State accountability systems are designed with the goal of furthering student learning, but the metrics they rely on often differ fundamentally from the data that actually drives school improvement. This article distinguishes between data used for public accountability and data used for school improvement, to help decision-makers deepen their holistic understanding of indicators of student progress.
Effective improvement data must be timely, clear, malleable, and predictive—properties that annual standardized accountability metrics frequently lack. Standardized test scores, for instance, change so slowly that even schools making exceptional gains may appear stagnant, while high-level outcome measures like graduation rates offer little actionable guidance for educators working within a school year. The article examines how malleability and predictive validity should serve as non-negotiable criteria when selecting accountability indicators, and describes the potential risks of bringing school-level improvement data into public accountability frameworks, where the pressure to perform can suppress honest inquiry and incentivize superficial fixes. State boards of education are encouraged to seek evidence of both malleability and predictive power for any proposed indicator, set realistic annual targets grounded in what schools can genuinely achieve, and invest in the data infrastructure and collaborative time that meaningful improvement requires.
When state boards treat accountability metrics as tools rather than labels—asking not just whether a number moved, but what it means and what conditions shape it—they can build educational ecosystems that serve both transparency and genuine school improvement.