Title: An IRT model for evaluating university students' satisfaction about the on-line activities during the COVID-19 pandemic
Authors: Serena Arima - University of Salento (Italy) [presenting]
Cristina Mollica - Sapienza Universita di Roma (Italy)
Abstract: The transition from the second to the third decade of the new century was sanctioned by the serious health crisis caused by COrona VIrus Disease (COVID-19). In order to contain the spiral of infections and thus avoid the collapse of the health structures, governments imposed a series of restrictive measures which, during the 2020 springtime, culminated in a period of lockdown. As a consequence, the traditional academic institutions reformulated their training activities and administrative services remotely, through the massive use of digital communication technologies. Motivated by an international project promoted by the University of Ljubljana in Slovenia, aimed at studying the COVID-19 effects on the life of higher education students, the performance evaluation of the Italian academic education institutions in coping the COVID-19 pandemic is considered, based on the satisfaction experienced during the lockdown by a sample of university students enrolled in a degree or doctorate courses. We propose a Bayesian item response model, a mixed-effects graded response model, accounting for the different nature of the possible answers. The latent trait, e.g. the satisfaction of each student, is hierarchically modelled as a function of specific characteristics of the corresponding university. Universities are then ranked using stochastic dominance criteria.