The use of an algorithmic approach for the assessment of research quality

Abstract

Recent years have seen a growing interest in the use of quantitative parameters for assessing the quality of research carried out at universities. In the UK, university departments are now subject to regular investigations of their research standing. As part of these investigations, a considerable amount of quantitative (as well as qualitative) information is collected from each department. This is made available to the panels appointed to assess research quality in each subject area. One question that has been raised is whether the data can be combined in some way to provide an index which can help guide the panels' deliberations. This question is looked at here via a detailed examination of the returns from four universities for the most recent (1992) research assessment exercise. The results suggest that attempts to derive an algorithm are only likely to be helpful for a limited range of subjects.