The Emotional Side of AI: Assessing ChatGPT’s Hedonic Usability and User Perceptions

Abstract

The primary focus of this study is to investigate the emotional and hedonic usability of ChatGPT in educational and research environments, analyzing user experiences among university students. A mixed-method research approach was used, incorporating statistical analysis, surveys, sentimental classification, correlation analysis, regression modeling, and institutional comparisons. The key usability factors examined include emotional satisfaction, frustration, enjoyability, academic usefulness, ease of use, accuracy, response speed, and engagement. The major findings indicate that ease of use, accuracy, and response speed significantly impact academic usefulness, while technical issues and contextual inaccuracies cause frustration. Engagement significantly predicts enjoyment, frustration has a negligible impact, and institutional differences affect emotional satisfaction levels. The study also found that user engagement varies significantly based on experience levels with AI technology. Users who regularly interact with ChatGPT demonstrate higher confidence in the tool’s capabilities and greater reliance on its suggestions, whereas users with limited exposure often express skepticism regarding its accuracy. Recommendations include developing personalized interfaces, improving contextual awareness, integrating ChatGPT with educational platforms, and building adaptive learning pathways that provide tailored guidance based on users’ experience levels.