The Illusion of Usability

Minard2

Usability is a key topic and frequently thought of from the perspective of the human-technology interface. How easy is your oven to use? Is your exercise bracelet elegant, but hard to put on your arm? The illusion of usability is related to my previous post, since usability depends on the person using it. There is no such thing as generalized usability. Let's take a famous graphic--shown above--created by Charles Minard. This graphic is a map of Napoleon's disastrous Russian campaign of 1812. This is a visually striking map and elegantly shows a number of variables (6 types). The diagram is probably adorning walls and in coffee table books everywhere. According to a Wikipedia article, "modern information scientists say the illustration may be the best statistical graphic ever drawn." If your expertise is statistical graphics, knowledge of such a map is undoubtedly of significant academic use. What if you just wanted to know how many soldiers died during the walk to Moscow and back? What if you were not interested in knowing the temperature, or the lat/lon geographic coordinates? Perhaps you'd rather experience what it was like in 1812 during the slow frozen march? Representations are only as good as the explicit, and unique, purpose required by a viewer. An eight year old might do better with a verbal description, or a reproduction of a painting that is not too graphic. I attended a talk recently, where the speaker spent some time crafting a story using a slide containing Minard's graphic--the illusion was that the image was telling the story, when in fact, one had to consider the medium being used, the speaker's voice, the elements left in, and left out of the story. The image itself has no story to tell; one requires human interaction to achieve that effect--it was his theatrical performance that brought Minard's graphic to life. Models are just like Minard's masterpiece--you have to design them very carefully for one specific type of user.  Ideally for a single person. Is there a simulation model equivalent to Minard's graphic? The notion that one size fits all is false at best, and dangerous at worst.