State assessments can be useful for both instruction and accountability—if we design them that way

Dale Chu’s recent critique of through-year assessment revisits a familiar concern: that state assessments cannot serve both accountability and instruction without compromising one purpose or the other. It is a concern worth taking seriously. Accountability requires comparable, defensible claims about student performance; instruction requires timely, actionable evidence that helps teachers respond to student learning. If poorly designed, through-year assessment can blur those purposes and satisfy neither.

But that is not an argument against through-year assessment. It is an argument for better design.

Nathan Dadey and his colleagues at the Center for Assessment have made the more careful version of this point. They define through-year assessments as systems intended to support both “the production and use of a summative determination” and at least one additional goal. The most commonly cited additional goal is instructional utility: providing information during the year that helps educators and students make better decisions while there is still time to act.

That is exactly the claim through-year assessment systems must be held accountable for.

So what is required for instructional utility? The Center has identified five essential features: coherence with the enacted curriculum, tasks that support deeper thinking, results at the right grain size, timely feedback, and results that inform instruction. The fifth is both the most important and the hardest to achieve. It is not enough to display data faster or more often. Assessment becomes instructionally useful only when it helps educators understand what students are thinking, where they are struggling, and what to do next.

That is the design problem we are working to solve with the Montana Office of Public Instruction.

In the Montana MAST through-year model, shorter assessment events are designed to align flexibly to local curriculum scope and sequence while still aggregating to a comparable statewide summative measure. The goal is not to make a single test “do everything.” It is trying to replace a fragmented assessment ecosystem—state summative tests over here, commercial interims over there, curriculum-embedded checks somewhere else, and disconnected remediation tools on top of all of it—with a more coherent system of instructional decision making.

This is not a “New Shimmer” problem—asking one product to do two incompatible things. Granted, the accountability concern is real. Because through-year testlets contribute to the final summative score, accountability pressure does not disappear; it is distributed across the year. That makes design especially important. These short testlets, administered in a class period every four to six weeks, are not intended to replace classroom formative assessment practices or curriculum-embedded assessments. They are intended to monitor students’ progress toward mastery in a way that applies accountability not simply to the final outcome, but in support of better instructional practice throughout the year: reinforcing the adopted curriculum rather than narrowing it, testing students on what they have had an opportunity to learn, returning instructionally useful feedback while it can still shape instruction, and supporting cycles of inquiry that help teachers adjust instruction before learning gaps harden.

We are seeing evidence that students and educators recognize the value of this approach as they become more familiar with it. In the most recent annual survey of Montana teachers and students, 79 percent of students and 59 percent of teachers surveyed agreed that students learn more when they take several smaller assessments during the year instead of one large test at the end. And teachers overwhelmingly value the misconceptions reporting because it gives them something specific and actionable they can address in the context of their taught curriculum. That’s something they just aren’t getting from the rest of the test pile.

Dale’s argument largely recycles an older debate as if the field has stood still. Innovative assessment leaders are moving beyond the false choice between instructional utility and comparability. A growing number of states—including Montana, Indiana, North Carolina, Missouri, Delaware, and now Texas and Oklahoma—are exploring ways to make state assessment more useful without abandoning comparability or rigor. The details differ, but the motivation is shared: State leaders recognize that the current model often produces information too late, too far from instruction, and too disconnected from the decisions educators need to make during the year.

That is why through-year assessment should be evaluated not as a new version of the traditional summative test, but as an effort to redesign the state assessment system around a clearer theory of action. The recent analysis of MAST published by the Montana Office of Public Instruction and Education First makes that point well:

Through-year assessments are best understood not as incremental modifications to existing summative models but as fundamentally different systems that require clarity of instructional vision, deeply contextualized theories of action, intentional incentive structures and a reimagined role for state education agencies. When these elements are misaligned, TYAs risk amplifying reactive or incoherent instructional responses. When they are thoughtfully aligned, however, TYAs have the potential to provide information that is genuinely value-added: supporting educators in making more informed, timely and productive decisions for their students.[1]

State assessments should not promise to be everything to everyone. But they can be better than the status quo. Through-year models should be judged by what they make possible in practice: teachers using more actionable evidence in planning conversations, districts using data to target professional learning and curriculum implementation supports, and states gaining clearer visibility into where investments are needed to help students meet rigorous expectations. The challenge is real, but it is one that state assessment systems are increasingly positioned to meet.