Descriptive statistics

Simply looking at summary statistics for the individual judges reveals interesting information about the judges. The low mean scores of the Romanian, Chinese, and Russian judges indicate that they tend to score more harshly than the others; the high mean scores of the South Korean, French, and American judges show that they are more generous; and the Italian judge's mean score shows her to be middle-of-the-road. Moreover, the low standard deviations for the French and Chinese judges' scores mean that they mark deviations in performance quality on a finer scale than the others, while the high standard deviations for the American and Russian show that they make free use of the range of scores. A blunder or exceptional leap that may cause a change of 0.2 or 0.3 points in the French judge's score will cause a change of 0.3 or 0.4 points in the American judge's score.

However, while the judges seem to be quite different in their methods of scoring, the individual summaries don't give the whole story. It is entirely possible that although judges 4 and 5 give quite different scores for the same performances, their pattern of scoring is quite similar. Check the intraclass correlation coefficients to test this.

Next