Bibliometric indicators as items

Abstract

The question of what bibliometric indicator indicate has been discussed for several decades. Over that period, the use of indicators has increased, the number of indicators too, but the question of what the indicators exactly measure remains to be debated. In this paper we propose to approach it from the perspective of scale construction. Basically, this means that we interpret the publication-based and citation-based indicators as items that measure aspects of the scientific quality, but at the same time we accept that all these indicators are characterized by error. However, several indicators together, may lead to a valid and reliable variable, representing a latent quality dimension. This approach should not be confused with composite indicators, such as deployed in university rankings.