Correlation Is Increasingly Being Written Like Causation, Study of 194,631 Papers Finds

A large analysis of 194,631 research papers suggests that scientists are increasingly using language that presents correlation as if it proves causation. The distinction matters because two factors can move together without one directly producing the other. Researchers say imprecise wording may overstate findings, influence public interpretation, and complicate efforts to reproduce results. The study highlights the need for clearer reporting, stronger statistical reasoning, and careful explanation of limitations when presenting associations in academic research and public communications for everyone.