Fitts’ Law in the Wild: A Field Study of Aimed Movements

This paper presents the first field study of aimed movements in graphical user interfaces, designed to get better insight into pointing in the real (electronic) world and assess the validity of Fitts’ law in the wild. We unobtrusively collected kinematic data from 24 users over several months, and segmented it into a table of 2 million movements. We show that Fitts’ law is indeed robust for modeling pointing performance, provided that an adequate noise reduction process is applied. We introduce the length-distance index (LDI) to take into account the fact that many movements are not straight, and we introduce an extension of Fitts’ law that includes an LDI term. We also show evidence of the effect of cognitive tasks in pause and click time, and the sensitivity of differences in performance, e.g., across input devices, to LDI. Altogether, these findings provide a firm ground to better assess the validity of results obtained in the laboratory. ACM Classification