The Use and Understanding of Keyword Searching in a University Online Catalog.

This article reports on a study of keyword and Boolean searching by users of an online catalog at the University of Oklahoma. The primary aim of the study was to discover if typical users of a keyword Boolean OPAC possess a basic understanding of how the system processes keyword and Boolean searches. A complementary objective of the study was to learn of the frequency of keyword searching on this university OPAC, as compared to the traditional author, title, and subject heading catalog searches. Two data collection methods were employed: a survey questionnaire and computer transaction logging software. The key findings of the study were: users of this online catalog search more often by keyword than any other type of search, their keyword searches fail more often than not, and a majority of users do not understand how the system processes their keyword searches. In addition, it was found that users are not fond of reading printed instructions on-screen or off-screen. Users also expressed a strong interest in search assistance from the system while conducting their searches. Perhaps we have made our second-generation keyword and Boolean OPACs (Online Public Access Catalogs) deceptively and dangerously easy to use. The warnings have been there for a number of years. Since the earliest days of OPAC use and user studies, both librarians and researchers have expressed concern over the high degree of user satisfaction with OPACs that do not perform well, especially when processing subject searches. The danger of today's easy-to-use, implicit Boolean OPACs lies in the fact that they can and are being used heavily by legions of searchers who apparently have little understanding of how these bibliographic retrieval systems process the keyword and Boolean searches submitted to them. Is such knowledge a critical factor in successful searching of OPACs? Researchers think it is. Borgman has given us the most comprehensive analysis to date of the complex nature of searching in online catalogs and the kinds of knowledge required to search effectively in these retrieval systems.[1] Effective searching by keyword requires knowledge of the type and content of the index or indexes targeted in any given keyword search. Formulating effective Boolean queries requires conceptual knowledge of how search terms may be combined and semantic knowledge of how a particular system processes Boolean queries. A low level of this conceptual and semantic knowledge has been shown to be a major contributing factor in poor search performance.[2] This article reports on a study conducted at a major U.S. university that sought to discover if typical users of a keyword Boolean OPAC possess a basic understanding of how the system processes keyword and Boolean searches. A complementary objective of the study was to learn of the frequency of keyword searching on this university OPAC, as compared to the traditional author, title, and subject heading catalog searches. OPAC User Search Behavior and Experience In this second decade of the OPAC one would be hard pressed to find an OPAC in use that did not allow keyword searching and at least some rudimentary form of Boolean retrieval. Recent analyses and surveys of OPACs, published by Cherry, et al., Tillotson, and Akeroyd,[3] report the widespread availability of keyword searching. Many OPAC researchers have promoted this form of access as the best solution to the well-documented problems users experience with subject searching in OPACs.[4] Many studies of end-users of OPACs and other information retrieval (JR) systems were published between 1985 and 1995. Reviews of this literature may be found in Larson, Hildreth, Connell, Thorne and Whitlatch, Nahl and Tenopir, Spink, and Walker and Hancock-Beaulieu.[5] These studies have enlarged our knowledge of user search behavior, needs, and difficulties experienced in the use of second-generation retrieval systems. …