Evaluation of final examination performance at Czech University of Life Sciences during the COVID-19 outbreak

Abstract

The COVID-19 pandemic outbreak has upended the educational system worldwide, possibly with severe long-term consequences as most training institutions were forced to move to an online environment. Given the sudden transition to remote education, the main objective of this contribution is to evaluate the impact of distance education on examination results. We investigated the examination results of tax related subjects collected at the Czech University of Life Sciences in Prague during the period from 2014 to 2020. The sample consists of examination results of 120 different classes within 6 years with a total amount of 7268 observations. Firstly, we pivoted the data into the long format and performed binary logistic regression. Our findings suggest that the odds that student successfully passes the exam increases if the student was examined online compared to in-person. Additionally, we used KNN regression which enables us to predict success rate for an upcoming semester. According to our analysis, it is expected that on average 82 students out of 100 will successfully pass the exam. The model was calibrated using cross-validation to choose optimal K.