PREDICTING PERFORMANCE FROM PHYSIOLOGICAL MEASURES
Watts, P., Clure, C., Hill, R., & Lish, A. (1996). Applied prediction of cross country skiing performance from physiological test data. Medicine and Science in Exercise and Sports, 28(5), Supplement abstract 794.
Physiological measures (derived from a skiing treadmill) when plugged into regression equations were used to predict performance times in classical and freestyle cross-country skiing races. A small homogenous sample of trained female skiers and a single criterion for performance were used.
Measures were used to predict 10 km freestyle and classic races and regression equations computed. Those equations were then used to predict a second set (5 km freestyle and 10 km classic) of races. There was a significant difference between predicted and actual race times in the second set of races. This indicated that physiological measures, when regressed against one set of criteria most likely will not be accurate for another set of criteria. The predictive validity of physiological measures for cross-country skiing competitive performance is low.
It was concluded that while regression equations may exhibit high multiple correlation valued and provide insights concerning important physiological characteristics, the applicability of such equations to predict further performance has limited value.
Implication. Physiological measures describe the status of various parameters in athletes but they have little predictive association with future performances.
Return to Table of Contents for this issue.