Why is usability so hard? [Observations from the information retrieval perspective]

I had to move this one back to the top because after someone's comment about this thread on other blogs, I started to think about this article from the library and information science/information retrieval perspective. In Human-Computer Interaction Resource Network (HCIRN): Reflections, Adam Smith tackles the difficult question, "Why is usability so hard?" Answers Smith's answer is right on the mark. He says These are the cornerstones of user interface design -- people and goals/tasks/strategies existing within a context. To create usable, effective designs that people like, you need to understand all of those on their own terms and as a functioning whole. And that's why designing usable systems is so hard. Information goals and context That's a nice summary that makes good sense. Smith also says that "the first trick to designing usable systems is actually to not think about the computer at all, but to focus instead on the goals." It should come as no surprise to people with information retrieval backgrounds, that his statement is not trivial in anyway. I realize that for many users information seeking on the Web can be for experiences that are ephemeral and not directed at use outside of the electronic world, e.g. chat, entertainment and surfing. Smith's statement interests me most as it applies to users that seek knowledge to do something offline -- the users browsing and searching large information collections with the goal of extracting knowledge for use in the physical world ("meatspace" as opposed to "cyberspace", as I've read it refered to.) Information retrieval professionals obsess over the part about goals and context. In a good session with a reference librarian, you will be asked about the intended use for the information you seek and for some context -- the what and why of your information need. i.e. if you are a pre-med student doing research on current children's vaccinations for a term paper, the type of information you require may not be the same as for a parent looking for current information on the same topic. Or maybe it would. The context surrounding your need has to be fleshed out before the search intermediary -- the librarian -- can direct you to a good place to look. To complicate things, often times users cannot express easily in language what the information need is -- what Nick Belkin called an "Anomalous State of Knowledge". [1] If the information returned is not appropriate, the user and the intermediary iterate through the process. (I once illustrated this concept -- the classic information retrieval model -- crudely in an article I published on image retrieval [2].) So uncovering the goals and context to find the information can be a bit more complicated then just asking, "What are you looking for?" Information seeking on the Web OK, you say, that's nice for libraries, but what does that have to do with creating usable Web sites and information retrieval systems? On the Web, the information seeking episode replaces the human intermediary with the interface and the search form. The user interface is our main tool for presenting access points to our information through navigation and the search funtionality presents a method for posing questions to our body of stored knowledge. One of the problems with designing interfaces or systems, however, is that if you have not anticipated the kinds of questions your audience will ask, then it is quite difficult to present the information topology to them in a way that makes sense -- to map the access points to your body of knowledge with the user's way of thinking of that body. And that, to me, is where the system may tend to fail. This may not be such a problem when dealing with a shallow corpus of content, but when the body of information you are attempting to provide access to is extremely large, as is the case with Digital Libraries, representation of the knowledge contained within that body is extremely difficult. In the LIS field, a lot of the information retrieval literature that I have read focuses on users' information seeking behavior to uncover how people search and to point out how systems fail users in being able to use that information to do things in the real world. This cannot be very different from what Usability professionals do. With the availability of so much of our knowledge being on the Web these days, it's not surprising that so many people focus on the issue of the difficulty in creating Usable Web experiences. For references, click "Read more" below. [1] N. J. Belkin, R. N. Oddy, and H. M. Brooks. Ask for information retrieval: Part I. background and theory. Journal of Documentation, 38(2):61-71, June 1982. [2] M. Angeles. Information Organization and Information Use of Visual Resources Collections. VRA Bulletin, 25 (3), 51-58, Fall 1998.