The number of people with low vision globally has been steadily increasing, reaching 1.1 billion as of 2020. Due to their visual impairment, people with low vision face higher risks during disasters such as earthquakes. While research on disaster responses among people with low vision has primarily focused on specific aspects, such as escape strategies or perceptual challenges, few studies have examined the complete evacuation process and decision-making logic during disasters. To address this gap, this study proposes a structural coding method based on interview texts, integrated with user experience (UX) analysis and quantitative evaluation techniques. By utilizing UX to analyze the actions and decisions of people with low vision, this approach enables researchers to efficiently construct a comprehensive framework for understanding their disaster responses and draw valuable insights to improve disaster management strategies for this vulnerable group.