Evaluation is an integral part of research that provides a true measure of effectiveness. This paper presents a study conducted to evaluate the effectiveness of CAS, a knowledge acquisition authoring system developed for generating the domain knowledge required for constraint-based tutoring systems with the assistance of a domain expert. The study involved a group of novice ITS authors composing domain models for adding two fractions. The results of the study showed that CAS was capable of generating highly accurate knowledge bases with considerably less effort than the effort required in manual composition.