Scientometrics

, Volume 26, Issue 1, pp 21–35

Reference standards for citation based assessments

Authors

  • A. Schubert
    • Information Science and Scientometrics Research Unit (ISSRU)Library of the Hungarian Academy of Sciences
  • T. Braun
    • Information Science and Scientometrics Research Unit (ISSRU)Library of the Hungarian Academy of Sciences
Invited Papers

DOI: 10.1007/BF02016790

Cite this article as:
Schubert, A. & Braun, T. Scientometrics (1993) 26: 21. doi:10.1007/BF02016790

Abstract

One of the most crucial points of citation-based assessments is to find proper reference standards to which the otherwise meaningless plain citation counts can be compared. Using such standards, mere absolute numbers can be turned into relative indicators, suitable for cross-national and cross-field comparisons. In the present study, three possible choice of reference standards for citation assessments are discussed. Citation rates of publications under study can be compared to the average citation rates of the papers of the publishing journals to result inRelative Citation Rate (RCR), an indicator successfully used in several comparative scientometric analyses (see, e.g. Refs 1–5). A more “customized” reference set is defined by therelated records in the new CD Edition of theScience Citation Index database. Using the socalled “bibliographic coupling” technique, a set of papers with a high measure of similarity in their list of references is assigned to every single paper of the database. Beside of being an excellent retrieval tool, related records provide a suitable reference set to assess the relative standing of a given set of papers as measured by citation indicators. The third choice introduced in this study is specifically designed for assessing journals. For this purpose, the set of journals cited by the journal in question seems to be a useful basis to compare with. The pros and cons of the three choices are discussed and several examples are given.

Copyright information

© Akadémiai Kiadó 1993