Portal de Periódicos da CAPES

An analysis of decathlon data

2002; Wiley; Volume: 51; Issue: 2 Linguagem: Inglês

10.1111/1467-9884.00310

2517-6153

Trevor F. Cox,

Data Mining Algorithms and Applications

Summary. Data in the form of times, distances and scores for athletes competing in the decathlon at the meetings of the International Amateur Athletic Federation from 1991 to 1999 were collected and then subjected to statistical analysis. The first part of the analysis concentrated on the multivariate description of the data by using cluster analysis and classical scaling. The second part of the analysis questioned whether the decathlon could be considered fair or unfair to particular athletes, e.g. the track specialist or the field specialist, since the spirit of the decathlon is that all 10 events are equally regarded. Statistical techniques used were based on Bartlett's test of homogeneity of variance, Kendall's coefficient of concordance, Spearman's rank correlation and cumulative sum charts. The tentative conclusion is that the decathlon favours those athletes who do well at the field events.