A new way to analyse computer mouse trajectories
It is based on an information-theoretic framework, where a set of entropy-based measures is used to quantify the spatial information in the empirical data. Movement features present in empirical trajectories, such as location, direction, and space exploration are extracted adopting an original entropy decomposition.
EMOT offers a robust methodology in case of noisy trajectories due to uneven movements of the cursor and can work with raw x-y trajectories registered with any recording software.
Publications
- Calcagnì, A., Lombardi, L., & Sulpizio S. (2017). Analysing spatial data from mouse tracker methodology: An entropic approach. Behavior Research Methods. In press [download]
Getting EMOT
Conditions: The EMOT package is offered free of charge to anyone interested in using the EMOT methodology provided that it is not used for financial profit and that you properly cite the software and the EMOT paper (see above). Requests to use the software for any commercial purpose must be directed to the authors. You are not allowed to redistribute a downloaded copy of the software to others. If you want others to use the EMOT package, please refer them to this website.
EMOT package v. 1.0 [05/01/2017]
About
EMOT analyses the spatial information of computer-mouse trajectories conveyed by cognitive processes in choice/decision tasks. Unlike other methodologies, EMOT extracts dynamic features of the hand-movement by looking at trajectories in terms of fast movements and motor pauses.
In a two-choice categorization task
hand-movements reveal several scenarios involving different
cognitive processes. Fast movements represent
those motor sub-processes executed after a decision has been made.
By contrast, motor pauses unfold sub-processes
related with decisional conflicts in categorization, decisional
uncertainty, goal formulation and/or reformulation.
EMOT analyses trajectories in a single-trial fashion.
The noisy x-y trajectory is initially transformed into distances
and angles. Next, movement features like location, directions, and
amplitudes are modeled in terms of a histogram model.
The histogram model is used to determine a quantification
of spatial events involved in the original movement
path (e.g., competition between competing and target cues,
attraction of competing cue) in terms of fast movements and motor
pauses. The quantification is solved adopting a Non-linear
Programming (NLP) approach with a Cumulative Residual Entropy
(CRE) decomposition.
Contact
For further information, requests, feedback, or reporting bugs, please contact the manteiner:
Antonio CalcagnìDepartment of Psychology and Cognitive Sciences, University of Trento (Italy)
Email: ant [dot] calcagni [at] gmail [dot] com
Phone: +39 0464 808644