‘Spell-checker for statistics’ reduces errors in the psychology literature. A study comparing journals that require researchers to use a web application for detecting statistical errors in papers and that do not. The former reduced errors by 4.5%, while the latter by 1%.