Effects of the COVID-19 pandemic in higher education: A data driven analysis for the knowledge acquisition process

Abstract

The COVID-19 pandemic abruptly changed the classroom context and presented enormous challenges for all actors in the educational process, who had to overcome multiple difficulties and incorporate new strategies and tools to construct new knowledge. In this work we analyze how student performance was affected, for a particular case of higher education in La Plata, Argentina. We developed an analytical model for the knowledge acquisition process, based on a series of surveys and information on academic performance in both contexts: face-to-face (before the onset of the pandemic) and virtual (during confinement) with 173 students during 2019 and 2020. The information collected allowed us to construct an adequate representation of the process that takes into account the main contributions common to all individuals. We analyzed the significance of the model by means of Artificial Neural Networks and a Multiple Linear Regression Method. We found that the virtual context produced a decrease in motivation to learn. Moreover, the emerging network of contacts built from the interaction between peers reveals different structures in both contexts. In all cases, interaction with teachers turned out to be of the utmost importance in the process of acquiring knowledge. Our results indicate that this process was also strongly influenced by the availability of resources of each student. This reflects the reality of a developing country, which experienced prolonged isolation, giving way to a particular learning context in which we were able to identify key factors that could guide the design of strategies in similar scenarios.