Physics Is Hard but Can Be Fun: A Survey of Indonesian High School Students’ Learning Difficulties, Technology Perceptions, and AI Readiness in Physics

Abstract

Despite growing interest in integrating digital technology and artificial intelligence (AI) into science education, limited empirical data exist on how Indonesian high school students perceive these tools for physics learning. This study investigates students’ perceived learning difficulties in physics, their perceptions of digital technology and simulations as learning aids, and their readiness to adopt AI tools such as ChatGPT and Gemini. A cross-sectional survey combining Likert-scale items and open-ended questions was administered to senior high school students from diverse schools across Garut Regency, West Java, Indonesia. The instrument covered four constructs: learning attitudes and difficulties, digital technology perceptions, AI readiness, and learning approach preferences. Results reveal that most students hold positive attitudes toward physics, yet moderate difficulties persist in comprehending abstract concepts, connecting formulas to physical phenomena, and interpreting graphical representations. Digital tools, particularly simulations such as PhET, were regarded as effective for visualizing abstract content. A paradox emerged regarding AI: while most students had used generative AI and valued it as a learning aid, many simultaneously worried that AI dependency could diminish independent thinking. Frequent AI users perceived greater benefits and less anxiety than infrequent users. The strongest consensus was students’ desire to learn responsible AI use, a call that received virtually no opposition. Students preferred real-life examples and hands-on experiments over other learning approaches. These findings point toward a scaffolded pedagogical framework integrating digital simulations and AI tools within context-rich physics instruction.