Balancing varied assessment functions to attain systemic validity: Three is the magic number

Accountability-oriented reforms demand immediate and continual gains on achievement test, for all students, and without diminishing other outcomes or undermining instruction. This paper describes a framework for aligning classroom assessment and external testing with the aim of negotiating these seemingly contradictory goals. The framework varies the sensitivity to instruction and the representations of knowledge across approaches to assessment. Cycles of design- based studies successively refine relationships between a curriculum and the frame that each assessment provides. Doing so, we argue, leverages the unique formative and summative balance across assessments in order to scaffold learning and demonstrate the "consequential" validity of our strategy without compromising curricula, instruction, or the "evidential" validity that warrants their continued use.