CORRESPONDENCE

Dispense with redundant P values

University of Amsterdam, the Netherlands.
Contact

Search for this author in:

In the same issue in which you propose ‘Five ways to fix statistics’ (Nature 551, 557–559; 2017), seven of the nine research papers in the biological sciences include P values with their figures. In my view, this convention is questionable in most cases, given that the large statistical effects are immediately evident from the respective plots.

To help to fix statistics, I urge authors to instead report a P value only when it conveys useful information and is essential for the interpretation of results.

Nature 554, 31 (2018)

doi: 10.1038/d41586-018-01314-9

Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.