RELIABILITY AND VALIDITY OF JUDGING IN MEN’S ARTISTIC GYMNASTICS AT THE 2009 UNIVERSITY GAMES

Abstract

Ensuring reliability and validity of judging at artistic gymnastics competitions is difficult. Despite the FIG Men’s Code of Points being changed, there is little evidence to show that these changes have had an effect on judging standards. After the last change to the Code of Points (2008) the second biggest men’s artistic gymnastics competition took place in 2009 - University Games in Belgrade. Data based on judges’ scores were analysed. By last change of the Code of Points the sum of the Difficulty score and the Execution score form the Final score. For the Execution score, which is evaluated by 4 or 6 judges (4-in qualifications and all around, 6 in finals) reliability and validity were calculated (intraclass correlation coefficient, Cronbach’s alpha, Kendall coefficient of concordance W, and a theta coefficient; differences in mean E scores between judges were tested using repeated measures ANOVA. All data was analyzed using SPSS Statistics 17.0. Results show very high reliability (e.g. Cronbach alfa range from 0.92 up to 0.99). Systematic bias in individual judge’s scores and judges’ panels were frequent. Invalidity tends to decrease as competitor numbers increase. Despite good reliability and satisfactory validity of judging at the University Games it should be emphasized that judging quality differs between apparatus, sessions and judges.