Implementation of path analysis and piecewise structural equation modelling to improve the interpretation of key performance indicators in team sports: An example in professional rugby union.

Novak, AR; Impellizzeri, FM; Garvey, C; Fransen, J

Implementation of path analysis and piecewise structural equation modelling to improve the interpretation of key performance indicators in team sports: An example in professional rugby union.

Novak, AR Impellizzeri, FM Garvey, C Fransen, J

Permalink

Publisher:: TAYLOR & FRANCIS LTD
Publication Type:: Journal Article
Citation:: Journal of sports sciences, 2021, pp. 1-8
Issue Date:: 2021-06-20

Closed Access

	Filename	Description	Size
	Implementation of path analysis and piecewise structural equation modelling to improve the interpretation of key performance indicators in team.pdf		1.03 MB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Novak, AR
dc.contributor.author	Impellizzeri, FM
dc.contributor.author	Garvey, C
dc.contributor.author	Fransen, J https://orcid.org/0000-0003-3355-1848
dc.date.accessioned	2021-11-09T04:57:15Z
dc.date.available	2021-11-09T04:57:15Z
dc.date.issued	2021-06-20
dc.identifier.citation	Journal of sports sciences, 2021, pp. 1-8
dc.identifier.issn	0264-0414
dc.identifier.issn	1466-447X
dc.identifier.uri	http://hdl.handle.net/10453/151449
dc.description.abstract	Analysis of key performance indicators (KPIs) in team sports has frequently involved multiple univariate analyses and modelling of direct associations between each KPI and match outcomes. This study aimed to show a more appropriate framework and modelling process to establish causal plausibility for future confirmatory studies. A cross-sectional design was adopted, using 337 team-match observations of Australian Super Rugby performances. A tentative model was developed in consultation with a domain expert (national analyst) and analysed using piecewise structural equation modelling. Model fit was assessed using Fisher's C and the Akaike Information Criterion (AIC). Hypothesised relationships were modelled using linear mixed effects models and unmodelled pathways were investigated using tests of directed separation. The model was an acceptable fit overall, and adjustments were identified in collaboration with the national head analyst, improving the AIC from 127.15 to 120.77 (Fisher's C = 66.78; p = 0.382). Modelling the hierarchical data structure and developing models that contain more logical hypothesised associations (in consultation with domain experts) is a more useful and important step to analyse and interpret effects of KPIs on team performance. This analysis provides support to the plausibility of the causal structure and generation of new and more precise hypotheses.
dc.format	Print-Electronic
dc.language	eng
dc.publisher	TAYLOR & FRANCIS LTD
dc.relation.ispartof	Journal of sports sciences
dc.relation.isbasedon	10.1080/02640414.2021.1943169
dc.rights	info:eu-repo/semantics/closedAccess
dc.subject	1106 Human Movement and Sports Sciences, 1302 Curriculum and Pedagogy
dc.subject.classification	Sport Sciences
dc.title	Implementation of path analysis and piecewise structural equation modelling to improve the interpretation of key performance indicators in team sports: An example in professional rugby union.
dc.type	Journal Article
utslib.location.activity	England
utslib.for	1106 Human Movement and Sports Sciences
utslib.for	1302 Curriculum and Pedagogy
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Health
pubs.organisational-group	/University of Technology Sydney/Strength - CHSP - Health Services and Practice
utslib.copyright.status	closed_access	*
dc.date.updated	2021-11-09T04:57:14Z
pubs.publication-status	Published

Abstract:

Analysis of key performance indicators (KPIs) in team sports has frequently involved multiple univariate analyses and modelling of direct associations between each KPI and match outcomes. This study aimed to show a more appropriate framework and modelling process to establish causal plausibility for future confirmatory studies. A cross-sectional design was adopted, using 337 team-match observations of Australian Super Rugby performances. A tentative model was developed in consultation with a domain expert (national analyst) and analysed using piecewise structural equation modelling. Model fit was assessed using Fisher's C and the Akaike Information Criterion (AIC). Hypothesised relationships were modelled using linear mixed effects models and unmodelled pathways were investigated using tests of directed separation. The model was an acceptable fit overall, and adjustments were identified in collaboration with the national head analyst, improving the AIC from 127.15 to 120.77 (Fisher's C = 66.78; p = 0.382). Modelling the hierarchical data structure and developing models that contain more logical hypothesised associations (in consultation with domain experts) is a more useful and important step to analyse and interpret effects of KPIs on team performance. This analysis provides support to the plausibility of the causal structure and generation of new and more precise hypotheses.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/151449