심동적 영역에서 준거지향기준의 설정

Abstract

One of current issues in evaluation is with regard to the distinctions between, and relative advantages and disadvantages of norm-referenced and criterion-referenced measurement. However, criterion-referenced measurement is more emphasized with changing the educational circumstances in which a predeterminded standard of performance is used rather than a normative standard. Criterion referenced standard(CRS) represents desired levels of performance or status on a criterion domain or attribute, and provides diagnostic information about whether status or performance is adequate. The use of CRS in testing is to categorize students into master or nonmaster based on the CRS or the cut-off score. The problems of CRS are that they are arbitrary and that the consequences of misclassifications. A variety of methods for establishing CRS have been developed. Mehtods that are applicable in the psychomotor domain of physical education are judgemental, normative, empirical, and combination methods. The judgemental mehtod is to establish CRS by the experts' opinion. The normative method is to use normative data along with the information to setting the CRS. A CRS with empirical method is to base on empirical data of predeterminded master and nonmaster. The combination method is to use the combination of the above methods. An important aspect of developing CRSs is establishing their reliability and validity. The reliability and the validity of CRS are defined as the consistency and accuracy of classification, master or nonmaster. One type of reliability of CRS is rater reliability that is the same concept of objectivity. And P and Kappa coefficients are used as the indices of the CRS reliability in test-retest apporach. The validity of CRS is evaluated based on a probability of correct decision, Phi coefficient, and a utility.