Ruth Rosenholtz of the department of Brain and Cognitive Science at MIT writes:
We mostly do computational modeling of human vision. We try to do on the one hand the sort of basic science that fits in the human vision community, while on the other hand developing predictive models which might actually lend insight into design.
Your talk resonated with me in part because of this paper [Do Predictions of Visual Perception Aid Design?, by Ruth Rosenholtz, Amal Dorai, and Rosalind Freeman]. We went into our study thinking that people would like to have a quantitative tool to help analyze designs. But what we concluded, somewhat anecdotally, was that its main use seemed to be as a conversation-starter, and a means of communicating ideas about the design. And the reason it seemed to work is that our visualizations were the right level of a “puzzle” — challenging enough to be a bit fun to work out.
On another topic, check out the infographic from last weekend’s NYTimes magazine. Where are the actors in “Nashville” actually from? One might think this should be shown geographically, though perhaps not in a 2″x2″ space. Well, there’s only 6 actors, why not just use a table? Apparently, too boring. So they made a table (two columns: actor, where they are from), then scrambled the first column. Then connected names to locations with colored line segments. I said to my husband, “why would anyone do this?” He said, “I guess it makes you have to spend time to puzzle it out.”
Recall our recent discussion [link fixed] of the confusing and visually-appealing Christmas tree graph.