Andrew Dillon

Andrew Dillon

Andrew Dillon has quite a few responsibilities at Indiana University, where he is Associate Professor of Information Science and Informatics, is part of the Core Faculty in Cognitive Science, and is Adjunct Associate Professor of Computer Science and Instructional Systems Technology. (Academia really needs to do something about its labeling systems already.) Educated in Britain and Ireland, and one of the stars at last April's ASIS Summit, Andrew is yet more proof that you can find incredible talent away from the coasts.
-- Lou Rosenfeld

Lou: Based on the e-mail I get, information architects are way more interested in suggestions regarding which tool is the best for developing site blueprints than knowing which graduate program will provide the strongest theoretical foundation for their work. As an academic, what's your take? Are budding information architects bypassing programs like Indiana's for a welcoming job market, or are you seeing growth in the numbers of students there to learn about information architecture (IA )?

Andrew: There is no doubt that the job market is very hot at the moment and people with the right mix of skills have no difficulty finding employment. However, our programs here at Indiana are also in high demand and classes in such areas as human-computer interaction, interface design, usability testing, and Java are very popular.

I think many students view what we teach as a means of transitioning from their undergraduate degree in the humanities or the sciences to a technology-oriented career. We are about to launch a new Master's degree in Human-Computer Interaction through our new School of Informatics, and if initial demand is any indication, I suspect we will have a lot of people looking at Indiana University as the place to train as an Information Architect.

Lou: So you feel that academia can teach the right skills for information architects who practice in the "real world"?

Andrew: Undoubtedly! We have state of the art facilities and a first rate faculty who research as well as teach IA, even if they have always called it computer science, information science, library science, or something else. At Indiana we have worked hard at building a cross-disciplinary program in Informatics that bridges these disciplines so that we can best teach and study issues fundamental to IA.

While it is easy enough for someone to learn basic skills on their own or through stand alone classes, we all know that being a good information architect requires knowledge of human beings, how they think, how they communicate and how they behave in context. Such skills are hard to pick up on the job, and while experience can add polish, the fundamentals of IA rest on the knowledge that universities like ours create and pass on.

Lou: What kinds of skills can a budding information architect pick up in an academic setting? And which skills should we just forget about acquiring in academia?

Andrew: A tension exists between what students want and what I believe they need, at least initially. Most students come in wanting to master as much technology as they can, feeling that this is the key to their job placement. I work hard at convincing them that while technological competence is fine, technology changes rapidly and they will be better equipped in the long term by understanding more about users, more about the process of design and more about usability testing methods.

Our goal is to equip information architects with the skills to have a career, not just a job. Thus, we encourage students to think about theories of human behavior and how they apply to interaction design, and we teach them how to test out ideas and use statistics to estimate confidence in their findings. Such skills are essential for information architects and are really only likely to be acquired in an academic setting.

Lou: How does research fit into the picture for the rest of us? Are information architects and others who design web sites condemned to reinvent the wheel, or is there a "silver bullet" that will convince practitioners to investigate decades worth of relevant literature?

Andrew: That is a really intriguing problem. There are decades of relevant findings about design, navigation, reading from screens, and so on that many new information architects just ignore. Part of this is the sheer "unusability" of much of that knowledge, which is typically found in rather dry academic outlets like conference proceedings and academic journals; smart information architects will immediately appreciate this irony.

But another part of the problem is the technological bias of many people drawn to the field; this bias causes many to view findings from previous technologies or previous decades as irrelevant to current concerns. If I had a dollar for every time I was asked if any of the work on hypermedia was relevant to design of the web, I think I could retire today. The trick is to see the process of interaction as a human one and then to see how the context of use that was tested relates to one's current questio n. In that light, ten year old research on menu structures or navigation aids in hypermedia is directly relevant to web design, as long as it was good research in the first place (which of course, is a quality control function that is offered by good journals and conference proceedings).

So there is no silver bullet, but I think researchers are now more aware that their results might need to be distilled into several forms to have the most impact on practice.

Lou: Have you noticed any particular research that has had a direct impact on how information architects (or, more broadly, web developers) are doing their work?

Andrew: I think the writers who build bridges between the literature base on HCI and current needs of design folks fulfill a really important role--and I think the community has responded to the works of Jakob Nielsen, Ben Shneiderman and Don Norman in a way that shows there is a hunger for such knowledge. And of course, your own book on information architecture is practically required reading for anyone in this field. (Lou: thanks for the plug!)

The downside of course is that it is too easy to take any one writer's views as the only word on a subject, whereas my view is to let evidence or data from real users speak loudest. Unfortunately it is not always clear from some books just how the data were collected or how evidence supports a certain design recommendation. I think within IA there is a need for a dedicated forum for research findings and design recommendations, and I suspect we will see such a forum emerge before too long.

Lou: Let's talk about your research. One concept you explore deals with the shape of information spaces. Shape certainly has architectural relevance in tangible self-contained information spaces, like books, where shape is visible and directly useful. But does shape have practical value when it comes to developing usable architectures for more abstract, multi-dimensional information spaces like web sites?

Andrew: Well I certainly believe that the concept of shape is practically as well as theoretically relevant to IA. As you say, shape can be made visible to users and this has always been so with books. I suspect many of the navigation problems users experience in digital worlds are directly related to poor shape cues and affordances, and in one sense, IA is all about giving information a shape that enhances users' movements through that space.

I coined the term "shape" after listening to users talk their way though various interactions with digital spaces, it seemed to capture the mix of spatial and semantic issues they dealt with. Also, I use it to break down the idea that navigation is completely distinct from comprehension of information, as in my view that is just not true. Comprehension of any sizable information space requires navigation and the cues users rely on often have semantic relevance to the information space, not just visual or spatial relationships such as arrows or forward/back buttons. After all, a good link is loaded with semantic information as well as a visual cue that it is selectable. And of course, users do not go to web sites just to navigate, do they?

Lou: What site are you visiting lately that other information architects should check out? Why?

Andrew: I use the web a lot and am always receiving pointers to new sites and interesting designs from my students. However, I think the best test is to look at the sites I routinely use because they offer something that is part of my everyday life. To that end, I confess to being completely reliant on Soccernet for news about my favorite game, and I bank online with a local credit union which allows me to bank at night or while away from home. Neither site wins design awards for aesthetics, and I have even offered to run tests for both but was greeted with a wall of silence in response, but both have provided something important to me that I cannot easily get anywhere else . So you could say they have both got something fundamentally right in their designs for this user, look and feel notwithstanding.

I also made great use recently of the Edmunds car site where I found a wealth of accessible information on vehicles. As you can see, I like sites that are content rich and empower the user. A little more of that in our designs, and less concern with flashy images would signal a triumph for the field of IA, in my view.

Lou: Speaking of soccer, how is Manchester United doing this year?

Andrew: Ha! Well, the club is doing great. The only downside is that they have had their most successful period in years during the time I have been living in the States (correlation does not equal causation!) so I never get to see a game these days. Check out their site and you might give thanks (as I do) that their players are better than their web designer!