A List Apart
Brightly Colored Food
City of Sound
Croc o' Lyle
Digital Web Magazine
Dive Into Mark
Guide to ease
Joel on Software
Noise Between Stations
Off the top
Signal vs. Noise
Mark Hurst has written an interesting discussion about web pages and how people navigate. In it, he reminds us of something he wrote in 1999,
On any given Web page, users will either… click something that appears to take them closer to the fulfillment of their goal, or click the Back button on their Web browser.
The interesting part of his message here, I think, is that the IA/designers’ focus on aspects of the UI such as navigation consistency is less important than the supporting of users in getting them to their intended goal. He says provocative things such as “users don’t care where they are in the website”. If you can get your head past that idea, 3 bullets summarize what this should mean for you in practice:
I’ve posted additonal personal opinions on this topic elsewhere on my weblog. Peterme discusses Mark’s ideas as well, pointing out that he shouldn’t dismiss the value of wayfinding cues in order to make the point that empasis should be placed on user needs and behaviors supporting those needs. Christina doesn’t see the harm in Mark’s oversimplification and suggests that informational cues such as breadcrumbs put the burden of mental strain on the user. It’s nice that she also suggests alternatives identified in her Widgetopia to helping users identify alternate paths related to their current task, addressing a point that I think is important — “Where can I go” is perhaps more important than “Where am I?”. Manu Sharma adds that both Peter and Mark are probably both right in this debate, but the difference in perspectives is explained by their different experiences.
I wrote about some research we're doing in my organization to observe user interaction with navigation by tracking where users click on the page (body, local navigation, breadcrumbs, global navigation). Our observations aren't dissimilar to what Michael Bernard observes in usability testing -- links to content are most often searched for/clicked in the body of pages. Navigating our site (a digital library) consists mainly of browsing through a directory (a-z lists are available as are a poly-hierarchical directory listing), so what we were mainly interested in was how people made use of the links in the local navigation. I'd be interested in seeing if other people have done this and what they were looking for. I find, as an in-house site developer, that being responsible for a site for a long term (as opposed to just launching one and going on to a new project) gives one good opportunity to observe and assess the site for usability. Your can assess patterns of use over long periods of time. You can make contact with users and keep the lines of feedback open with them over time. Clearly there is something unique about being involved in the evolution of a singular site, which I am only beginning to appreciate.
I recently wrote an article for CNET's Builder.com on recommendations for web site navigation. I'm targeting software developers like myself, who may not be as familiar with the principles of site design and architecture as usability experts or web designers.
At the SLA conference this year I got to demo Anacubis. I don't often see anything too interesting in the exhibition halls of library & info. sci. conferences, but this tool caught my eye. Based on the investigative software used by police in England and developed by i2 Group, Anacubis is a visualization front-end for data sets. The demos that showed included a front-end for Dun & Bradstreet company research, Lexis Nexis Legal databases, and intellectual property databases. In the D&B example, once you are viewing a company's information you can browse customer, competitor and subsidiary companies, as well as view officers visually. I've seen some application of similar visualization tools -- mostly it seems social network and map-based stock market visualizations have been around -- but this seems to be the first major commerical entry in the area of commercial information provision. I can see major advantage in the visualization of patent information, for example. This kind of information can be invaluable to companies looking to protect their patents and visualization tools can certainly help exploit our visual senses, which are more efficient/quicker when it comes to picking out patterns of information.
An article in New Scientist reports that new research shows mice "make signposts out of leaves and twigs so that they do not get lost in fields".
"The wood mice might need to use signposts because the fields where they live are very bland - one patch of ploughed field looks much like another (..) And while some other mice use scent markers, wood mice are wholly visual". Much like humans.
Joshua Kaufman collects some examples of contextual navigation using paging.
Change Sciences has an archive of best practices whitepapers they've produced. Free registration required. Topics include writing for the web, navigation and orientation, search, checkout, user registration, and two interesting 'design paradoxes' articles. Most interesting to me is the recent task design article, and the two older, but still valuable ROI & Investing in User Experience papers.
It sure would be nice if the best of sigia-l was culled periodically. Scott Berkun does this from time to time. Maybe the signal to noise has gotten better on the list?
The Guardian has a good review of the UK site Upmystreet.com, which allows people to seek information/services within a neighborhood by entering a postal code. The site has gone a step further by connecting people in within that locale as well. The ability to mix information seeking and interpersonal interaction seems like an interesting idea. When you consider that mobile devices will can be used to access services like this, new possibilities as well as new concerns are inevitable. Apparently there are some issues of privacy and safety, such as concern over the safety of children using the service. Nevertheless, a cool new way of making connections via locale.
Lucian pointed to the short Syntagm article by William Hudson on right-side navigation. Hudson, responding to Bob Bailey's HFI newsletter article on the topic believes that we need more data before we can know that moving navigation to the right will be a real improvement.
I am in a discussion with a programmer about ways to offer navigation using a poly-hierarchical arrangement of nodes. He brought up the concept of directed acyclic graphs (DAG), which is from Mathematics. I learned from the Free Online Dictionary of Computing that the idea is that a directed graph would contain no cycles, i.e. if there is a route from node A to node B then there is no way to cycle or loop back. I can see some applications benefitting from this algorithm, such as in forward citation searching. I think I may not understand the concept entirely, but I am guessing that in an information environment, this means that you'd lose context the deeper you find yourself in a directed path. Or perhaps it simply means you navigate forward to point A from point B and has nothing to do with providing backward movement.
The problem we're experiencing is that we have been dealing with a legacy of organizing by collections/products/services, which is reinforced in our site navigation. Oddly, we don't have problems post-coordinately displaying term combinations in database search results. Rather, in search results we display other terms from the subject taxonomy to narrow results by subject. The problem we have is with the legacy of hierarchical arrangements of access points organized by: collections, services, topics (this uses slices of the subject taxonomy). It's a very library-centric view that we've been dealing with changing, and if you ever worked in a library (corporate, private, special or public) you might know how difficult it is create this type of change.
I've pointed out that the concept of surfacing more facets of index terms would be helpful for browsing. Jim Anderson at Rutgers helped me to buy into this idea while I was in library school, and before I knew much about the web, I advocated this idea in an image index I proposed in 1997. That naive and over-ambitious Filemaker Pro screen shows how I envisioned it. It's funny. Today, I'm wondering how we can support the display of polyhierachical classifications such as our subject taxonomy and other database fields. We have some ideas floating around, but I feel like a toddler trying to topple an elephant.
Some follow-up. We're kicking around the idea of a) showing multiple breadcrumbs, and b) showing local navigation for one of the hierarches where the node exists. With the local navigation, we're going to check where the user came from in order to determine which tree to show. If they came from a bookmark or an email (most of our pages are also lined to from email alerts) we will show nothing, unless the node only has one parent, then we will show that tree. This is the theory. We need to test, but interested in opinions. Have you done something like this in a better way?
webgraphics is discussing the touchscreen interface used in the Georgia elections this week. The UI is simulated on the Georgia site for your clicking pleasure (or pain). James found a related on article on Wired, High-Tech Voting Gets Thumbs Up.
Chad and Tanya pointed to Important Works for Web Navigation, David Danielson's annotated bibliography of essential, foundational literature for the study of web navigation. Danielson has published some other HCI work related to web navigation behavior and design completed during his Masters program at Stanford.
Braunarts' 3d music (requires Shockwave plugin) is an interteractive performance that blends music and a zoomable interface to create a 3 dimensional environment in which people explore the musical compositions. Interesting, but somehow, I feel uncomfortable in 3d or ZUI web environments like this. It's funny, because I used to play video games that rendered space in 2d and 3d and felt comfortable enough in those spaces, knowing that there was a goal to arrive at -- destroying the Death Star or getting around that pylon to shoot a tank down -- but exploring 3d spaces with ZUI's on the web just seems so slow and boring to me. Somehow something gets lost for me in the translation of the experience from the gaming world.
Victor and Joshua are both talking about story telling as a method for communicating possible actions or paths when interacting with web sites. Victor mentioned an IBM seminar he attended about story telling. Haven't done much reading in this area and would cetainly like to learn more if I can find the most salient literature. I did find Curt Cloninger's A List Apart article, a Case for Story Telling to be interesting as well. Cloninger makes the case for considering the narrative possibilities when designing for the web as a communications medium. He's right, web sites are often not just databases and the design should consider aspects of human experiences with sites not merely as transactional database interactions as such. I like Victor's process of mapping actions or attributes of the narrative to interactions with the site. Interesting. More obvious I guess is the development of the characters, plot, setting, etc. and flowing that into elements of design process -- personas/characters, scenarios.
Adaptive Path is cornerning the market on IA articles and mind share lately. Superstars write a lot.
Progress Paralysis: Eight steps to get your Web site moving again by Peter Merholz in New Architect.
The Culture of Usability: How to spend less and get more from your usability-testing program by Janice Fraser in New Architect.
Site Navigation: A Few Helpful Definitions by Indi Young in Adaptive Path Publications.
Jef Raskin's The Humane Evironment has been making the rounds. Works on OS X. I wonder if anyone has checked it out using CVS and installed. Haven't read anyone's observations yet.
The Matts (Jones and Webb) pointed to the Spring desktop for Mac OS X, an alternative interface for navigating the stuff on/with your computer.
If you're into wayfinding, design, and labels then Public Lettering: a walk through central London is an charming tour through typography in public spaces.