Robertson, Samuel ORCID: 0000-0002-8330-0011 (2014) Lessons from a National Golf Program: Performance Testing Considerations for Talent Identification and Development. In: Golf science: a World Scientific Congress of Golf, 05 September 2014-07 September 2014, Gold Coast, Australia.

Abstract

Purpose: Performance tests are commonly used within talent identification (TID) and long term athlete development (LTAD) golf programs. However, debate exists regarding the usefulness of such tests due to considerations relating to their design, measurement properties and suitability for use with adolescent players. In this presentation, results from a series of projects undertaken between 2011 and 2014 which utilised participants from the Australian National Golf Squad are outlined. Specifically, this research focused on: a) Developing a performance testing battery with the intention of better informing the identification of talented junior Australian golfers, and b) Investigating the relationships between player results in this battery and their data obtained from actual tournament performance. Methods: A range of questions relevant to both TID and LTAD of golfers were identified throughout the undertaking of this research and the design of the battery. These questions included: a) When assessing adolescent golfers, what is the ideal number of trials required in a test in order to obtain a ‘true’ indication of their ability? b) How can the ecological validity/representative design of performance tests be optimised, without experiencing subsequent reductions in test-retest reliability? c) Can golf performance tests be designed that are able to accurately predict future player performance? Results: To date, seven peer-reviewed publications have resulted from these projects. Based on this work proposed solutions to the questions stated above are provided: a) Monte Carlo simulation techniques are a novel statistical method of modelling repeated test performance a priori. Even when implementing familiarisation procedures, more than 5 repeated trials may be needed when testing adolescents in order to obtain indicative performance . Th is recommended in order to avoid player non- or de-selection errors in TID settings. b) By designing tests that i) replicate conditions experienced in competition and ii) utilise contextual interference techniques, the predictive validity of a performance test battery can be optimised. This also can allows for the battery to be used to inform player practice specificity as well as monitor longitudinal player development. c) Utilising non-linear and data mining statistical analysis techniques such as decision trees and support vector machines allows for relationships between testing and athlete performance to be better understood . In sports such as golf, this allows for the identification of multiple players profiles and individual rates of development to be better understood and catered for in coaching and practice scheduling. Further, decision tree output has the advantage of being easy to interpret by players and coaches. Discussion: A framework for assessing the methodological quality of performance tests is provided and discussed in the context of results stemming from this body of research. Practical Application/ Clinical Relevance: The challenges discussed in this presentation are relevant for those designing or adapting performance tests in structured golf TID and LTAD programs. Through consultation of published, valid frameworks and guidelines such as that provided here, the efficacy of golf TID and LTAD programs can be optimized.

Search Google Scholar

Repository staff login