Background/Context: The target of assessment validation is not an assessment but the use of an assessment for a purpose. Although the validation literature often provides examples of assessment purposes, comprehensive reviews of these purposes are rare. Additionally, assessment purposes posed for validation are generally described as discrete and static.
Purpose/Objective/Research Question/Focus of Study: This paper has two goals. The first is to synthesize recent frameworks that differentiate among purposes of large-scale standardized assessments. The second is to describe the forces that shape the purposes of any particular assessment over time. The author highlights the tendency of assessment purposes toward purpose drift, where purposes evolve and differentiate from purposes originally intended. The author argues that this process is predictable enough that validation practices that ignore drift are incomplete and insufficient.
Research Design: This paper is comprised of a synthesis of past validation frameworks and a case-based argument in favor of expanding these frameworks.
Conclusions/Recommendations: If predictable forces cause the purposes of an assessment program to expand beyond the purposes originally validated, then conventional validation approaches proposed in the assessment literature are incomplete. Validation efforts should follow known vectors that propagate purpose drift and address the appropriateness of the purposes at the destinations of these vectors. Vectors include the aggregation of scores to support group-level inferences, the disaggregation of student results into scores for diagnostic or formative purposes, and rising stakes on metrics like “adequate yearly progress” and “teacher value added.”