This section explores empirical findings related to writing complexity, emphasizing demographic, linguistic, and economic factors, as well as potential influences from Large Language Model (LLM) usage. Analysis using the Flesch-Kincaid Reading Grade reveals significant variations in readability across different groups. Native English-speaking countries score higher, indicating complex sentence structures. Notably, gender disparities emerged, with female-authored abstracts generally having better readability. Geographic disparities show South Asia with the lowest scores, while regions like East Asia & Pacific excelled. Trends from 2012 to 2024 indicate general improvement in writing quality, with a notable dip during the COVID-19 pandemic. The analysis also highlights how internet access and native language impact writing complexity. Additionally, LLM usage is growing, particularly in non-native English-speaking countries, suggesting a shift in academic writing styles and possibly enhanced readability in regions like China, which is rapidly adopting these AI-driven linguistic patterns.
Source link
