Artificial Intelligence in E-Learning: An Empirical Study of ChatGPT as a Learning Tool

Abstract

Artificial Intelligence (AI) is transforming not only traditional learning processes but also e-learning methods. This paper aims to statistically analyze the impact of AI tools on the academic performance of higher education students. An experimental study was conducted with active students of the European University of Tirana.  Students participated in a structured learning session utilizing ChatGPT to engage with specific concepts from the curriculum, followed by an assessment designed to evaluate their comprehension and knowledge acquisition. The sample yielded 26 valid test responses. The data gathered through this process is used to measure the impact of AI tools on learning performance.  The students’ actual average grade was also treated as a controlled variable and measured as part of the study. The study aims to determine if there is a relationship between learning performance through ChatGPT and overall academic performance. The findings yield descriptive statistical measurements, concluding that learning performance is not very high when using an AI tool.  The data were transformed into categorical variables, and a contingency table was used to conduct a test of independence to determine whether students’ academic achievement influences learning performance facilitated by an AI tool. Additionally, a correlation analysis is conducted to examine the potential relationship between variables. The findings indicate that the variables are independent, suggesting that other factors may affect learning performance. The study is of high importance for transforming the traditional learning process in the education field.