Correlating Libcitations and Citations in the Humanities with WorldCat and Scopus Data

The term libcitations was introduced by White et al. (2009) as a name for counts of libraries that have acquired a given book. Somewhat like citations, these library holdings counts, which vary greatly, can be taken as indicators of the book’s cultural impact. Torres-Salinas and Moed (2009) independently proposed the same measure under the name catalog inclusions. Both articles sought an altmetric for authors of books in, e.g., the humanities, since the major citation indexes, oriented toward scientific papers, have not served them well. Here, using very large samples, we explore the libcitation-citation relationship for the same books by correlating their holdings counts from OCLC’s WorldCat with their citation counts from Elsevier’s Scopus. For books cited in two broad fields of the humanities during 1996-2000 and 2007-2011, we obtain positive, weak, but highly significant correlations. These largely persist when books are divided by main Dewey class. The overall results are inconclusive, however, because the Scopus citation counts for the books tend to be very low. Further correlational research should probably use the much higher book citation counts from Google Scholar. Nevertheless, a qualitative analysis of widely held and widely cited books clarifies the libcitation measure and helps to justify it. Conference Topic Indicators Introduction Journal-oriented scientists have long had citation counts as an indicator of the impact of their articles, and journal-based citation indexes cater to them. But the same indexes cover citations to books less well, and book-oriented scholars in the humanities and softer social sciences feel themselves at a disadvantage, especially if citation measures are going to be used in performance evaluations and funding decisions (see Kousha, Thelwall, & Rezaie 2011 for a review). White et al. (2009) responded to this lack by proposing that one measure of a book’s cultural impact could be the number of libraries that hold it. The idea behind this altmetric was that librarians who acquire a book are somewhat like scholars who cite it, in that both acts involve assessment and choice on behalf of communities of readers. To bring out the parallel, White et al. called the librarians’ formal act of acquisition a libcitation (first syllable as in “library”). They wrote that the libcitation count (also known as a library holdings count) for a particular book “increases by 1 every time a different library reports acquiring that book in a national or an international union catalog. Readers are invited to think of union catalogs in a new way: as ‘librarians’ citation indexes’” (p. 1084). OCLC’s WorldCat was mentioned as a prime example of a union catalog—that is, one that pools the cataloging records of OCLC member libraries and reports how many of them hold each cataloged item. At the same time and wholly independently, Torres-Salinas and Moed (2009) made an identical proposal. Their name for libcitations (our term here) was catalog inclusions, and they, too, stressed the parallel between such inclusions and citations to journal articles (p. 11). They, too, named WorldCat as a potential source of library holdings data. Moreover, both