Closing the Loop: Providing Test Developers with Performance Level Descriptors So Standard Setters Can Do Their Job

Abstract

Standard setting panels are tasked with recommending one or more performance level standards for assessments that are used to classify students into ability categories. These assessments are sometimes developed with the performance level descriptors known and other times without these descriptors. Based on an analysis of 11 state, educational, alternative assessments, this chapter investigates the effects on the standard setting process of developing a test both with and without these descriptors. The results suggest that the standard setting panelists are more consistent with one another and more aligned with empirical data when the items were developed with the descriptors in mind.