Logistic regression of online risks on academic stress and performance undergraduates

Abstract

The progression of technology could lead to online risks, such as accidental exposure to harmful content by undergraduates. Investigating how this exposure affects their mental health, particularly academic stress, and in turn, their academic performance, is critically important. This study aims to examine the impact and predictability of online risks on undergraduates’ academic stress and performance, using logistic regression as the main method of analysis. The findings show that online risks have a significant effect on academic stress (p<0.05), but there is no significant impact on academic performance (p>0.05). Students who frequently encounter scam or bullying content are 2.317 and 2.400 times more likely, respectively, to suffer from academic stress compared to those who encounter it less. Additionally, demographic factors, especially gender, are significant (p<0.05) in terms of academic stress and performance. The analysis predicts that females are 4.210 times more likely to experience academic stress than males, while males are 2.768 times (in model 4) and 2.601 times (model 5) more likely to achieve cum laude honors than females. This research provides valuable insights for academic policy makers to improve education quality and offers a basis for further studies in this area.