Trajectory Metrics¶
This section covers various trajectory metrics used to analyze the users trajectory, such as path length or AUC, which provide insights into movement efficiency and accuracy during user interactions in a mathematically rigorous manner.
Path¶
The path or total distance traveled during user interactions is a key metric for analyzing movement efficiency [Katerina and Nicolaos, 2018, Rhim et al., 2023, Kieslich et al., 2019, Seelye et al., 2015].
The function path() calculates the path length for interaction points from a DataFrame or session traces.
This function computes the path length based on the Euclidean distance between consecutive points.
The distance between consecutive points is calculated as:
AUC Ratio¶
The Area Under the Curve (AUC) is a metric that quantifies the overall movement efficiency during user interactions in computers [Katerina and Nicolaos, 2018, Kieslich et al., 2019] or even in mobile devices [Rhim et al., 2023].
The function auc_ratio() calculates the AUC ratio for each session.
That is, the real area under the curve and the optimal area (straight-line) and returns the ratio between their difference and the rael area.
Given the real area \(A_{real}\) and the optimal area \(A_{opt}\) the AUC ratio is computed as:
This small epsilon prevents division-by-zero when the optimal area is 0.
Metrics are also provided via auc_ratio_metrics(), which computes mean, max, and min AUC and AUC ratio for each session.
Maximum Absolute Deviation¶
The Maximum Absolute Deviation (MAD) is a metric that quantifies the maximum deviation of the actual movement path from the optimal straight-line path during user interactions [Katerina and Nicolaos, 2018, Rhim et al., 2023, Kieslich et al., 2019].
The function deviations() calculates the MAD for each session or interaction trace, returning both the mean, maximum and minimum MAD values.
The MAD is calculated as the maximum perpendicular distance from any point on the actual path to the straight line connecting the start and end points of the movement.
Diagram illustrating Maximum Absolute Deviation (MAD)¶
Average Absolute Deviation¶
The Average Absolute Deviation (AAD) is a metric that quantifies the average deviation of the actual movement path from the optimal straight-line path during user interactions [Rhim et al., 2023].
It consists of the average value of the perpendicular distances from each point of the trajectory to the straight line connecting the start and end points of the movement (optimal trajectory).
The function deviations() calculates the AAD for each session or interaction trace, returning the AAD value along with MAD metrics.
Diagram illustrating Average Absolute Deviation (AAD)¶
References¶
Tzafilkou Katerina and Protogeros Nicolaos. Mouse behavioral patterns and keystroke dynamics in End-User development: what can they tell us about users' behavioral attributes? Comput. Human Behav., 83:288–305, June 2018. URL: https://doi.org/10.1016/j.chb.2018.02.012.
Iftikhar Ahmed Khan, Willem-Paul Brinkman, Nick Fine, and Robert M Hierons. Measuring personality from keyboard and mouse use. In Proceedings of the 15th European conference on Cognitive ergonomics: the ergonomics of cool interaction. New York, NY, USA, January 2008. ACM.
Jungwook Rhim, Jiwon Kim, and Gahgene Gweon. Using geometric features of drag-and-drop trajectories to understand students' learning. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, volume 19, 1–14. New York, NY, USA, April 2023. ACM. URL: https://dl.acm.org/doi/10.1145/3544548.3581143.
Masud Karim and Md Hasanuzzaman. A study on mouse movement features to identify user. Sci. Res. J., 08(04):77–82, April 2020.
Dirk U Wulff, Pascal J Kieslich, Felix Henninger, Jonas M B Haslbeck, and Michael Schulte-Mecklenbeck. Movement tracking of psychological processes: a tutorial using mousetrap. Behav. Res. Methods, 57(11):307, October 2025.
Martin Schoemann, Malte Lüken, Tobias Grage, Pascal J Kieslich, and Stefan Scherbaum. Validating mouse-tracking: how design factors influence action dynamics in intertemporal decision making. Behav. Res. Methods, 51(5):2356–2377, October 2019.
Eduard Kuric, Peter Demcak, Matus Krajcovic, and Peter Nemcek. Is mouse dynamics information credible for user behavior research? an empirical investigation. Comput. Stand. Interfaces, 90(103849):103849, August 2024.
Pascal J. Kieslich, Felix Henninger, Dirk U. Wulff, Jonas M. B. Haslbeck, and Michael Schulte-Mecklenbeck. Mouse-tracking: a practical guide to implementation and analysis. In A Handbook of Process Tracing Methods, pages 20. Routledge, 2nd edition edition, 2019. URL: https://www.taylorfrancis.com/chapters/edit/10.4324/9781315160559-9/mouse-tracking-pascal-kieslich-felix-henninger-dirk-wulff-jonas-haslbeck-michael-schulte-mecklenbeck.
Christian O'Reilly and Réjean Plamondon. Can computer mice be used as low-cost devices for the acquisition of planar human movement velocity signals? Behav. Res. Methods, 43(1):229–238, March 2011. URL: https://link.springer.com/article/10.3758/s13428-010-0030-3#notes.
Adriana Seelye, Stuart Hagler, Nora Mattek, Diane B Howieson, Katherine Wild, Hiroko H Dodge, and Jeffrey A Kaye. Computer mouse movement patterns: a potential marker of mild cognitive impairment. Alzheimers Dement. (Amst.), 1(4):472–480, December 2015. URL: https://alz-journals.onlinelibrary.wiley.com/doi/10.1016/j.dadm.2015.09.006.
Beatriz Pariente-Martinez, Martin Gonzalez-Rodriguez, Daniel Fernandez-Lanvin, and Javier De Andres-Suarez. Measuring the role of age in user performance during interaction with computers. Univers. Access Inf. Soc., 15(2):237–247, June 2016. URL: https://link.springer.com/article/10.1007/s10209-014-0388-6.
Clayton Epp, Michael Lippold, and Regan L Mandryk. Identifying emotional states using keystroke dynamics. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems. New York, NY, USA, May 2011. ACM.
Preeti Khanna and M Sasikumar. Recognising emotions from keyboard stroke pattern. Int. J. Comput. Appl., 11(9):1–5, December 2010.
Lisa M Vizer, Lina Zhou, and Andrew Sears. Automated stress detection using keystroke and linguistic features: an exploratory study. Int. J. Hum. Comput. Stud., 67(10):870–886, October 2009.
A F M Nazmul Haque Nahin, Jawad Mohammad Alam, Hasan Mahmud, and Kamrul Hasan. Identifying emotion by keystroke dynamics and text pattern analysis. Behav. Inf. Technol., 33(9):987–996, September 2014.
Maarten Dijkstra. The diagnosis of self-efficacy using mouse and keyboard input. PhD thesis, Utrecht University, 2013.