We use a simple, sociophysical, agent-based model to gain insight into the relationship between the strength of academic research groups and the researchers they contain and to investigate the optimal size of research groups in a number of academic disciplines. The model suggests a linear relationship between group research quality and group quantity up to an upper critical mass. It further suggests that above this size, research quality no longer increases significantly with group quantity due to a tendency for large research groups to fragment. We use the results of the UK's Research Assessment Exercise to establish empirical evidence that, although the best research groups tend to be large, they frequently tend to have characteristics more akin to small and medium groups. This may be interpreted as indicating that the optimal size of research groups is slightly in excess of the upper critical mass. We speculate that a strong degree of cohesiveness lies behind the success of such groups. The principle managerial message coming from this analysis is that emulation of such success involves optimization of the quality and quantity of communication links between research group members.
Bibliographical noteThe full text is currently unavailable on the repository.
- research quality
- optimal size of research groups
- research evaluation
- Ringelmann effect