Design of 2D Fitts' law experiments: An odd thing about targets

We explored the ISO 9241-411 multi-directional tapping (2D) test for evaluating computer pointing techniques. Based on Fitts' law, the test evenly distributes targets around a circular layout and requires the same movement distance between targets for all trials in a sequence. We demonstrate that this is only possible when using an odd number of targets. We reviewed 200 research papers using the 2D task and found that about a third of them use an even number of targets, and consequently report incorrect values for throughput, reinforcing the need to advertise this overlooked but relevant issue. Our findings imply that adhering to an odd number of targets and to the correct definition of movement distance ensures methodological consistency, leading to valid throughput values. Consequently, we argue for revisiting and updating the ISO 9241-411 standard to explicitly state that an odd number of targets is required for the 2D task.