How to avoid common data distortions with better bar graphs

A new study published as a preprint from HWNI member Markita Landry and graduate student Teng-Jui Lin has attracted the attention of the journal Nature, who published a technology feature article about it. The study is an analysis of nearly 3,400 biological research papers with at least one bar graph. Landry and Lin found that close to one-third of these papers included bar graphs that distorted the data in a way that could lead to misunderstandings, especially by the general public. The preprint provides recommendations on how to make bar graphs that avoid these distortions for more accurate communication of research findings. Read more from Nature.