ABSTRACT

Human motion is inherently variable while statistics involves the study of variability. So it is natural that statistics should be applied to the study of human motion. Of course, statistical methods are widely used in ergonomics and human factors research, but they are limited in a certain way. Consider quantities such as the time taken to complete a task, the angle formed at the knee in a static posture, and the maximum velocity of the hand during a reach. Th ese are all univariate, scalar quantities for which there is a large repertoire of statistical methods, many of which are available in standard statistical soft ware. Th is chapter assumes the reader has some knowledge and experience with these textbook statistical methods, but we want to present here some more advanced techniques particularly appropriate for modeling motion.