Fitts ’ Throughput and the Remarkable Case of Touch-Based Target Selection

Abstract. The method of calculating Fitts ’ throughput is detailed, considering task characteristics, the speed-accuracy trade-off, data collection, and data aggregation. The goal is to bring consistency to the method of calculation and thereby strengthen between-study comparisons where throughput is used as a dependent measure. In addition, the distinction between indirect and direct pointing devices is elaborated using the examples of a mouse as an indirect pointing device and a finger as a direct pointing device. An experiment with 16 participants using a smart phone was conducted as an empirical test of direct touch-based target selection. Overall, the throughput was 6.95 bps. This is a remarkable figure – about 50 % higher than accepted values for a mouse. The experiment included task type (1D vs. 2D) and device position (supported vs. mobile) as independent variables. Throughput for the 1D task was 15 % higher than for the 2D task. No difference in throughput was observed between the supported and mobile conditions