An Extended Framework for Measuring the Information Capacity of the Human Motor System

Fitts' law is a fundamental tool in understanding the capacity of the human motor system. It measures information throughput in terms of the tradeoff between the speed and accuracy of motor responses. Although immensely popular, the paradigm in which Fitts' law is the principal keystone is confined to relatively simple responses in strictly prescribed stimulus-response conditions. Our goal is to generalize the framework into completely unconstrained movement. The proposed new metric is based on a subject's ability to accurately reproduce a learned movement pattern. It can accommodate recorded movement of any duration and composition, and involving contributions of any part of the body. We demonstrate the proposed method by analyzing publicly available motion capture data. Possible applications include human-computer interaction, sports science, and clinical diagnosis.