In keeping with its commitment that the new web site be grounded in user-centered design, the Web Advisory Group conducted focus groups with students, faculty, and staff members.  The results provide preliminary input on users' perceptions and preferences for WAG members to consider along with forthcoming data from surveys and usability tests, in making informed decisions as the redesign process moves forward.

This report covers the focus group goals, participants, summary results, and suggestions for next steps. Separate documents, not posted on the web, include compiled comments from all the sessions, detailed notes on each session, and information on the focus group process.

The goals of these focus groups were to get qualitative information and insights on:

• What web sites people in our target groups use most often for research.
• What features they like in research-related web sites.
• What they would ideally like to see and be able to do on the library web site.
• How they perceive and use the library web site now.
• Their likes and dislikes regarding the library web site.
• Their suggestions for improving the library web site.
• … plus incidental information on terms they use, perceptions of the library they have, etc. that might be different from ours.

We conducted three focus groups in March and April 2003, with a total of 19 participants. The first group included 11 library staff from RRC, Instructional Services, several subject specialty libraries, and Technical Services. The second group included 2 upper division undergraduates and 2 graduate students. The third group included 4 members of the Academic Senate Library Committee. Subject interests of students and faculty participants were Public Health, German, Film Studies, Political Science, Biostatistics, Spanish, and Engineering.

Several themes emerged from these sessions. Participants consistently expressed their desire for a site that is rich in content, that integrates multiple resources into a single search and/or menu system, and that adapts to the needs of users at different skill levels or in different disciplines. They also placed a high value on logical organization, consistent navigation, uncluttered design, and understandable terminology. Their heartfelt and oft-repeated advice to us was "simplify, simplify, simplify."

Despite our efforts to define the scope of the discussions, participants tended to conflate all Internet sources, often not making distinctions among the Library web site, catalogs, databases, etc. This is not a mistake on their part; it is how they perceive the online research process. They also envisioned some features that may be technically beyond what we can do at the present time. This report includes their comments without making judgments on relevance or feasibility.

Comments varied based on both the subject and the library expertise of the participant.

Undergraduates tended to seek specific items identified by their professors, or absent this, often started with Google. One undergraduate, whose work is topical, often begins with SFGate.

Graduate students and professors were more likely to try to identify the experts and core materials on a topic. The latter process often involved a process to identify seminal works (on a new subject) and/or to identify resources with extensive bibliographies to provide clues. Some graduates and faculty reported heavy use of professional society web pages or other web sites developed by persons knowledgeable in the field (including the library’s subject-related page). Several mentioned the importance of listservs and newgroups for staying current with both who is working in the field and for staying apprised of new developments and thinking.

Several experienced scholars reported using Google-type web searches not so much to identify the center of discussion on a topic, but to identify "windfall apples" and "atypical responses". One professor reported routinely using Google, "to get an idea of what students may have seen."

Participants placed more emphasis on a site’s content than on its organization, navigation, functionality, and visual design. Users want sites that have breadth and depth of content with minimal link redundancy. They commented that if a site has too many pages that are out of date, they stop going there. Participants had often identified sites rich in their subject and had these pages permanently bookmarked (e.g., Harvard’s catalog, a particular subject-based site at Stanford, subject-specific encyclopedias) bypassing the UC Berkeley Library’s web site altogether. Some of our subject specialty library pages were themselves identified as rich resources, and were used as the primary entry point to library resources in preference to the library’s home page.

Some specific content items that participants look for or would find useful on the library site: open hours for buildings and services; location of various libraries; abstracts; online tables of contents, indexes, prefaces, and other tools to help them determine remotely whether an item was actually relevant to their search; databases of journal articles; location of the nearest copy machine; staff directory information; easy access to library policies; easy access from the library’s site to the campus homepage (on the other hand, one user at least uses the search box on the campus web to connect to MELVYL); help, especially context-sensitive help; and tutorial information to guide the new user through the basics (e.g., finding a book).

Participants were very interested in a site that could provide them with integrated subject access to materials in their fields, including "key sources," survey papers, and specialized encyclopedias. Participants want access to applicable resources whether or not they are traditional library materials, e.g., wondering why the library site doesn’t connect to the "statistics data resources on campus?" One student expressed a desire for more electronic reserve materials, to "eliminate need to buy readers."

Participants both described features that they found useful on existing web sites, and also brainstormed those features that would make the library’s site a really excellent tool. Some of these suggestions might be addressed with content, while some of them require underlying technology – hence the term used here, features.

• pre-sifted, cued by where the user is and what s/he is doing, or
• subject to being sifted, allowing the user to apply search terms within "help" in order to focus in on the exact help they want. (as in MSN.com)

The general characteristics attributed to desirable sites included: well-organized; clear; limited set of options; not repetitive; organized as a user rather than a librarian would; same organizational principles consistent throughout the site.

Comments indicated that participants find the organization of the library’s home page as it now stands to be confusing. Participants asked about the grouping of links, suggesting that we consider a more logical arrangement, even alphabetical order. The apparent duplication between the two menus, "Libraries and Collections" and "Internet Resources by Academic Discipline" also drew criticism. Participants suggested reducing the number of links on the home page by grouping them behind labels that made sense to the user. Three suggestions were made for top-level organizational schemes:

• Collections or Research; Services
• Collections; Services; Search the Internet
• Bibliographies & Catalog; Library Locations & Policies; Recent Acquisitions; and [here the participant hesitated] E-mail & Electronic Reserves, Computer.

Each focus group surfaced the idea of offering "views" that sifted and reduced all possible links into those links most relevant to an individual user. Some wanted several rather generic views (e.g., based on a user’s affiliation with the campus or subject area of interest). Some wanted a very specific view (e.g., based on a more complicated sign-on to identify the user exactly, perhaps remembering a profile of prior search strategies and found resources).

Participants commented that the site should be clearly organized and should "give progressively more detailed information as the user drills down."

Users expressed the following opinions and preferences with regard to navigability:

• Show the main categories for the entire site on each page ("don’t make me go up, to go over")
• Help me to know where I am in the site, and to remember how I got here.
• Return me to the place that I left. (Pathfinder doesn’t always do this; rather it often returns you to a list one above the last one you displayed.)
• The site "shouldn’t open too many new windows that then have to be closed."
• Drop-down menus are good if you know what you are looking for and know the terms used, and if you can see the entire menu on the screen; otherwise they can be frustrating.
• Site maps are helpful if you don’t know the site and the terms; you can see at a glance how the site is organized (though one participant pointed out that the ideal site "wouldn’t need a site map, because the navigation was so obvious and good from the start").

Participants were keen on two points regarding web system functionality: it has to be fast and work well for dial-up access, and it has to be free of ongoing, irritating glitches (e.g. fly-overs that stick; opening too many new windows that then need to be closed; linking to sites that disable the browser’s Back button). Participants expressed the need to print (pages from the web, citations from the catalogs, content from databases) and reported frustration in this regard. Other kinds of functional requirements are implied above, under Features, Organization, and Navigation.

Participants agreed that terminology used should be as consistent as possible throughout the entire site, and stressed repeatedly that the site should use normal language rather than library in-speak. Terms mentioned as potentially confusing included CDL Directory, GLADIS, telnet, CD-ROMs, and Sunsite. Addressing Jill Garland who was attending as an observer, one faculty participant commented "You are new [to the library], but in even a few hours, you learn a lot of jargon -- but it's meaningless to me."

Participants want a site that is uncluttered, attractive, and professional looking. Many were attracted by the simplicity of Google, even while admitting that the library’s site would need to be more complex. Participants commented that the current library home page is "way too busy", even "overwhelming." One professor reported, "I’ve never used the library site – took one look and said ‘no’."

Participants suggested that a better use of font, color, reducing the number of links, and not underlining links, might each contribute to a site that was easier to read (less cluttered) and easier to understand (information architecture made clear via use of visual cues and more informative buckets.) One suggested that we use graphics in an "intelligent and sparing" way (i.e., only if they carry either information or are themselves functioning icons.) Another underlined the importance of his being able to change the font size. Some participants find long pages that require scrolling to be troublesome.

As mentioned above, participants tended not to make distinctions between the library web site and the online systems accessible through it. For example, several expressed desires for features that we think of as related to the library catalog, such as better access to the tables of contents and indexes of books, and perhaps prefaces and introductions ("like Amazon," as one student put it). One student wanted the catalog to "just list the copies that are available to be checked out now," or simplify the results display with details given under each location for an item. Another student wanted to "check out books online, then go pick them up [at the circulation desk]."

Likewise, participants felt that a search that works in Google should produce similar, or at least some, results in the library catalog. One reported with some frustration that Google gave 47,000 hits on his search term, while Pathfinder gave zero. Some participants voiced concern about whether Pathfinder gives results consistent with, or as complete as, what one would get from GLADIS; others commented on confusing search choices and "back button" behavior in Pathfinder.

It was evident from these discussions that abnormalities and frustrations that users encounter with display, downloading, or printing of records in the catalog or full text items in article databases are closely associated with, if not actually attributed to the library web site. This underlines the importance of negotiating with vendors to provide services at a level of excellence that we ourselves would want to provide.

The web development process
Several suggestions were made about how we should proceed in developing the web site. One professor recommended that we tap into the talents of the many computer system experts on campus. Staff participants suggested that we hire (or find in-house) a good graphic designer, provide for orderly growth of the site by using templates and guidelines, and move to a rapid prototyping environment.

The focus groups offered a number of ideas that could be accomplished with existing resources as part of the planned redesign project. We recommend that WAG evaluate participants’ comments and suggestions carefully and supplement them with data to be gathered from user surveys, card sorting exercises, terminology tests, and usability tests.

In order to share this report with library staff and other libraries that may find it useful, we recommend that this document (without appendixes) be posted on the WAG web site and made known within the UC System, and that participant comments related to the library systems and services other than the web site be brought to the attention of staff in charge of those areas.

In the longer term, we also recommend that WAG address the question of how the library might deal with the expressed needs and preferences that focus on the catalog, require technology or resources not currently available, or otherwise fall outside this group’s current purview. Examples include participants’ desires for enriched catalog records, a radically improved search engine with the ability to integrate results from a variety of sources, and the ability to log onto the site once per session and see options and resources tailored to their individual profiles. This situation invites "out-of-the-box" thinking so that the next phase of development can move farther toward the web site that many users desire, and some already expect.

Submitted to WAG – May 30, 2003
Discussed and accepted by WAG – June 16, 2003