The Impact of Covid-19 on Online Final Exam Scores Among Computer Science Students Malgorzata Rataj and Joanna Wojcik University of Information Technology and Management, Rzeszow, Poland mrataj@wsiz.edu.pl jwojcik@wsiz.edu.pl Abstract: The main goal of this research is to identify the impact of COVID-19 on online final exam scores among Computer Science students. The correlation matrix we used indicates the interrelationships among learning outcomes and student profile, type of classes and student online behaviour. Six courses were taken under consideration: Practical Algorithms, Discrete Mathematics, Software Engineering, Programming, Team Projects and Artificial Intelligence. A total of 4,988 final exam results were examined. After a deep analysis of the literature on the topic, we expected two scenarios. The first scenario constituted a decline in passing grades due to challenges such as: learning platform failures, poor internet connections or poorer quality of lessons due to teachers’ lack of online competence. We hypothesized the second scenario as extraordinary student performance compared to their prior exams, but due to their dishonesty. The results of the study revealed that neither of the scenarios took place. It turned out that the challenges that seemed to be the most difficult ultimately did not matter. The present study finds that there is not a significant difference in the students’ final exam performance between their online and traditional courses. Our strategy as described in this article has demonstrated a smooth transition from traditional to online teaching and assessment in terms of the final assessment. Keywords: online exam, assessment, delivering education, e-learning 1. Introduction Nowadays society has become highly dependent on computer networks, particularly after the Covid-19 crisis demanded it on a global scale. The development of information technologies not only opens up great opportunities for new ways of delivering education but also provides us with all the tools we need for carrying out higher education fully online. It is clear that successful implementation of university courses is no coincidence, but rather requires great analysis and planning (Rataj & Wojcik, 2020). Every professional learning initiative requires deep analysis to identify the learning needs so as to best shape the proposed solution. This article describes the seventeen-month period of university closure from March 2020 to August 2021, comparing it to the same semester conducted in the traditional form a year earlier. It is worth noting that in both years the subjects were conducted by the same teachers. Among the many challenges in the transition from face-to-face education to online education, the assessment – namely the switch from oral/written exams to online exams – can be considered as one of the biggest hurdles. The main goal of this research was to identify the impact of the COVID-19 on final exams scores of Computer Science students. Traditionally, these courses were delivered in the classroom: all classes took place face-to-face with the teacher and all exams were fully supervised in person. Despite the circumstances in 2020 calling for major changes, we were not allowed by the Ministry of Science and Higher Education to change the curriculum or learning outcomes (final grades); we were only allowed to change the form of the exam from open question to multiple-choice question or from solving a task to answering true/false questions. These questions were randomly selected and assigned by the system to each student. The fact that each student received randomly selected questions minimalized the possibility of the exchange of answers among students. We made sure that the difficulty level of the exam was the same for every student, and all students took the exam simultaneously. Students did not have the option to return to previously attempted questions, which also served to minimize the possibility of “collusion” between students. Additionally, students had to have their pc cameras turned on during the exam so as to be supervised. In the case of large groups, they were divided into smaller groups, each of which was supervised by additional teachers. All exams went ahead as planned from home via the Blackboard system. 955 Proceedings of the 23rd European Conference on Knowledge Management, ECKM 2022