A Bayes net approach to modeling learning progressions and task performances

Abstract

A major issue in the study of learning progressions is the linking of student performance on assessment tasks to the progressions. This paper describes the development and use of learning progressions (LPs) in the field of computer networking, centering on Bayes nets built around LPs. We discuss how Bayes nets can be built to 1) accommodate responses to tasks keyed to targeted levels of learning progressions and 2) define ordered student-model variables to characterize students’ standing on LPs. The ideas are illustrated with exemplar Bayes nets built on Networking Academy LPs, and tasks designed to obtain evidence in their terms. Extensions to observations from complex tasks keyed to multiple learning progressions are described.