Variability of throughput in pointing device tests: button-up or button-down?

Object selection and activation in graphical user interfaces is usually connected to the changes in the state of the buttons on the mouse. In input device testing the position of the mouse cursor at the time of button-down or button-up events is used. We measured the effect the choice of event has on throughput in a standard Fitts' paradigm pointing device test. 12 participants used two mice with button-down and button-up events as the target selection trigger. Button-down events produced 0.2 bps higher throughput than button-up events. Because of this, it is important to choose the relevant event to be used in pointing device tests and to report the choice along with the results. Otherwise results may be misleading and inter-study comparisons of throughput values may be inaccurate.