Some observations on the use of bibliometric indicators in the assignment of university chairs

Abstract

In their article, “Can the Assignment of University Chairs Be Automated?”,Nicolini, Vakula, Balla, andGandini describe the results of their initial attempts at using multiple bibliometric indicators in order to eventually automate the assignment of University chairs at the full and associate professor levels. The indicators utilized consist of each candidate's age, years of scientific activity, number of published articles, citation rate and the quality and type of publishing and citing journals. Data concerning these indicators are obtained from both SCI databases and the curriculum vitae of 76 test-candidates. Although the ranking of candidates is shown to be affected by both subfield differences in citation patterns and the weighting factor assigned to the bibliometric indicators tested, some predictive validity is found between the use bibliometric indicators and the independent peer-review of candidates. While these results are encouraging, the authors readily acknowledge the limitations of their efforts and the need to further refine the use of bibliometric indicators, before their employment in the automated assignment of University chairs.

My observations concerning the work ofNicolini et al. will take the following form. First, I wish to comment briefly on what I view to be the philosophy behind the efforts ofNicolini et al. Second, I want to mention several positive and potentially negative procedural issues associated with the proposed use of bibliometric indicators in assessing individual scientific accomplishments. And finally, I would like to comment on what I see as some broader, latent consequences, potentially associated with the use of bibliometric measures in the automated assignment of University chairs.