engleski

Cluster analysis of sport interests in elementary school children

With the aim to provide the cluster analysis of sport interests, the sample of 1294 elementary school pupils were given list of 52 sports to be evaluated on five-point scale. The data were analysed by k-means clustering algorithm with k set to five. The number of clusters was chosen to enable the presentation of five most often recognized factors of sport interests. Discriminant analysis of clusters confirms that cluster analysis brought five significantly different groups. Univariate analyses of variance show that all 52 sports were included in cluster discrimination. The first cluster includes subjects preferring sports with marked aesthetic component, some sports on water and equestrian, and rejecting the number of "male" sports. The second cluster includes pupils preferring dance, tennis, roller-skating and badminton, and rejecting "male" sports including. The third cluster includes subjects with above average values of preferences on all sports. The fourth cluster includes pupils with marked above average interests in football, ice hockey, rowing, shooting and moto-sport, and evident under average interests in dance, aerobics, synchronized swimming and rhythmic gymnastics. The fifth cluster is composed of subjects prefering football and basketball, being neutral to tennis, table tennis, handball, cycling and moto-sport, and rejecting all the other sports. The description of clusters doesn't reproduce the factors of sport interests and results suggest gender stereotyping being involved.

cluster; sport interests; pupils

nije evidentirano