Validity by Design

Abstract

Lissitz and Samuelsen (2007) argue that the unitary conception of validity for educational assessments is too broad to guide applied work. They call for attention to considerations and procedures that focus on “test development and analysis of the test itself” and propose that those activities be collectively termed content validity. The author of this article describes work that makes more explicit the underlying principles of assessment design, thereby providing conceptual foundations for familiar practices and supporting the development of new ones. By structuring design activities around assessment arguments, the test developer accrues evidence in passing for what Embretson (1983) calls “construct representation” argumentation for validity.