Latent growth curve modeling is a structural equation modeling technique for longitudinal dataset. lt is a confirmatory data analysis procedure to learn about the variations of an endogenous variable over time. The technique is characterised by estimating intra- and inter-individual growth trajectories enabling researchers to predict future development. This modelling procedure has been applied in other scientific fields, such as Social Sciences but never attempted before in Sport Sciences or sports performance as much as we are aware of. Indeed, all structural equation models reported in the sports performance literature were developed from cross-sectional data (e.g. Morais et al. 2012; Barbosa et al. 2010). Swimming performance is a multi-factorial phenomenon, where the interactions between several exogenous variables determine the endogenous one. Young swimmers' performance is strongly related to the kinematics (e.g. Uitt et al. 2009; Morais et al. 2012) and hydrodynamics (e.g. Toussaint et al. 1990; Marinho et al. 2010; Saavedra et al. 2010). A few follow-up and tracking studies suggested that at different moments of a season, performance would be mostly dependent from different determinant factors (e.g. Mora is et al. 2013). Nevertheless, until now such relationship was never modeled. The aim of this research was to compute a latent growth curve model for young swimmers' performance and its biomechanical determinant factors. lt was hypothesised that different exogenous variables would have the main direct effect on the performance improvement (endogenous variable) over time.
© Copyright 2014 XIIth International Symposium for Biomechanics and Medicine in Swimming. Veröffentlicht von Australian Institute of Sport. Alle Rechte vorbehalten.