ABSTRACT
It is only a slight exaggeration to describe the test theory that dominates educational measurement today as the application of 20th century statistics to 19th century psychology. Sophisticated estimation procedures, new techniques for missingdata problems, and theoretical advances into latent-variable modeling have appeared—all applied with psychological models that explain problem-solving ability in terms of a single, continuous variable. This caricature suffices for many practical prediction and selection problems because it expresses patterns in data that are pertinent to the decisions that must be made. It falls short for placement and instruction problems based on students' internal representations of systems, problem-solving strategies, or reconfigurations of knowledge as they learn. Such applications demand different caricatures of ability—more realistic ones that can express patterns suggested by recent developments in cognitive and educational psychology. The application of modern statistical methods with modern psychological models constitutes the foundation of a new test theory.
