Logo image
Back
A novel method to guide classification of para swimmers with limb deficiency
Journal article   Peer reviewed

A novel method to guide classification of para swimmers with limb deficiency

Luke Hogarth, Peter Van de Vliet, Mark Connick and Brendan J Burkett
Scandinavian Journal of Medicine & Science in Sports, Vol.28(11), pp.2397-2406
2018
DOI: https://doi.org/10.1111/sms.13229
Appears in  UniSC Diversity and Inclusion Research Collection
url
https://doi.org/10.1111/sms.13229View
Published Version

Abstract

disability freestyle machine learning paralympic partial least squares UniSC Diversity Area - Disability and Inclusion
Abstract The International Paralympic Committee has directed International Federations that govern Para sports to develop evidence-based classification systems. This study defined the impact of limb deficiency impairment on 100 m freestyle performance to guide an evidence-based classification system in Para Swimming, which will be implemented following the 2020 Tokyo Paralympic games. Impairment data and competitive race performances of 90 international swimmers with limb deficiency were collected. Ensemble partial least squares regression established the relationship between relative limb length measures and competitive 100 m freestyle performance. The model explained 80% of the variance in 100 m freestyle performance, and found hand length and forearm length to be the most important predictors of performance. Based on the results of this model, Para swimmers were clustered into four-, five-, six-and seven-class structures using nonparametric kernel density estimations. The validity of these classification structures, and effectiveness against the current classification system, were examined by establishing within-class variations in 100 m freestyle performance and differences between adjacent classes. The derived classification structures were found to be more effective than current classification based on these criteria. This study provides a novel method that can be used to improve the objectivity and transparency of decision-making in Para sport classification. Expert consensus from experienced coaches, Para swimmers, classifiers and sport science and medicine personnel will benefit the translation of these findings into a revised classification system that is accepted by the Para swimming community.

Details

Metrics

3 File views/ downloads
926 Record Views
22 Times Cited - Web of Science

InCites Highlights

These are selected metrics from InCites Benchmarking & Analytics tool, related to this output

Collaboration types
Domestic collaboration
International collaboration
Web Of Science research areas
Sport Sciences
Logo image