A1199
Title: Quantile regression for 2022 New York City marathon data
Authors: Anneleen Verhasselt - Hasselt University (Belgium) [presenting]
Abstract: Statistical methodology for complex data has evolved significantly to accommodate the diverse data types encountered in real-world applications. For longitudinal data, a large proportion of adapted methodology has traditionally focused on mean regression. More recently, attention has turned to longitudinal quantile regression, and even then, primarily in a theoretical context. Copula-based longitudinal quantile regression methodology is applied to data from the 2022 New York City Marathon, which includes intermediate time recordings. Despite the ubiquity of such longitudinal race data in increasingly popular running events, they have typically been reduced to one-dimensional data or used solely for ANOVA analyses. Quantile-associated curves are introduced and applied to several metrics quantifying runners' speed and pacing behavior. The benefits of this approach are twofold: it demonstrates the potential of this methodology in similar contexts, and it provides novel insights into runners' racing behavior that can assist athletes, coaches, and sports science experts.
