Departures from the formal of actual students' university careers: an application of non-homogeneous fuzzy Markov chains
F Crippa, M Mazzoleni, Mariangela Zenga · Journal of Applied Statistics · 2015
As in most higher education (HE) systems, the Italian university organisation draws paths of credit progression in formal curricula, which aim at framing the acquisition of knowledge and competencies within each specific major. The resulting yearly syllabi therefore develop in a sequence of examinations that are to be successfully passed, and formal administrative registration allows access to the following academic year. In general, there is a divergence between formal and actual career progression because each university student can proceed at her/his own pace, sketching her/his own trajectories, free to depart from the formal progression. Even if applied to various HE settings, Markov chain models do not fit the aforementioned situation. A methodological extension has been introduced, whereby progression levels are considered as fuzzy states. Markov chains with fuzzy states identify the latter with specified academic years and express each student's situation as a relational link to present and past academic attainments. This link is operationalised by means of a membership function, which is here discussed with reference to the Italian HE system.