Abstract
How should universities evaluate scientific research? This paper critically assesses the quantitative approach to the evaluation of scientific outputs based on publication metrics. First, I provide an overview of the standard indicators, such as Impact Factor and h-index. Secondly, I show that one limitation of the metrics system is that it lacks adequate criteria to distinguish research fields that should be kept separate for evaluative purposes. Finally, I claim that this limitation negatively affects the use of such metrics. In particular, it risks to hinder the development of normal science in a Kuhnian sense in some of such fields.
