한국무용 실기고사의 심사자간 신뢰도

Abstract

Traditionally, dance performance was assessed in term of holistic rating method which describes overall score instead of several discrete evaluation domains. This method, however, can be influenced by the judge`s severity and inconsistency unlike the dance performance measures. The trimmed mean has been used in dance performance assessment as a summing method of judges` assessment results. The trimmed mean, however, may not account for the severity and leniency in the assessment of the remaining judges who are not involved in summing process. The purpose of the study was to examine the inter-judges` reliability of Korean dance performance assessment collected from three years. Five experts who worked at the University more than 10 years were selected for rating students` dance performances for three years from 2004 to 2006. Intraclass correlation coefficient (ICC) was calculated to examine the internal consistency of five judges` assessment. The range of ICCs obtained from dance performance assessment for three years was .840 to .851. The range of ICCs when a pertinent judge was deleted was .780 to .917 In conclusions, judges` dance performance ratings based on holistic rating method was moderately high, but not enough for requirement for competition judgment (ICC=.900) and appeared inconsistent assessment among judges. It is suggested that analytic rating method is recommended when assessing dance performance for the purpose of college entrance examination and national level dance competition.