Over 30 years of using functional data analysis in human movement. What do we know, and is there more for sports biomechanics to learn?

Abstract

Functional data analysis (FDA) is a contemporary area of statistics designed for analysis of functions or curves. FDA has grown in human movement applications over the last three decades, with it being applied across a range of sport applications including rowing, weightlifting, diving, race-walking, jumping and running. Functional principal components analysis (fPCA) has been the most commonly used technique in sports biomechanics, often being applied to better understand characteristics of variability present in curves from biomechanical variables sampled from sporting movements. Given that FDA is an area of statistics with specific techniques for processing and analysing data, it provides one valuable platform for biomechanists to understand and think about their data more holistically. Further, the visual interpretability that FDA techniques provide, there is great potential for FDA to be used beyond research contexts, as a suite of practical tools to assist practical sports biomechanists in making decisions in sport. This review aims to demonstrate some methods yet to be applied in sports biomechanics, with simple sports biomechanics data applications taken from rowing. This article aims to showcase the value that FDA may have in assisting practitioners as they make decisions with athletes regarding their movement characteristics.