I completed the MS in HCI at Depaul with my last class being an independent study to build a tool for testing/researching navigation and how people acquire/find information from them. Actually, I have been researching this for the last year on and off so this was in a way more of a culmination of thoughts and reading. The main outcome was the development of a web based tool that could be used to present tasks to a user and collect data as they interact with a menu navigation system. Ultimately, the project was a continuation of Dr. Craig Miller’s research into how users conceptually navigate a hierarchical menu structure. He had developed several prototypes that used java and flat text files to present the data.

As with most academic endeavors of this kind I, the student, started with a fairly broad goal and needed to narrow it down. However, my goal was not exactly the same as Dr. Miller’s as I wanted and still want a general purpose tool that could be used for a variety of research activities related to testing website navigation and usability. Where as he and another researcher are seeking to better understand the conceptual model of users as they learn a menu and find information organized in that menu. For me, my practical work experience has lead me to see the need for an easy to use tool to help better organize/prepare complex menus for websites, especially after a site has launched. So I am first assisting Dr. Miller and the other researchers in developing a tool that will work for their testing, but also I am trying to develop it in a way that I can use to do basic testing of web usability in general.

I have produced a prototype tool and released it as an open source project. It can be found at the address below. It is in alpha (or even pre-alpha) phase so any recommendations, thoughts, or contributions will be welcomed.

UI Nav Test Tool

The project was written in PHP and uses a MySQL back-end. There is a demo and you can download the source plus the setup files from the site. If you have questions you can contact me via the project site. Essentially, the UI Nav Test tool is a very stripped down CMS with features for presenting tasks to a research participant and it also has a basic reporting page and allows the researcher to download the raw data for further analysis. While the appearance is very basic, this is intentional. As the purpose is to identify problems with labels and organization, not whether something looks pretty or not. The demo is set to use one of Dr. Miller’s content domains which is focused on understanding how people conceptualize the navigation steps they will take to find items.

I just finished the capstone course. Normally the last class to be taken, but due to it only being offered one time a year, I have one remaining course to complete my masters degree in HCI. It was a tough quarter in part due to a family problem, but also the from amount of work that was needed. I had to go to Japan for a week about midway through, but I was able to finish all the work required. We had to write 3 papers reviewing published research/articles by well known HCI practitioners as well as conduct our own research project and do a write up.

The first review was of a writing by Bruce Tognizzini. Tognazinni is part of the Neilsen, Norman group and a former Apple employee. Several years ago he put out an article called “A Quiz designed to give you Fitts”. In the article he goes over questions that every HCI expert should be able to answer. Well you also have to be a Mac user and being an engineer would probably help (note having a degree in HCI does not necessarily make you an engineer). He uses the article to basically lament the passing of the old Mac OS and how superior it was to the new Mac OS X (and of course Windows). To which I have to say, he is full of carp (sic). While I love Macs and started out in System 5, Mac OS X is light-years beyond the previous OS. He is also on the record as hating the dock (but that is another article), which I also love. So while I respect him for his contributions to the GUI world, I think his emphasis on Fitts is a throw back to mechanical engineering and is not entirely applicable to the web today.

Here is the full review of Tog: Tog: Week 1 [pdf 103k]

The second review was of Chapter 3 from Jeff Raskin’s “The Humane Interface”. Raskin another Apple alum, outlines the need to for modeless interfaces and that he believes there is no such thing as an expert user. Again I can respect the man and his great contributions, but we are given utopian HCI platitudes, that haven’t gained traction despite being known for a decade or more (which is a long time in tech). I completely disagree with not differentiating users and providing different ways to do things. To me this is akin to the idea that all learners learn the same way, which has been shown untrue. While certainly nobody probably knows every feature of a program (except the developers), they can be experts in the use of specific features. Which is in my opinion enough to warrant accommodation different from beginners. Learnability is a key principle of usability and should take into account that people learn in different ways.

Here is the full review of Raskin: Raskin: Week 2 [pdf 106k]

I will wrap up in another post on my final project later….