The impact of the COVID-19 pandemic on higher education: Assessment of student performance in computer science

Abstract

The COVID-19 pandemic had radically changed higher education. The sudden transition to online teaching and learning exposed, however, some benefits by enhancing educational flexibility and digitization. The long-term effects of these changes are currently unknown, but a key question concerns their effect on student learning outcomes. This study aims to analyze the impact of the emergence of new models and teaching approaches on the academic performance of Computer Science students in the years 2019–2023. The COVID-19 pandemic created a natural experiment for comparisons in performance during in-person versus synchronous online and hybrid learning mode. We tracked changes in student achievements across the first two years of their engineering studies, using both basic (descriptive statistics, t-Student tests, Mann-Whitney test) and advanced statistical methods (Analysis of variance). The inquiry was conducted on 787 students of the Lublin University of Technology (Poland). Our findings indicated that first semester student scores were significantly higher when taught through online (13.77±2.77) and hybrid (13.7±2.86) approaches than through traditional in-person means as practiced before the pandemic (11.37±3.9, p-value < 0.05). Conversely, third semester student scores were significantly lower when taught through online (12.01±3.14) and hybrid (12.04±3.19) approaches than through traditional in-person means, after the pandemic (13.23±3.01, p-value < 0.05). However, the difference did not exceed 10% of a total score of 20 points. With regard to the statistical data, most of the questions were assessed as being difficult or appropriate, with adequate discrimination index, regardless of the learning mode. Based on the results, we conclude that we did not find clear evidence that pandemic disruption and online learning caused knowledge deficiencies. This critical situation increased students’ academic motivation. Moreover, we conclude that we have developed an effective digital platform for teaching and learning, as well as for a secure and fair student learning outcomes assessment.