Additional outcomes/statistics

[PhilippWanner]

Outcome/Feedback - if possible adding the following outcome parameters/statistics:

• Peak velocity and peak acceleration
• Statistics at peak velocity: movement time, total time, movement distance, RMSE movement, and directional error (difference between the ideal trajectory and the real trajectory) at peak velocity
• Hand path area (only possible when the option “automatically move cursor to center” is deselected --> i.e. participants have to do out and back movements): the area enclosed by the hand path (i.e. out and back movements)
• Normalized hand path area : hand path area divided by the squared movement length
• Spatial error: the linear distance from the movement end point and the center of the target

[lkeegan]

Spatial error: at timeout linear distance from cursor to target minus target radius

[lkeegan]

First step:

• Implement some/all of these additional statistics in a Jupyter notebook starting with exported data from a trial
• Ideally this notebook would also include the existing statistics that are calculated & exported
  • These would be both tests of the existing implemented statistics and documentation/examples for users to start from when doing their own analysis

Next step:

• Add some/all of these to the statistics to those that are calculated, displayed and exported by the software
• Probably also need to restructure how the statistics are chosen & displayed for this
• Add definitions/description of each statistic to the documentation

[lkeegan]

More possible stats:

• Angle at peak velocity (directional error at peak velocity): The angle at peak cursor velocity compared to the target/center using vectors (e.g. https://www.geeksforgeeks.org/angle-between-two-vectors-formula/).
  For statistics to target: the angle between the vector A including the center of the central target to the position of the cursor at peak velocity and vector B including the center of the central target to the center of the outer target.
  For statistics to center: the angle between vector A including the center of the outer target to the position of the cursor at peak velocity and vector B including the center of the central target to the center of the outer target.
• Error at movement reversal (spatial error):
  For statistics to target: the shortest distance (scalar distance) of the outer cursor reversal point (i.e. movement reversal/endpoint) from the center of the outer target.
  For statistics to center: the shortest distance (scalar distance) of the inner cursor reversal point (i.e. movement reversal/endpoint) from the central target.
  The outer and inner movement reversal points could be calculated using the minimal velocity of the cursor movement after peak cursor velocity (according to the above-mentioned new definition of t(final))
• An option that can be selected in the statistics for smoothing and filtering data. These techniques shouldn’t be applied online (i.e. during task execution), but for the calculation of the statistics (especially for the statistics export). For example, a cubic spline data interpolation would be a good option for data smoothing and a low pass filter, for which the frequency/cutoff can be entered (e.g. in Hz), for data filtering.