Background: The heptathlon is a women's multi-event competition that determines the best overall athlete in track and field. The interrelations among the seven events reflect their unique contribution to the overall score as well as the contribution of possible latent abilities that may affect performance in more than a single event. Heazlewood (2011) examined the performances in the IAAF top list in 2010 (N=173) using both structural equation modeling (SEM) and traditional factor analysis (FA). SEM did not provide an interpretable description of the hypothesized heptathlon structure while FA resulted in four distinct factors (`sprinting', `jumping', `throwing' and `endurance') that explained 73% of the total variance. Fanshawe (2012) analyzed the results of the 2008 Olympic Games (N=33) and reported only two factors (`throwing' and `running-jumping'). Both studies were based on relatively small samples that may have yielded unstable extracted factors. Aim: The purpose of the current study was to establish the structure and magnitude of latent abilities in the heptathlon based on a larger multi-year sample of heptathlon athletes. Methods: The sample consisted of all the scores from the IAAF top list of the 2004-2013 seasons. Athletes who were listed more than once in the top lists are represented by their best performance during this period. Out of 453 best scores, only 450 athletes were listed with complete records. These records were subjected to PCA with varimax rotation. Results: Two factors were identified with Eigenvalues of 2.2 and 1.5, explaining 52.7% of the total variance. The first factor displayed high factor loadings for the 200m (0.81) and the 100m hurdles (0.78) races and the long jump (-0.72). It may reflect the trait of `running speed'. The second factor displayed high factor loadings for the shot put (0.82) and the javelin throw (0.80) and reflects a general `throwing' ability. The high jump was partially loaded on both factors (-0.40 and 0.42 respectively) while the 800m race was partially loaded on the running speed factor (0.42). The 800m and the high jump are characterized by low communalities (0.17 and 0.34 respectively), indicating a large unique contribution to the total variance while all other communalities (range: 0.58 - 0.68) indicate that more than 50% of the event variance is explained by latent traits. Conclusions: The previously described `throwing' factor was reiterated in the current study. The other factor was clarified as being dominated by `running speed'. No other general traits were unveiled. These results bear a range of implications with regard to training and talent identification.
© Copyright 2014 The 3rd Wingate Congress of Exercise and Sport Sciences. Veröffentlicht von The Wingate Institute for Physical Education and Sport. Alle Rechte vorbehalten.