Human ingenuity has developed a myriad of ways in which information can be encoded for later retrieval. Information encoding strategies differ widely in terms of how well they serve their function. Dimensions by which such strategies can be measured include expressiveness and effectiveness.

Expressiveness is how well a display encodes all and only the desired relations among data. Effectiveness, on the other hand, is a measure of how well (rapidly and accurately) a graphical display is decoded by a human observer.

The animation project explores why animated displays are sometimes more effective at presenting information than static displays. We show subjects several different static displays, each one, in some sense, informationally equivalent to a target animated display, and we have been studying the performance differences that emerge.

In a replication of an experiment by Kaiser and Proffitt, subjects saw either animations or diagrams (with textual descriptions) of a pendulum, and were asked to rank the naturalness of five different outcomes, given that the cord of the pendulum was severed when the bob was either at its apex or its nadir. If subjects saw the diagram first, they tended to answer incorrectly, and if presentation of the diagram was followed by presentation of the animations, they would then answer correctly. However, when the order of presentation was reversed, subjects would tend to answer correctly after seeing the animations, but then when asked to answer the same questions immediately afterwards, given a diagram, they would answer incorrectly. Why the insights lent by animation did not transfer is one of the most vexed questions raised by this line of research.