Interview with John Kupersmith
Reference Librarian
University of California, Berkeley
John Kupersmith is a reference librarian at the University of California, Berkeley, where he is taking part in several web redesign projects, editing a current news item on the library home page, and co-teaching a web searching class. In the past he has been Internet Services Librarian at the Washoe County Library (Reno, NV) and Networked Information Services Coordinator at the University of Texas at Austin General Libraries. John’s interests outside of work include hiking, kayaking, and photography. His web site includes material on web design, library terminology, technostress, and other topics.
I spoke to John recently about his thoughts on library web site design and how librarians are thinking out of the box.
Michael Whang: Why did you create a clearinghouse of library terms in usability testing and other studies?
John Kupersmith: Because these studies confirm that many users don’t understand our specialized language. Even terms such as “library catalog” can cause problems. When students ask for help finding the “search engine” to look up a book, this suggests we have some work to do. In the print-only era, users who couldn’t understand our signs or handouts might still have been able to find their way to a physical destination. But to use a web site effectively, the user has to understand what’s presented, at least well enough to make appropriate choices. With more and more libraries conducting usability studies and redesigning their web sites accordingly, it should be possible to identify some useful data and patterns, including terms that are relatively well understood. To gather and share that information, I’ve started a site called Library Terms that Users Understand.
MW: You mention “reducing cognitive barriers caused by terminology”. I take it this means using a term that users can easily recognize and identify with as opposed to a term that forces users to stop and think about its meaning or its relationship to the thing that they’re looking for.
JK: That’s the ideal—instant recognition with no ambiguity, or as Steve Krug puts it, “Don’t make me think!” In practice, if users don’t understand the concepts behind the terms or behind some of the distinctions we make, we may not be able to realize this ideal completely. But in many cases we can still get closer than we are now.
MW: Finding a term or label that everyone finds obvious is just one aspect of designing a good navigation and way finding system. What else can Web developers do to help users understand how to use the site, know where they are, and know what the site contains?
JK: It’s important to choose terms carefully based on data about what users understand, but also to structure the choices we present so that they fit a user’s frame of reference, and reveal more information to expand that frame as you go along. For example, a home page link might say “Find Articles” and lead to a page with further options. Also, if we know that users are likely to make an inappropriate choice in a predictable way, why not put something in place to catch and guide them? An example might be a “Find Articles” link on the library catalog’s opening screen, leading to a selection of article databases.
MW: What are terms that you have found to be most successful on your library Web site?
JK: We’ve found that users tend to select links containing what I call target words – such as “Find Books” – versus tool names such as “Library Catalog”. Other libraries have reported the same thing. We’ve also seen users gravitating to “Library Services” for a wide range of content. These data are from testing small numbers of people, but what if there’s a pattern across multiple studies in multiple places? That could be useful.
MW: How did you go about identifying those terms? ?
JK: Mostly as a byproduct of conventional usability tests (observing users as they do tasks). We’re just starting to plan for systematic terminology testing, and right now I’m trying to figure out the best way to do this. The conventional wisdom is that you do card sorting, but I think that’s likely to generate better data on site organization and grouping of elements than on terminology. An alternative method is to give users a hypothetical situation and ask them which of several alternative link wordings they would choose. A third possibility is to interview users and ask them in depth what they think various terms mean. All these methods have advantages and drawbacks, and I’m still looking for the best option.
MW: Did you have a difficult time getting buy-in from librarians on the choice of terms used for the site?
JK: In my experience, most librarians recognize that we have a genuine problem with terminology. Some believe that terms such as “indexes and abstracts” are part of what students need to learn in a college or university, saying we should not “dumb down” our site. I actually have a lot of sympathy with that view. The problem is, since we can’t reach all students with instruction programs, should we design the site for those who have already learned our language, at a cost to those who have not? To get around this, we’re experimenting with providing multiple paths: target words for the novice, plus more specific, tool-oriented choices for the expert user.
MW: For librarians new to Web development and to designing a positive online experience, what are some resources that might help avoid any design pitfalls?
JK: I think one of the best places to start is Donald Norman’s book The Design of Everyday Things. Even though it dates from before the web, it’s a great introduction to the usability mindset, with classic (and often hilarious) analyses of such things as light switches and telephones. Edward Tufte’s books on visual design are useful in the same way. In the web area, there’s Steve Krug’s Don’t Make Me Think, Louis Rosenfeld and Peter Morville’s Information Architecture for the World Wide Web, and the various works of Jakob Nielsen.
MW: A majority of libraries design by committee. Were there any lessons that you learned during the process? (What did you learn and can share with your readers? What would help other librarians / developers achieve success in this kind of situation?)
JK: I’ve been reminded lately that people from different specialties within the library have very different perceptions of user behavior and needs, and may express these in different ways. Public services and web usability are themselves specialties – it’s important to recognize that your own knowledge base and language may not be familiar to others. I think design by committee works best when everyone accepts and respects these differences. It’s also important to realize that all librarians are somewhat contaminated as designers because we do know the language and underlying structures. It takes some effort to put aside our preconceptions and focus on data about the users themselves.
MW: How do you think the Web has changed perceptions and use of the library?
JK: There are several forces at play here. Students are under a lot of stress and time pressure. It’s very easy for them to bypass the library completely when they’re online, going instead to free, easy-to-use search engines. The library has much more to offer in terms of quality, but we’re at a real disadvantage in terms of convenience. It’s interesting to hear students describe library web sites and catalogs as “overwhelming” – the same way they’ve always talked about physical libraries. So, we have a marketing challenge to make students aware of what we offer, how it’s special, and why it’s worth a little extra effort. We also have a design challenge to make our web sites and catalogs more usable, and to encourage database vendors to do the same with their products.
MW: Are there current projects where you applying good design strategies?
JK: UC Berkeley has recently redesigned the web site for its principal social sciences and humanities libraries in a way that reflects usability testing and incorporates some of the ideas I’ve described. We’re now looking at the design of the library’s main site. I’m also updating a 1998 article on technostress that includes a lot of material on online systems design, and hope to have that posted on my site this summer.