+9+7) = 5.32 The league table on this page for subjects is as follows : Column 1 : Number of staff assessed under this subject heading across all insitutions Column 2 : Average weighted score per member of staff The league table on page xvi for institutions is as follows : Column 1 : Rank order for universities. Membership of the Committee of Vice Chancellors and Principals has been used as the criterion to define a university.These institutions arebeen listed in bold.

Column 2 : Total number of staff assessed Column 3 : Percentage of total academic staff this represents Column 4 : Average weighted score per member of staff