Michael John MCGuffin
Fitts' law is a model of pyschomotor performance that was introduced in 1954 and has been verified in hundreds of subsequent studies. Fitts' law allows us to predict, for example, the time it takes a user to click on an on-screen target (a target could be, for example, an on-screen button). One of the basic principles revealed by Fitts' law is that large targets take less time to click on. From a design perspective, this implies a tradeoff between packing lots of small buttons into a single screen, and making buttons large so they can be selected faster.
An interesting idea to avoid this tradeoff is to have buttons that are initially small (so a large number of them can fit into a screen), and to expand a desired button to a larger size whenever the user wants to select it. However, from previous knowledge of Fitts' law and motor control theory, it was not known whether this scheme would really work. In particular, it was not known whether a user's visual and motor systems would be able to take split-second advantage of an enlarged target after the user had already started to move towards it.
We conducted experiments (described in Chapter 3) to determine how quickly users can click on "expanding targets", i.e. targets that grow as the user approaches them. We found that users were able to click on these targets faster than on targets that do not grow, and that the advantage in performance is approximately as large as one could possibly expect, given Fitts' law. However, in our experiments, the users only interacted with a single expanding target at a time.
If multiple expanding targets are present on the screen, problems arise. For example, when one target expands, it may cover up or hide other nearby targets, making them harder to click on. Our challenge, then, is to use expansion to make each of a set of targets easier to click on, without incurring a net penalty from making other targets harder to acquire. We propose many design strategies (in Chapter 4) for dealing with this problem. Some of our strategies are based on new methods of mathematical analysis that we developed.
The results of this thesis have implications for psychomotor theory and the design of graphical user interfaces.