Faculty Relevance Criteria: Internalized User Needs

REFERENCE LIBRARIANS, ONLINE SEARCHERS, system designers, and other information professionals work to incorporate user-based relevance criteria into information services and systems. Genuinely utilizing the relevance criteria that patrons employ requires, as a first step, the development of an in-depth understanding of those criteria. This study progresses toward that understanding by providing new data on the criteria used by members of a rarely studied interdisciplinary population and by developing a usercentered methodology. Each of five Women’s Studies faculty members was interviewed concerning her immediate reactions to different resources provided in direct response to real, on-going information needs. The criteria identified by this approach went beyond topicality, currency, and other basic aspects of relevance criteria to include elements such as interdisciplinarity and theoretical perspective. INTRODUCTION The concept of relevance remains central to library and information studies work, the gold standard by which services and systems are judged. Those that put relevant information into patrons’ hands succeed; all others fail to some extent. The practical aspects of relevance work may be loosely divided into two approaches, system-centered and user-centered. The system-centered approach seeks to develop the competence of an external agency (such as a database or search engine) that can recognize relevant information by its similarities to the semantic representation of an information need. A series of OPAC searches, for example, should retrieve relLynn Westbrook, School of Library and Information Studies, Texas Woman’s University, P.O. Box 425438, Denton, TX 76204-3438 LIBMYTRENDS, Vol. 50, No. 2, Fall 2001, pp. 197-206 02001 The Board of Trustees, University of Illinois 198 LIBRARY TRENDS/FALL 2 0 0 1 evant items on the basis of semantic clues in the search statement (e.g., words used, context of word use, and word relationships) by placing them in relationship to the document representation (i.e., the cataloging record). The system-centered work builds on an assumption that the information embodied in a document or its surrogate can be objectively, if somewhat imprecisely, matched to the actual information need as articulated in a structured query (Mizzaro, 1997, p. 812). This approach leads to developments in document representation and information retrieval (e.g., Saracevic, 1969; Bookstein, 1979; Janes, 1991). The user-centered approach to relevance work holds that such a complex, highly personalized array of factors comprise relevance for a single individual in a certain situation at any given point in time-that no outsider can accurately judge relevance. Reference librarians often observe this phenomenon in action when users reject a number of “relevant” items in favor of what appears to be an irrelevant item. This user-centered approach leads to a taxonomy of user-generated criteria and an enhanced understanding of the elements involved in making relevance judgments (e.g., Cool, Belkin, & Kantor, 1993; Park, 1993; Barry, 1994). Reference librarians who understand the full range of possible relevance criteria can effectively serve their patrons. Both approaches strengthen the effort to link individuals with the information they need.