Generalizability of assessing fundamental aquatic skills of preschool children: A three-facet design
Abstract
This study is focused on reliability of assessing fundamental aquatic skills in preschool children. They are still not able to swim at least 10 meters independently, and therefore, they must be tested by several fundamental aquatic motor tasks.
The degree of task accomplishment is evaluated by a group of experienced independent raters. It is insufficient to present only inter-rater reliability as information about reliability of test scores because different aquatic tasks are also one of the sources of measurement error.
In the present paper, the generalizability theory (GT) is introduced as an appropriate tool for solving such a situation. It extended the classical reliability theory by recognizing and estimating measurement error arisen from more than one source.
The whole measurement procedure was repeated after two weeks.