definition of t(move) and t(final)

[lkeegan]

We just realized that we have problems with the calculation/definition of t(move) and t(final). It seems that t(move) is defined as any movement even if the movement is very small. For example, if “fixed target display interval” or “delay before target display” are selected the participants make very small cursor movements before the target is displayed and while the cursor is not moved to the target. It seems that this very small movement is defined as t(move). t(final) is currently defined as time when the cursor reaches the target (or timeout occurs). Parameters like reaction time or movement time could be incorrectly calculated because when participants reach the target there is always a small cursor movement as it is not possible to immediately stop the movement, or the participant makes small movements before the next (outer) target is displayed. It would be good if t(move) is defined as the start of the movement that aims to reach the target and t(final) as the point of movement reversal/stop (if the target is reached in time t(final) should be shortly after target reach). According to similar research/motor learning tasks this could be done, for example, by using the cursor velocity profile. The cursor typically has a bell-shaped temporal profile of movement velocity (peak velocity occurring around 50% of the movement time) when performing a single out or in reaching movement. Similar to similar tasks, t(move) and t(final) could be calculated backwards from peak cursor velocity. Using this calculation t(move) would be defined as point when velocity is minimal (in some cases zero) before peak velocity. t(final) would be the point of movement reversal calculated as minimal velocity (in some cases zero) after peak velocity. There is only one exception, which is when the person makes movement corrections. In this case, there are several “bellies” in the velocity curve. This method of using the velocity profile would also mean that in some cases the cursor movement before t(move) (or after t(final) and before t(move)) is not considered for the calculation of the statistics (the raw data should still be saved in the x and y cursor position per timestamp export, but not used for the calculation of statistics). It would be great to also use these definitions of t(move) and t(final) as start and end points for the calculation of all other parameters (i.e. distance, RMSE, area, peak velocity, and peak acceleration).