Building Assessment Systems That Work

We build assessments from the ground up, and we fix ones that aren’t working. This includes formative assessment processes, classroom assessment tools, summative state tests, kindergarten readiness measures, and credentialing and adult learning instruments. Every instrument we build or improve is designed to produce evidence that students, families, educators, and policymakers can use to inform decisions.

We conduct validity and fairness studies, apply modern psychometric methods, and provide the technical backbone for programs where the results matter and the methodology will be scrutinized. That includes equating, standard setting, alignment, classification accuracy, linking, score reporting, and computational psychometrics.

High-quality assessment data is valuable only when the people using it can turn it into meaningful decisions about programs, policies, and instruction. We help educators and leaders choose the right tools, interpret the results with confidence, and turn assessment evidence into practical next steps they can apply right away to their practice.

States, districts, and assessment providers face challenging decisions with assessment systems: Which screeners should they adopt? Should they adopt new interim assessments? How should they balance accountability requirements with local assessment needs? How can they best integrate AI?

We help education agencies and organizations think through these questions with technical expertise and independent perspective to build systems where the pieces fit together.

Our team develops machine learning models for screener comparability, applies machine learning to study whether assessments measure the same thing for students across overlapping demographic groups, builds AI tools for item and passage development, fine-tunes AI models for assessment alignment, and studies how large language models may introduce bias into measurement.