Abstract
Purpose of the research. This study explores COVID19-imposed online education at one of the universities in Kazakhstan to understand whether and in what way it has affected students’ academic performance. Methodology. We use several methodologies, such as the principal component analysis, the nearestneighbour propensity score matching and the latent variable multilevel model. Originality / value of the research. To the best of our knowledge, this is the first attempt to estimate the effect of COVID19-imposed online teaching on academic performance in Kazakhstan. Findings. Though our data suggest that the students exposed to online education had slightly lower academic performance than the earlier cohort of similar students who studied in class, this result is not statistically significant. In fact, for both cohorts of students, their past academic performance reflecting their academic abilities and attitudes, such as university entry test score and past accumulative GPA, matter much more than the mode of study. However, we found out that certain technical variables which characterise the “quality” of Zoom® sessions affect students’ academic performance. Specifically, using a smartphone instead of a personal computer to enter a class conducted in Zoom® negatively affects students’ achievement. Some other characteristics also produce expected effects though they turned out to be statistically insignificant possibly due to aggregation and a small dataset.