Reducing Construct Irrelevant Variance in Large-Scale Science Items Using Principles of Universal Design for Learning and Evidence-Centered Design

• In Event: Understanding the Cognitive Processes Tapped by Complex Assessments

Abstract

The No Child Left Behind Act requires that students with disabilities be included in state assessments and accountability, giving rise to a number of issues related to fairness and test validity. This paper describes: 1) a design methodology for improving the validity of inferences about the performance of students with disabilities on large-scale science assessments; (2) results of an analysis of extant science items to identify sources of construct-irrelevant variance related to cognitive demands; (3) examples of item revisions to mitigate construct-irrelevant task demands; and 4) results of a field test of revised items administered to both general education students and those with learning disabilities.

Authors

• Britte Haugan Cheng, SRI International

• Elizabeth Murray, CAST - Sr Research Scientist

• Geneva D. Haertel, SRI International

• Liliana Ructtinger, SRI International

• David Rose, CAST, Inc.