Abstrakt: |
This article presents a standard-setting procedure for performance assessment in a foreign language, through which some of the major problems facing performance assessment in criterion-referenced testing can be addressed. The procedure, which was geared to revealing and accommodating inter-judge variability, employed the synergy of multiple approaches to the data analysis, including Generalizability Theory (Brennan, 1983; Shavelson and Webb, 1991) and Multifaceted Rasch Analysis (Linacre, 1994; McNamara, 1996), whereby the investigator was better informed in handling the problems associated with performance assessments in criterion-referenced testing. The study was motivated by the need to develop a viable standard-setting procedure and to demonstrate its potential. [ABSTRACT FROM AUTHOR] |