Motor Learning Metrics for HCI Research

To quantify and understand how users learn movement-based techniques in HCI, we build four metrics based on motor learning literature. Refinement time proportion (RTP) and refinement space (RS) quantify the user prediction error. Normalized path error (NPE) and normalized jerk error (NJE) measure the closeness of the current movement to an ideal movement.

Paper

Difeng Yu, Mantas Cibulskis, Erik Skjoldan Mortensen, Mark Schram Christensen, and Joanna Bergström. 2024. Metrics of Motor Learning for Analyzing Movement Mapping in Virtual Reality. In Proceedings of the CHI Conference on Human Factors in Computing Systems (CHI ’24), May 11– 16, 2024, Honolulu, HI, USA. ACM, New York, NY, USA, 18 pages. https://doi.org/10.1145/3613904.3642354

How to Install

library(devtools)
install_github("Davin-Yu/MotorLearningMetrics4HCI")

or use our binary package in the repository

install.packages(path_to_the_.tar.gz_file, repos = NULL, type="source")

How to Start

1. Please read the paper to get an understanding of how the metrics work.
2. After installing the package in R, use help(package = "MotorLearningMetrics4HCI") to navigate the package. You can run the example codes included in the documentation.