Hybrid unfolding models to Likert-scale data to assess distance learning perception in higher education

The Covid-19 pandemic forced students of any age and level of education to change the learning process, from learning in presence, to Distance Learning (DL). Such a relevant switch to DL has not been seamless for students, from both practical and psychological perspectives. In fact, students adaptation to DL process also depends on Covid-19 induced stress. Aim of the paper is to analyse an Italian university students survey on DL perception and Covid-19 related psychological effects, such as stress. The proposed approach implements a hybrid method that synthesizes the DL perception items into an ordinal response that is then regressed on the remaining items, to study the the effects on DL perception of further aspects (such as stress) and identify the most relevant covariates. The modeling phase consists of the implementation of the adjacent categories models to take into account the intensity of the opinions (students’ feeling) at each end of the spectrum.