Sarah C. Williams, Angela Bonnell, and Bruce Stoffel
Print version (Adobe Reader required)
Illinois State University’s Milner Library conducted focus groups in the summer and fall of 2007 as part of its user-centered approach to implementing a federated search engine. The feedback supplemented the comments from usability testing conducted in the summer of 2006. The purpose of the focus groups was to learn about students’ use of and satisfaction with the federated search engine and to gather their ideas on how to incorporate it into the library website. The focus groups provided qualitative information that Milner Library used to guide decisions regarding website design and federated searching instruction. A list of best practices from the user perspective is also drawn from the findings. The unique aspects of this article include the use of focus groups to gather feedback on federated searching and the discussion of incorporating a federated search engine into a library website. This article is based on preliminary findings presented at the Internet Librarian 2007 conference in Monterey, California.
For more than a decade, federated searching—the ability to simultaneously search multiple online library databases or Web resources—has been one component in the arsenal of information retrieval tools available to libraries. Since its inception, several thousand libraries across the United States have started providing some form of federated searching on their websites.1 In April 2005, after nearly two years of review of the leading federated search solutions, the statewide consortium to which Milner Library at Illinois State University (ISU) belongs announced its decision to select WebFeat as its federated search provider.2 The consortial package available to member libraries presented an opportunity as well as a dilemma for our library. Illinois state budget shortfalls forced drastic cutbacks throughout all areas of the university. We canceled print subscriptions to periodicals and evaluated databases for redundancies. We were hesitant to invest limited funds and precious staff time implementing a federated search tool we were unsure patrons would use or find useful for their research, even if it was offered at a reduced rate.
By 2005, much had already been written about libraries grappling with federated searching. Case studies reported the challenges libraries faced as they implemented and customized federated search interfaces. Despite these challenges we decided to take advantage of federated searching benefits. We proceeded with the expectation that we could mitigate perceived shortfalls by customizing the product to meet our patrons’ needs. User-centered assessment was one way to determine if the product’s expense was justified.
We dedicated considerable time and energy in customizing the federated search engine, which we branded Search It. In June 2006, Search It moved off the development site onto the library homepage for a soft rollout of the tool, as seen in figure 1.
In our first attempt to assess use of Search It by students at ISU, we scheduled usability testing in August 2006. The purpose of the testing was to study ease of use for new users. Participants were first asked to perform five research scenarios and then asked six open-ended follow-up questions on their search experience.3 User comments were useful in identifying problems with descriptive language, search content and search options, and navigation from the federated searching interface to the native interfaces. Highlights from student feedback revealed that none of the participants realized the Quick Search option searched only twelve resources. To correct this confusion, we inserted the number twelve into the Quick Search subheading so it would read, “Search 12 Selected Library Resources Simultaneously.” In the usability testing, none of the participants chose to search the library catalog using the Advanced Search option. In a concerted attempt to promote the use of books in the collection and to take advantage of simultaneous searching in a variety of sources, we chose to automatically include the library catalog into most subject categories in the Advanced Search option. Some participants had difficulty finding and understanding the sort feature in the results display. We adjusted the alignment and the wording in the dropdown menu to better represent the sort options.
Revisions to the search interface and results page would be the first changes made to our customizations. Through the fall and spring semesters, use data revealed students were using the new tool, but we were interested in supplementing quantitative data with qualitative feedback from students. Following a year of use, we planned to assess the tool in focus groups.
In the summer and fall of 2007, Milner Library conducted focus groups consisting of students who identified themselves as users of Search It. Students were asked to provide feedback on their use of and satisfaction with Search It. We also gathered their ideas regarding Search It’s placement on our website.
The user-centered feedback convinced us that federated searching was a beneficial tool for student research and that it deserved a more prominent location on our website. The qualitative feedback we gathered from our focus groups fills a gap in existing research on federated searching. It supplements existing literature from federated searching usability testing that typically focuses on navigation of federated search engines. Unlike other studies, our research offers in-depth user input from focus groups on the value of this tool.
Most articles about federated searching in academic libraries have one of three focuses: why an academic library should offer federated searching, how federated searching has been introduced by academic libraries, and what users and librarians think of the tool. Much of the user satisfaction data is based on conversations with research participants in controlled settings rather than with actual users. Feedback from users has been based on surveys rather than in-depth dialogue. While numerous studies have focused on interface design and navigation, none has addressed broader questions of where and how users expect to find federated searching on university websites.
Roy Tennant’s 2001 Library Journal article was one of the first to describe the possibilities of federated searching for simplifying academic research by eliminating the need for researchers to familiarize themselves with multiple databases. While acknowledging technical challenges posed by federated searching, Tennant identifies promising early implementations, including Searchlight and WebFeat.4 In a 2003 Library Journal column, Tennant declared that federated search tools are “the correct solution for unifying access to a variety of information resources,” while recognizing challenges in designing such a tool. Chief among them were providing results in a manner that would not overwhelm or underwhelm, finding a way to display results in relevance order, and providing users ready access to full-text publications.5
Pages: 1 2 3 4 5 6 7