The subject of Information Systems is included in the descriptions of two RAE units of assessment: UoA 25 Computer Science (CS) and UoA 61 Library and Information Management. In 2001, these two panels convened a joint sub-panel to consider submissions in this area.
Although Portsmouth made no RAE submission to either CS or LIM in 2001, there has been continued active research in Information Systems. This includes the work of the Healthcare Computing group, three members of which were part of the Subjects Allied to Medicine 11a submission that got a grade 3b in 2001.
The standard to reach to obtain a grade 4 is "Quality that equates to attainable levels of national excellence in virtually all of the research activity submitted, showing some evidence of international excellence."[1] In practice, this meant that fewer than 10% of staff were not of "international" or "national" excellence. In their overview report[2], the LIM panel said that in assessing whether work was of international or national excellence, they "took account of the extent to which the work advanced the subject, and increased understanding and knowledge. Consideration was given to the originality, impact, range, accuracy and clarity of the work. Originality was taken to include the use of new sources of data or significant re-interpretation. Work deemed to be of international excellence was outstanding in terms of almost all these qualities." The panel noted, however, that "at international level, there was no shortage of international conference proceedings returned but the overall quality was disappointingly low and many showed little evidence of research." In other words, for "international" excellence, it is not sufficient for the journal or conference to have "international" in its title - the work described must be of international standard. A good test of whether international excellence is to ask whether someone in Australia would have heard of you and your work.[3]
By 2006, we expect at least 6 staff to meet RAE grade 4 quality thresholds. These on their own would constitute a small but viable submission. However, depending on University strategy, there are active researchers from other areas who could be included in LIM, increasing the volume of research and strengthening the case. These could be in other departments both within and outside the Faculty of Technology.
How big a submission do we need? In 2001, the smallest submission that obtained a grade 4 or better was from Napier University. They got a 4 with only 6 staff submitted. It is also not necessary to have a very high proportion of staff submitted. Napier submitted 37.5% of their eligible staff; Leeds Metropolitan got a 4 with only 34.35% of their staff (9 out of 26.2 FTE). With approximately 50 staff in the three departments, our proportion would be smaller, but there may be tactics we can use at submission time to counter this.
The basic standard for staff to be submitted is that they have at least one publication that provides evidence that their research meets the standards for at least "national" excellence. In the 2001 RAE, nearly all staff who were submitted had the maximum four publications, but the LIM panel commented that it was not compulsory to return four publications and they would have preferred that some outputs had not been submitted. The panel also stated that there would be no discrimination between different forms of output. It is interesting to note, however, that while to the UoA as a whole 59% of outputs submitted were journal articles, 21% were conference proceedings and 9% were chapters in books, among grade 5* and 5 institutions, the proportions were 82%, 10% and 3% respectively, suggesting that panel found the best research predominantly in journals.
In 2001, the amount of research income per member of staff varied enormously across institutions. While the big boys (Sheffield and Salford who got 5*) averaged £105k per member of staff, a grade 5 could be obtained with as little at £39k (Brunel) and a grade 4 with as little as £17k (UCL). Interestingly, Glasgow and Aberystwyth each earned about £115k per member of staff and still only got a 3a. Of universities most like us, Napier got most of its funding from industry; Leeds Metropolitan got most of it from the research councils; Manchester Metropolitan (4) and Brighton (3b) got most of theirs from central government and the EU.
Much research funding of course goes on employing research staff. The average number of research assistants per member of staff was 0.32 across the LIM UoA, increasing only to 0.36 among grade 4 and above institutions. Few institutions reported many technical or support staff, indeed there were only a total of 22.7 in all institutions put together.
In terms of research student numbers, we are already in a healthy position with 9 completions since 2001. Overall, the number of PhD completions per member of staff submitted (CPS) was 0.66. However, while 5* and 5 institutions had 1.43 and 1.05 CPS, grade 4 institutions averaged only 0.44 and grade 3a 0.49. Indeed, Napier got their 4 despite having no PhD completions at all during the census period!
The LIM panel also based its ratings on its assessment of an institution's research culture, evidenced by a coherent, well articulated strategy, vision and research investment.
A brief note on why we are targeting LIM rather than CS. CS is dominated by big departments in "old" universities. In 2001 LIM submissions were more from "new universities", and the LIM panel overview report explicitly mentioned that they had performed better than in the previous RAE in 1996. Also the indicators for a grade 4 submission in CS are much higher than in LIM: £128k income, 0.57 research staff and 1.14 PhD completions per member of staff.
Of course, the next RAE will be played to different rules than the last one. The following are some of the possible changes and how they might affect our prospects.
[2] Judith Elkin, LIM Panel Overview Report, http://www.hero.ac.uk/rae/overview/docs/UoA61.pdf
[3] Graham Moon, University Director of Research, at a meeting in November 2002.
Updated to