Czy można porównywać jabłka i gruszki? O danych bibliometrycznych w różnych dziedzinach nauki


  • Zbigniew Błocki Instytut Matematyki, Uniwersytet Jagielloński
  • Karol Życzkowski Instytut Fizyki Uniwersytet Jagielloński oraz Centrum Fizyki Teoretycznej PAN

Słowa kluczowe:

bibliometric indicators, evaluation procedure, Impact Factor, Eigenfactor Article Influence Score, inter-field comparison


Usage of biliometric data in the evaluation process of grant applications in various fields is discussed. In such a procedure bibliometric data can only be used as an auxiliary tool and should never replace the peer-review process. In order to compare bibliometric data for scientists from different fields one has to use numbers rescaled against the average values in each field. To make such a comparison easier we compute average bibliometric data using known statistics available from Web of Science and National Science Foundation. In this way average number of articles coauthored by a scientists in a given field and average number of citations obtained in one year are calculated. For instance a typical physicist is cited 16 times more often than an average mathematician and therefore his average Hirsch index could be approximatelly 4 atimes higher than in mathematics, although the detailed numbers strongly depend on the subfields of both branches. Similar statistics is obtained comparing the lists of the most cited scientists by fields, and also analyzing citations of Nobel Prize winners and Fields medalists. We discuss also the Impact Factor (IF) of a journal and compare it to the Eigenfactor Article Influence Score (AIS) which is based on the Page Rank algorithm of the Google search engine. We provide several arguments supporting the claim that the AIS index of a journal provides a more reliable estimate of the scientific value of a journal than the IF.