Cross-comparison of the performance of discrete, phase and functional data analysis to describe a dependent variable

The aim of this study was to assess and contrast the ability of discrete point, functional principal component analysis (fPCA) and analysis of characterizing phases (ACP) to describe a dependent variable (jump height) from vertical ground reaction force curves captured during the propulsion phase of a countermovement jump. A stepwise multiple regression analysis was used to assess the ability of each data analysis technique. The order of effectiveness (high to low) was ACP, fPCA and discrete point analysis. Discrete point analysis was not able to generate strong predictors and detected also erroneous variables. FPCA and ACP detected similar factors to describe jump height. However, ACP performed better than fPCA because it considers the time and magnitude domain separately and in combination and it examines key-phases, without the influence of non-key-phases.
© Copyright 2013 ISBS - Conference Proceedings Archive (Konstanz). Springer. Published by International Society of Biomechanics in Sports. All rights reserved.

Bibliographic Details
Subjects:
Notations:training science technical and natural sciences strength and speed sports
Tagging:reaktiver Sprung
Published in:ISBS - Conference Proceedings Archive (Konstanz)
Language:English
Published: Taipei International Society of Biomechanics in Sports 2013
Online Access:https://ojs.ub.uni-konstanz.de/cpa/article/view/5531
Volume:31
Issue:1
Document types:congress proceedings
Level:advanced