Optimal CAT

Optimal CAT is an easy-to-use production-grade Computer Adaptive Testing (CAT) engine. It's both for psychometricians to run simulations and for IT engineers to deploy to production.

It builds upon the shadow-test approach. It leverages a state-of-the-art Mathematical Programming solver (FICO Xpress Optimization) to deliver optimal tests to each student (hence the name Optimal CAT). It guarantees that test blueprints and constraints are always met, and measurement accuracy is maximized.

The engine is built as a microservice (stateless REST API interface) and can be easily integrated with many test delivery systems.

shadow-test approach

We believe every student deserves the most accurate measurement.

Latest Version (updated Nov 9, 2024)

Windows             OptimalCAT-1.1.5-windows.zip   (Manual)
Linux                   OptimalCAT-1.1.5-linux.tar.xz
                           
Core Team:         Jie Li, PhD
                            Wim J. van der Linden, PhD
                            Richard J. Patz, PhD

Support:              optimal-cat@outlook.com
              

CSEM of shadow test vs. theoretical limit