A vision for instructionally embedded assessments

Across the country, states are rethinking how large-scale assessments serve students and teachers. New federal flexibility and growing interest in through-year models have opened the door to designs that do more than measure learning at the end of the year. Instructionally embedded assessments are central to this shift because they allow educators to check progress as instruction happens.

We believe the future of assessment is shorter tests with deeper insight, data that inform teaching in real time, and results that meet accountability needs. Done well, instructionally embedded assessments can reduce testing fatigue, strengthen instructional alignment, and provide evidence that serves both classroom and policy needs.

Our perspective

Instructionally embedded assessment is not a new idea, but making it work at scale requires more than good intentions. The system must be grounded in research-based models of student learning and flexible in meeting the needs of students, teachers, and parents. Instructionally embedded assessment requires a commitment to usability because if results don’t help teachers teach, the model fails.

At ATLAS, we see an opportunity to move beyond compliance-driven testing toward systems that meaningfully support student learning outcomes. Our work over the past decade has shown that when assessments are designed around models of student learning and integrated into instruction, they can transform the way educators use data. By monitoring learning in real time, educators can identify misconceptions that might otherwise go undetected.

What experience tells us

 Through the Dynamic Learning Maps® (DLM®) Alternate Assessment System, ATLAS has spent nearly 15 years developing and delivering large-scale assessments based on research-informed learning maps. DLM assessments are for students with the most significant cognitive disabilities for whom general state assessments are not appropriate. The maps at the core of DLM assessments make the pathways toward student learning goals explicit, helping teachers understand what students know and the multiple paths toward where they need to go next. Those years of experience taught us that:

• Assessments must fit instruction, not interrupt it.
• Data must be immediate and actionable.
• Assessment systems should be efficient and purposeful.

When it’s clear what an assessment is meant to measure and how its results will be used, states can design systems that reduce redundant testing and help teachers use data to guide instruction.

Our recent work with instructionally embedded models in general education builds on these principles. The Pathways for Instructionally Embedded Assessments (PIE) project has shown that shorter, targeted assessments paired with clear reporting can give teachers the insight they need.

In a pilot study of the PIE system, teachers used assessment results to inform instructional strategies and activities while targeting skills students struggled with. Teachers described the learning pathways as “powerful and helpful” for seeing how math skills connect conceptually, how precursor skills build toward grade-level standards, and how progressions reveal a clear trajectory for their lessons. This is one of many ways educators use real-time data to adjust instruction and support student progress on challenging concepts.

The path ahead

Interest in through-year and instructionally embedded assessment designs is not a passing trend. It reflects a growing consensus that assessment should add value to teaching and learning. The challenge is turning that vision into reality for all students.

 We have already shown what is possible. DLM has delivered a peer-reviewed, instructionally embedded alternate assessment since 2015, and PIE has explored scoring models, data designs, and classroom configurations to guide operational decision-making.

Now the path forward is about scale and impact. States need models that maintain rigor while also meeting multiple purposes, from guiding daily instruction to supporting summative reporting. Our next steps focus on helping states define theories of action, designing systems around intended uses, and building teacher-friendly tools that make assessment results more actionable.

This is a pivotal moment for assessment policy and practice. Our research base, expertise in learning maps and evidence-centered design, and experience with alternate and general education systems position us to help states that value large-scale assessment but want an alternative to traditional end-of-year testing. Traditional summative models have not consistently served all learners well. Instructionally embedded designs offer a more rigorous, effective, and equitable path forward. Now is the time to apply these proven design principles to benefit all learners.

About ATLAS

ATLAS, a center within the Achievement & Assessment Institute at KU, promotes learning and improved outcomes by creating accessible and academically rigorous technology-based learning and assessment systems. Visit the ATLAS website to learn more.