Collaboration and productivity: an investigation into "Scientometrics" and "UHasselt" repository

Abstract

In this paper we investigate the following problem: for a fixed field or institute, can we prove that, the higher the number of papers of an author (calculated in the total way), the higher his/her fraction of co-authored papers (with at least one co-author) ? This is one of the possible formulations of the relation between collaboration and production. We investigate two different data sets: all articles in Scientometrics (from Volume 1 onwards, up to the end of 2007) and the institutional repository of the Universiteit Hasselt (UHasselt), containing 7,604 articles (Feb. 27, 2008). In the Scientometrics case we cannot prove the asserted relation while in the UHasselt case we can prove that high productivity leads to high fractions of co-authored papers (but low productivity can have low or high fractions of co-authored papers). Since the majority of the papers in the UHasselt repository is in the sciences while Scientometrics is more in the social sciences (certainly in the earlier years) this might be one explanation of the different findings. We also show that collaboration in the case of the UHasselt repository is much higher than in the Scientometrics case. We find that a determining factor of collaboration is the scientific environment and we present a production collaboration graph on which we can base ourselves for comparisons between fields and/or institutes.