USING THE MANY-FACET RASCH MODEL TO DETERMINE CUTSCORES AND RESOLVE FUNDAMENTAL STANDARD SETTING ISSUES

Abstract

This paper illustrates the use of the Many-facet Rasch model and the Objective Standard Setting Method in the construction of cutscores, and in dealing with fundamental standard setting issues. With the proliferation of standardized testing for accountability purposes, the tracking of educational growth within and across nations, and the high-stakes use of educational standards, issues pertaining to the selection of the “right” standard setting method for the construction of defensible cutscores has become more and more prominent. The Many-facet Rasch Model used together with the Objective Standard Setting Method allow for the modeling and adjustment of rater severity in the standard setting process, and facilitate clear and efficient identification of inconsistent judges and misjudged items. However, as with most model-based approaches, a clear understanding of construct theory is imperative to arrive at valid and defensible cutscores and performance standards.