Garfield and the Impact Factor : the Creation , Utilization , and Validation Of

INTRODUCTION Eugene Garfield laid one of the key empirical foundations for modern information science through the innovation of the citation indexing of science. He created the impact factor as part of the process of developing the Science Citation Index (SCI) produced by his company, the Institute for Scientific Information (ISI). In determining the coverage of this index, Garfield utilized two citation measures to analyze the structure of the scientific journal system: total citations and the impact factor. Both these measures of journal importance were subsequently incorporated into the Science Citation Index (SCI) and Social Sciences Citation Index (SSCI) Journal Citation Reports (JCR) published annually by ISI. Briefly defined, the total citations measure is the total number of references in the source journals indexed by ISI during the year covered by the JCR to all issues of a given journal dating from its origin. In contrast, the impact factor is a ratio calculated by dividing the number of references in ISI source journals to the issues of a given journal published in the two years preceding the JCR year divided by the number of citable source items in these issues. The difference in time framework makes total citations a measure of the historical significance of journals and the impact factor a measure of their current significance. Of the two measures, the impact factor came to be the one most widely applied and influential. The first part of this article (Bensman 2007) comprised an intellectual biography of Garfield, which had the purpose of tracing the evolution of his ideas, mapping out the structure of his thought, and determining the place of the impact factor within this structure. This second part has a complementary purpose. It is first to translate Garfield's bibliometric concepts into those of probability and statistics. Then, using the latter concepts, the paper statistically explains and tests Garfield's findings in respect to total citations and the impact factor as measures of journal importance. The statistical tests are conducted with a sample of 120 journals considered by chemists as relevant to their discipline. Chemistry is the field, in which Garfield began his career. Four measures are used to gauge the importance of these journals: a subjective scoring constructed from the data of a 1993 survey of the Louisiana State University (LSU) chemistry faculty on their journal needs; their 1993 usage at the University of Illinois (UI) at Urbana-Champaign Chemistry Library; and …