Investigating the Interface:by Nancy Newins
A New Role for Instruction Librarians
The credit course, the library lecture, workbooks, computer-assisted instruction — the list goes on. In twenty-five years as an instruction librarian I have tried all of these, yet I find myself increasingly dissatisfied with traditional approaches to reaching today's "plugged in" students and faculty. The most active instructional program rarely reaches more than a fraction of the total user population on our campuses. Also, library instruction delivered through traditional methodologies rarely reaches students at the proverbial "teachable moment."
Even the one-on-one instruction that is the heart of the reference transaction is becoming more infrequent as our patrons access information electronically from dorm rooms and offices. How can we assist faculty and staff to use electronic resources effectively when we are unable to reach the vast majority of users? Do our users really care about Boolean logic and search strategy or does this just stand in the way of their getting to needed information?
Rather than focusing exclusively on training the end user, I suggest that instruction librarians might also direct attention toward improving the information system itself. Let's assume a new role as "investigators of the interface," with the goal of increasing usability and making library systems (databases, web pages, online catalogs) transparent to the user and less instruction-intensive.
Other librarians have made this suggestion, and a recent article in Journal of Academic Librarianship presents a brief literature review. In an article titled "Way Beyond BI: A Look to the Future," the author suggests that instruction librarians need to look toward designing library systems "so easy to use and transparent that library instruction is not needed."(Herrington, 1998, p. 383). She cites other authors who have advanced similar suggestions (Campbell, 1993; Rettig, 1995 and Nielsen, 1986), and concludes: "The new library instruction model features a library system which is user-friendly, seamless, efficient, and effective." (Herrington, 1998, p. 385)
During a summer sabbatical sponsored by Randolph-Macon College, I began to study issues of interface design and usability through an introductory course in "Human-Computer Interaction" (HCI) at Indiana University-Bloomington's School of Library and Information Science (SLIS). I also participated in a workshop on usability testing to gain practical experience in doing actual testing in a usability lab. Both class and workshop were part of the Master of Information Science (MIS) as opposed to the MLS program in SLIS. Consequently, over half of the students in the class had no library background, which presented an interesting perspective.
HCI can be very broadly defined as a field that looks at all aspects of users interacting with a computer system or "tool" to do a specific task in a given environment. Usability testing, on the other hand, involves empirical testing of the effectiveness, efficiency and satisfaction of a given computer system (web page, software program, online database) with users who have been given a specific set of tasks to perform. From results of usability testing, designers can evaluate their work and determine changes that need to be made in the system design.
My first experience with usability testing was in the IUB SLIS usability lab as a subject in a test of a library home page developed by a large academic library in upstate New York. Imagine my dismay when, after completing the set of tasks which involved finding specific information using the home page, I was greeted with gales of laughter by the testers (non-librarians all). Their conclusion was that the page had been "designed by librarians for librarians" since, as the only librarian-subject, I was the only person who could use the home page to find information. What I viewed as a typical library home page was apparently unusable by non-librarians. I had learned the first lesson of usability testing: only testing with actual end users can determine the usability of a specific information system.
Testing typically occurs in a laboratory which can be as basic or as sophisticated as resources permit. A usability test can start with a test setup that consists solely of a video camera directed on the computer screen to videotape the subject's (user's) responses. A more formal usability lab, however, typically has two rooms: one for testing and the second used as a control room to monitor the testing. Two to three cameras, positioned appropriately in the room, record responses--one focusing on the computer screen, one focusing on the keyboard, and one taping the subject to record facial expressions and body language.
The following diagram shows OCLC's usability lab, which includes a third room for observation of tests by product designers, marketing personnel and managers. (Murphy, 1997, p. 21) I visited OCLC as a part of my sabbatical to view a usability test in a corporate setting. [See http://www.oclc.org/oclc/new/n229/ulab.html for full article on usability testing at OCLC.]
The usability test itself employs a defined testing protocol, including a script to brief subjects on what will occur in the testing, a set of tasks that the subject must complete, and a questionnaire to evaluate subjects' subjective satisfaction with the program. Verbal protocol--asking the subjects to talk about what they are thinking as they work through the tasks--also provides excellent data for analysis of user errors, and captures the users' general reaction to the program.
How can usability testing be of value to the practicing instruction librarian? First, as I discovered in my summer work, librarians need to usability test their home pages and web-based instructional programs with actual users. We often overlook this step in the rush to update the home page, or to add yet more information to our web site, without realizing that end users still cannot find what they need on our page. At Randolph-Macon College we plan to usability test future revisions of the library's home page with members of the college community.
I am also currently working with a professor in our Psychology Department on a usability testing project to determine the ability of general psychology students to use two interfaces to two different online versions of "Psychological Abstracts." This study has already provided useful information on student success (or lack thereof) with two very different interfaces, and the information will be used in making future selection decisions. The data will also be useful in focusing instructional efforts toward students' observed difficulties in use of the database.
In the competitive business of marketing online services database vendors are becoming increasingly conscious of the importance of usability testing. OCLC does extensive testing of products in development, and a 1997 issue of the OCLC Newsletter describes their operation (see http://www.oclc.org/oclc/new/n229/ulab.html). At the VIVA Users' Group meeting at the 1998 VLA Annual Conference, when asked which vendors were doing usability testing, all vendors present replied affirmatively.
Instruction librarians can provide suggestions to vendors for improved database interfaces based on formal or informal usability testing with patrons. The quality and usability of the interface should always serve as a major factor in any selection decision.
In Information Tasks: Toward a User-Centered Approach to Information Systems, Bryce Allen notes: "Libraries... seem neither to have thought through the needs of their users and the tasks they are accomplishing nor to have developed appropriate systems to address those needs and tasks." (Allen, 1996, p. xi) Instruction librarians need to take the lead in addressing issues of usability to enable patrons to use online resources efficiently, effectively, and with a minimum of instruction.
Allen, Bryce L. 1996. Information Tasks: Toward a User-Centered Approach to Information Systems. San Diego: Academic Press.
Herrington, Verlene J. 1998. "Way Beyond BI: A Look to the Future." Journal of Academic Librarianship 24:381-385.
Murphy, Bob. 1997. "OCLC Usability Lab Helps Staff Learn More about the User." OCLC Newsletter no. 229:21-27.
Nancy Newins is Head of Library User Services, McGraw-Page Library, Randolph-Macon College, Ashland, Virginia.