Capti Assess with ReadBasix – Regenerative Reading Assessment

The Phase I project designed, built, and refined a functional prototype of a regenerative reading assessment system that automatically creates, analyzes, and manages new ReadBasix items using a combination of GenAI, NLP, and psychometric validation. Key accomplishments included developing the Item Generator for six foundational reading skills, creating statistical Item Analyzer scripts, outlining the Item Manager dashboard, and enhancing the Capti assessment management platform for greater scalability and usability. Over 100 trial items per subtest were generated, fairness-reviewed, and incorporated into a national pilot with more than 10,000 students. Extensive concept testing with teachers, students, parents, and administrators informed improvements to UI design and reporting features. The project established clear feasibility and a strong foundation for scalable, self-renewing assessments that can expand item pools, support curriculum alignment, and reduce the cost and labor of ongoing test development.

In Phase II, over the two-year project period, the team will refine and fully develop three interconnected components: the Item Generator, which will create new items aligned to district curricula; the Item Analyzer, which will validate item quality, reliability, and fairness; and the Item Manager, which will allow educators to review and manage item banks. In Year 1, the team will focus on development, usability testing, and creating trial item sets. In Year 2, pilot studies will be conducted to test functionality and evaluate the reliability and validity of the new assessment forms, ensuring they are ready for large-scale use by schools and districts.