Performance in one-, two- and three-dimensional terminal aiming tasks.

Arm movement times were measured to targets with independent constraints (target sizes) in one, two or three directions (width, height and depth). In each case, modified forms of Fitts' law give a good fit to the data, with the best form being dependent on all constraints in the 'Weighted Euclidean' model of Accot, J. and Zhai, S., 2003. (Refining Fitts' law models for bivariate pointing. Proceedings of the SIGCHI conference on Human factors in computing systems CHI '03, 5-10 April 2003, Ft. Lauderdale, FL. New York: ACM Press, 193-200). The best predictor includes all the individual ratios of amplitude of movement to target dimension related to the various constraints, with the maximum portion of variance accounted for by the index of difficulty (ID) in the direction of motion. The suggestion by various authors that the maximum of the ID associated with the different constraints will determine the movement times is only partially successful. An explanation for the results is given in terms of control strategy and the form of target. STATEMENT OF RELEVANCE: In many tasks, workers move objects into limited spaces that allow only a single entry area. Termination of the movement may have constraints in one, two or three dimensions. Movement times for such tasks are important to predict industrial performance, and guide task designers in reducing physical constraints on performance.