Open Access

Jakub Jerzy Sommerfeld

• This study investigates several key questions related to health literacy, text difficulty, and AI-generated content. Our research explores whether there is a statistical significance in the perceived difficulty of texts based on their authorship (human-translated to layman language, AI-translated, and original excerpts from a medical book), the correlation between participants' evaluations of text difficulty and quantitative analysis of these texts using JASNOPIS, and the relationship between health literacy levels and the ability to recognise AI-generated work. Additionally, the study examines the impact of exposure to AI-made writing on recognising AI-generated content and the relationship between health literacy levels and perceived text difficulty. A cross-sectional survey was conducted with 94 participants, collecting data through online questionnaires between September 19, 2024, and October 9, 2024. Key findings include significant correlations between average perceived difficulty and several JASNOPIS variables. The study also highlights the lack of a significant relationship between exposure to AI-made writing and the ability to recognise AI-generated work. However, the findings indicate that individuals who scored higher on health literacy tests perceived the texts to be easier. The implications of this research emphasise the importance of enhancing health-related and educational materials using plain language principles to improve accessibility and comprehension. Future studies should consider using more objective measures to avoid the influence of personal biases.