Abstract
Imprecise wording in research articles can mislead readers to think that the findings are more powerful than they are. We empirically investigate how common it is for scholars to include statements in their articles which fail to accurately reflect the research design they used. We find that inaccurate statements are made in nearly half (47%) of all 2022 articles under investigation. Concerningly, the sections where such statements occur most frequently are the title, the discussion section, and the abstract – the very sections that are read most, especially by people who are only skimming through the literature. As a result, the risk of misinterpreting and misquoting findings is high, as is the risk that industry may adopt measures developed by scholars to change behaviour, even if there is no empirical evidence that such interventions indeed do change behaviour.