The women's junior and senior Heptathlon in the sport of Athletics is an event that is conducted over two consecutive days using seven event performances in the following order; day 1 consists of the 100m hurdles, high jump, shot put and 200m; and day 2 consists of the long jump, javelin throw and 800m. The individual event performances are converted to point scores using IAAF scoring tables, which are then summed to assess rank performance. Training recommendations are based on conceptual models and suggested exercise physiological links between the different events within the Heptathlon. In this model, sprint-power events such as the 100m hurdles, 200m, long jump and high jump should be correlated and some training transfer should occur between these events. That is, sprint training should assist long jump performance. The predominam research issue was do statistical analyses support the conceptual model, where athletes can expect some training and performance transfer between events or do the seven events represent unique constructs that have to be trained individually? The statistical method known as confirmatory factor analysis was applied to a set of IAAF data based on world athletic rankings for top 100 heptathlon scores in 2006 to uncover the latent structure or factors of the seven events. If significant correlations exist among the seven individual events, factor analysis can be utilised to reduce the larger variable set into a smaller number of latent non-measured variables or factors. This can suggest how the individual events are associated, such as do 200m, 100m hurdles, high jump and long jump represent a dependence on ATP-CP energy production and should cluster or load as one factor, whereas the 800m, which is though to represent an aerobic or endurance event may represent a second performance factor. The initial correlation matrix, a conceptual and statistical starting point in factor analysis, indicated very significant absolute correlations (.351 to .500, p<.001) between the 100m hurdles, 200m, high jump and long jump. Surprisingly, the shot put was significantly correlated (absolute values) with high jump, long jump, javelin and 800m (.243 to .390, p<0.15), although the common variance was small. The initial factor solution utilising principal component analysis derived three factors which explained 72.65% of the total variance. The initia] solution provided the most interprelable factor solution, which loaded 110m hurdles (.675). 200m (.750), high jump (-.635) and long jump (-.806) on factor 1; the shot put (.682) and javelin (.831} on factor 2; and the 800m (.739) on factor 3. A number of orthogonal factor rotations were applied to improve the factor loadings; however this approach did not enhance the interpretations from the initial solution. The rotated factor loadings indicated ihe 100m hurdles (.847), high jump (-.643), 200m (.668) and long jump (-760) loaded on the first derived factor (31.76% explained variance). Surprisingly, the shot (-.602) and 800m (.926) loaded on factor 2 (20.58% explained variance), and finally the shot (.544), 200m (.466) and javelin (.899) loaded with factor 3 (20.32% explained variance). The negative and positive factor loadings indicate inverse relationships. The factor loadings of the shot on factor 1 and factor 3, the 200m on factor 1 and factor 3 indicates factor complexity for these events. These findings suggest that training can be clustered into similar conceptual events, such as speed-power, strength-power and endurance events to enhance transfer based on different energy systems and strength-power-speed-endurance requirements that underpin each event in the Heptathlon.
© Copyright 2008 Proceedings of First Joint International Pre-Olympic Conference of Sports Science and Sports Engineering. Volume I: Computer Science in Sports.. Veröffentlicht von World Academic Union. Alle Rechte vorbehalten.