Time Prediction of Mouse-based Cursor Movements

SUMMARY This paper addresses the prediction of movement time of mouse-based cursor transfers between objects in a graphical user interface. We present the data from an experimental study aimed at obtaining an appropriate model for movement time prediction. As can be seen from the data, Fitts’ law which is widely used as a predictor for such movements fails with our data. This is not only for low values of Fitts’ index of difficulty as known from the literature, but also for small target areas like radio buttons, combo boxes, and buttons on toolbars such as typically occurring with graphical user interfaces. We present a new power model for movement time prediction derived from our data and compare this model to other models from the literature.