Generalizability of ratings of men´s all-around in artistic gymnastics with reference to various research purposes
Abstract
The main aim of this study is to introduce the generalizability theory (GT) as a flexible tool for three investigators with different approaches to the same data set taken from the men’s individual all-around final in artistic gymnastics world championship in Melbourne in 2005. GT extends the concept of reliability within the frame of classical test theory by recognizing and estimating the magnitude of the multiple sources of measurement error.
First investigator is interested in the relationships among all six apparatuses based on the final results. The main purpose of the second approach is to examine the reliability of rating scores of the group of six B-jury within each apparatus.
Finally, the last investigator is focused on the origin of gymnastic apparatuses. All three investigators use univariate and/or multivariate GT methods for answering their research questions.