ASSESSING THE EFFECT OF GENERATIVE ARTIFICIAL INTELLIGENCE ON ACADEMIC STRESS AND ACADEMIC PERFORMANCE AMONG MEDICAL STUDENTS OF KHYBER MEDICAL UNIVERSITY

Abstract

Background: Generative artificial intelligence (AI) is a new reality in the field of medical education that has added new factors to the academic stress and performance of students. This study examines these relationships among students at Khyber Medical University.

Methodology: The study was conducted as a cross-sectional study with 300 Medical and Allied health students and three validated scales. Data analysis consisted of reliability testing (Cronbach's alpha), normality testing, Pearson correlation test, independent-samples t-test, one-way ANOVA test, and simple and multiple linear regression.

Results: The internal consistency of all three scales was very high (α > 0.97). Parametric analyses were appropriate, as verified by normality testing. Simple regression found that academic performance was slightly positively correlated with AI use (R = 0.114, p = 0.048), and academic stress was slightly positively correlated with the use of AI (R = 0.165, p = 0.004). Multiple regression revealed that academic stress was a stronger predictor of performance (B = 0.396; p < 0.001) than using AI was when academic stress was included (B = 0.005; p = 0.911). The mean scores for academic stress and academic performance across academic years were found to be different by one-way ANOVA (p = 0.028 and p = 0.048), respectively. There were no significant differences between males and females on any of the variables.

Conclusion: Generative AI is an adjunctive learning tool that has little direct impact on learning outcomes. The factors for academic performance are mainly academic stress, and structured policies for AI use and comprehensive mental health support for students are crucial.