STRUCTURAL KNOWLEDGE ASSESSMENT - COMPARISON OF REFERENT STRUCTURES

Network representations of student knowledge are usually evaluated by comparing them to an expert ''referent'' structure. This study compared referent structures produced by the instructor, other experts, averaged experts, and an average based on the best students in the class. The referents were compared in their ability to predict exam performance in 2 college level computer programming courses and to differentiate among levels of expertise. Results showed that in terms of these criteria, (a) instructor-based referents were no better than other experts: (b) there was substantial variability among experts; and (c) structures derived from both averaged experts and averaged best students provided valid referents, but the expert-based referent was superior.