Library as Laboratory: Computer Science Students Practice Usability Engineering in an Academic Library
Oregon State University Libraries
Oregon State University
Copyright 2007, Margaret Mellinger. Used with permission.
This article presents a project-based collaboration between a usability engineering (UE) course and the campus library. Student teams developed UE projects based on usability issues they identified in the library's web-based and information technology services. Using the library as a focal point, students practiced common UE methods such as ethnomethodological observation, web usability evaluation, and an iterative prototype process based on cycles of design, testing and evaluation. At the end of the term, teams of students presented their novel solutions to library staff as part of their final grade. Other academic librarians may find inspiration to pursue similar partnerships on their own campuses.
Usability engineering (UE) is a user-centered, systematic approach for designing software systems and user interfaces that can be "readily comprehended, quickly learned and reliably operated (Butler 1996
)." UE encompasses a group of methodologies, including human-computer interaction, usability testing and evaluation, software and interface design, user centered design, and elements of human physiology and psychology (Preece 2002
). Usability engineering integrates usability throughout the development process.
One of the best ways to teach usability engineering is to provide students with authentic and appropriate hands-on project experience (Rosson et al. 2004). However, it is a challenge to find projects that are meaningful and can be completed within a term. Rosson and Carroll (2002) recommend scenarios and case studies as starting points for UE student assignments. This paper describes a similar, yet alternative approach of generating student projects from observations and analysis of a place, in this instance, the campus library. The collaboration benefits both students and the library.
In Introduction to Usability Engineering (CS352) at Oregon State University, students learn basic methods of user-centered design, including evaluation of software systems and user interfaces. The major assignment of the course is to work through a usability engineering process with a team. Over the span of the project, student teams 1) identify a usability problem with the library web site or information technology services, 2) propose a novel solution, 3) prototype the solution, 4) employ usability testing methods for evaluation of the product, and 5) refine their prototypes using the evaluation and feedback from the previous steps. The final prototypes are presented to library staff, with the understanding that the library may implement some of the solutions.
Identify Usability Problems: Ethnomethodological Study
To identify potential usability problems, students conducted observations in the library. After learning about ethical experimentation methods and taking an Institutional Review Board certification test, students did their "library field work" in one week. Teams chose specific places in the library for their observations, such as the Information Commons, a quiet study area, or a group study area. They spent hours in the library conducting their assessments. The professor instructed the groups to make an exhaustive inventory of four elements: the people using the space, the architecture of the space, the objects in the space, and the technology in the space. The students were advised to disregard preconceived notions of libraries and to write down their impressions, feelings, and ideas as they documented their observations.
The point of the exercise is to systematically analyze the library as an information system and to discern which functions and populations it supports. The assignment also helps students identify potential usability problems on which to base their final projects.
This activity employs elements of ethnography, but cannot be called an ethnographic study. True ethnography requires proper training, and a longer period of time than is available in a ten-week term, hence the activity is termed ethnomethodological (Crabtree et al. 2000). The approach, but obviously not the scope, is similar to one used by Given and Leckie (2003) in a large and comprehensive observational study of two Canadian public libraries. In that study, the researchers captured data on over sixty variables by making repeated "sweeps" of the library and noting what users were doing in particular spaces at specified times.
Students in CS352 took on the role of complete observer (Baker 2006). In other words, the student teams observed, but did not interact with library users. Student researchers were permitted to question or interact with library staff or faculty, if needed. Because of this level of detachment, student researchers were able to gain only a snapshot of users' activities in the library. However, a snapshot was enough to give them a sense of the power of ethnomethodological research in user-centered design, and to spotlight potential problem areas in the library's web-based services.
Propose a Novel Solution: Project Proposals
Incorporating analysis from the library observations, student teams generated preliminary project proposals. They presented those proposals in class and received feedback on the quality and feasibility of their ideas from the instructor and their peers. Each student team, with the input of the instructor, chose one proposal for its final project. This began an iterative process of design, testing and feedback. No programming is required in CS352; students are encouraged to use paper prototyping, a simple, yet powerful tool, and to focus on designing for usability.
Prototype the Solution: Design Gallery
Working in small teams, students designed prototype solutions to their chosen usability problems. Students brought their preliminary design prototypes to a design gallery. Design galleries are more commonly used in fields such as graphic design or architecture, but are well adapted to design work in computer science. The professor wanted computer science students to gain experience both in sharing their designs with others, and in delivering and receiving critiques of their work in an open forum. The first showing of the prototypes was in class, to give students some familiarity with the process before presenting to an outside audience. The teams set up their prototypes around the classroom and took turns describing them to other students and to several computer science graduate students with usability engineering experience. Teams used the feedback to assess the feasibility of their prototypes, to determine if the usability issues were addressed, and to go on to the next iteration.
After refining their prototypes with the feedback from the design gallery, students developed usability tests to evaluate their prototypes. These could include standard tests based on the GOMS method (goals, operators, methods and selection rules), originally devised by Card, Moran and Newell (John 1995
). Students could also choose to conduct short, controlled experiments. One class session was devoted to testing. Students both administered the usability tests they devised and participated as subjects in other teams' tests. As with the ethnomethodological observations, the usability testing phase was an abbreviated version of an iterative process that could take weeks or months for a commercial product.
Using the feedback from the web usability testing and evaluation, the teams refined their prototypes once again for the final project. At every stage of the usability engineering process, different sorts of feedback and evaluation are used to improve the designs. At this stage, some of the students chose to go beyond paper prototyping; one or two teams programmed a partially operational new interface for the library catalog, and other teams used illustration or presentation software to demonstrate their ideas in a more polished way than the paper prototype.
For the final exam, UE students presented their final prototypes and project descriptions as posters in an open house hosted by the library. Student teams described their prototype solutions to groups of library staff and to their professor, who visited each team's display to evaluate the students' work for a final grade. Library staff expressed keen interest in the projects and busily took notes as they chatted with students about their prototypes.
The library has now served two times as a sort of laboratory for the usability engineering course. The first time, the instructor gave students wider latitude for projects. The second time, projects were limited to web-based or information technology problems. While students in that first term identified real usability issues, some of their prototypes were too impractical to be pursued as operational solutions. For example, the former library laptop checkout system required the student requesting the laptop to wait near the circulation desk until their name was called to pick up a computer. When there was a queue for laptops, students uncomfortably waited around the crowded circulation area for their turn. UE students proposed a stand-alone circulation system for the laptops. This solution was not technologically feasible for the libraries. However, struck by the group's identification of the problem, library staff solved the problem in another way. The library implemented a restaurant-style beeper system. Students take a beeper, leave the circulation area to study or get a snack elsewhere in the library, and the beeper goes off when a laptop is available for them to check out.
The results from the second iteration of the class had a higher degree of feasibility. Projects included an online study room reservation system (it is currently a paper system), an improved interface display for a large plasma screen in the Information Commons, and several perspectives on improving the appearance and functionality of the online library catalog, especially for locating materials in the library. From this group of prototypes, it is likely that the libraries will address the room reservation system; it was identified as a problem in both the first and second set of student projects. Because the library requires the room reservation system to integrate with the patron information database, it is likely that the library will purchase software that works with the ILS to solve this problem, rather than building a system locally.
An unrealized aspect of this collaboration is that some of the students' ideas have strong potential to be developed and implemented. Ideally, there would be a venue for students to do so. One idea is for students to elaborate on their projects later in the curriculum, perhaps in the senior design course. Another idea is to build on the partnership by pursuing funding for some students to work with the libraries to implement their ideas. This would build the students' vitas, while improving the library's web-based services.
Students worked on real-life problems and learned about the library as an information system with digital, physical, cultural and social aspects. They experienced the power of a user-centered design process and gained an appreciation of the overall usability engineering process. Students were surprised and pleased to find a great deal of interest in their ideas and work as they engaged with librarians and staff to complete and present their projects. It was a positive learning experience.
The libraries also benefited from this partnership. Librarians and staff were astounded by the students' inventive, novel and sometimes provocative solutions to usability problems. The problems students addressed were often known to people in the library, but seeing the students' new approaches to seemingly intractable usability issues energized the library staff. Some of that energy has translated into improvements in library services.
Academic libraries have a wealth of expertise near at hand, and tapping into those riches strengthens both libraries and the communities they serve. Librarians in liaison roles are in great positions to scout for new ways to build partnerships with faculty and students. In what ways can your library reach out to partner with experts in your community?
The author wishes to thank Dr. Carlos Jensen, for connecting computer science students to library usability issues. Some of the content of this article was previously published in the OSU Libraries newsletter, The Messenger
Baker, L. M.
2006. Observation: A complex research method. Library Trends
Butler, K. A. 1996. Usability engineering turns 10. interactions 3(1): 58-75.
Crabtree, A., Nichols, D. M., O'Brien, J., Rouncefield, M. & Twidale, M. 2000. Ethnomethodologically informed ethnography and information system design. Journal of the American Society for Information Science 51(7): 666-682.
Given, L. M. & Leckie, G. B. 2003. Sweeping the library: Mapping the social activity space of the public library. Library & Information Science Research 25(4): 365-385.
John, B.E. 1995. Why GOMS? interactions 2(4): 80-89.
Preece, Jennifer. 2002. Interaction Design: Beyond Human-Computer Interaction. Danvers, MA: John Wiley and Sons, Inc.
Rosson, M.B. & Carroll, J.M. 2002. Usability Engineering: Scenario-Based Development of Human-Computer Interaction. San Francisco, CA: Morgan Kaufman.
Rosson, M. B., Carroll, J. M. & Rodi, C. 2004. Case studies for teaching usability engineering. In: Proceedings of the 35th SIGCSE technical symposium on Computer Science Education, pp. 36-40. Norfolk Virginia, USA: ACM.