Exploring Student Engagement with ChatGPT in Simulated Learning: A Multivariable Analysis of Pedagogical Preferences

This study investigates students' perceptions across different subjects and demographic profiles regarding the use of ChatGPT in educational activities, with a specific focus on its implementation in simulation-based tasks, compared to UrbanGame and Museum Assignment activities. A range of quantitative analyses—including ANOVA, parametric and non-parametric correlations, multivariate analysis, and chi-square tests—were employed to examine the influence of variables such as age, gender, and academic discipline on pedagogical interest and activity preferences. Results indicate significantly higher pedagogical interest among younger students and moderate correlations between Likert-scale learning dimensions. Despite being the least preferred activity, simulation demonstrates comparable educational value in terms of content comprehension, skill acquisition, and value awareness. No statistically significant differences in activity perception were found across age, gender, or subject, suggesting a broadly inclusive pedagogical design. These findings provide guidance for improving instructional strategies and integrating AI-based tools like ChatGPT into the classroom.