Abstract

This paper argues that the bibliometric methods that should be used to evaluate social science research differ from those that are used to evaluate scientific research. The argument is based on a review of several decades of research into the quantitative characteristics of the social science literature. Unfortunately, evaluation of social science research if done properly is more complex and expensive than evaluation of science. This is because social scientists publish in a broader range of literature and build their work on a broader base both topically and over a longer period of time than do scientists. A thorough understanding of the nature of social science literature will enable sound evaluations to be designed.