Nature published the 2005 study comparing Wikipedia and Britannica science articles.
The study appeared as a news investigation in Nature in December 2005. Researchers selected pairs of science articles from Wikipedia and the online Encyclopædia Britannica, then sent them to expert reviewers who were not told which encyclopedia supplied which text. The comparison examined factual errors, omissions, and misleading statements.
Nature reported 42 usable reviews. Reviewers found roughly four inaccuracies per Wikipedia article and three per Britannica article, while identifying four serious errors in each source. The result was widely summarized as evidence that Wikipedia came surprisingly close to Britannica in science accuracy, despite its open-editing model.
The study remains important but was not the last word on encyclopedia quality. Britannica disputed the methods and interpretation, arguing that the sample was small and that some excerpts were not directly comparable. Nature rejected the demand for a retraction and defended its investigation. The episode helped make Wikipedia’s reliability a major subject of public and academic debate.