Analysis of a pointing tasks on a white board

We study the variations in two dimensional (2D) pointing behaviour of a group of subjects by means of capturing their movement traces in an automatic way with the Mimio device. Such traces can provide detailed insight in the variability of 2D pointing relevant for example for the design of computer vision based gestural interaction. This study provides experimental evidence that for medium large distances Fitts' model, and Welfords and Shannons variants, continue to show a linear relationship between movement time (MT) and the index of difficulty (ID) with a high correlation for the ranges considered. The expected increased sensitivity to changes in ID for these distances are confirmed. Nearly all movements show three phases: a planning phase, a ballistic phase and an adjustment phase. Finally, we show that the arrival time at the target resembles a log-normal distribution.