Search Menu

P values in display items are ubiquitous and almost invariably significant: A survey of top science journals

P values represent a widely used, but pervasively misunderstood and fiercely contested method of scientific inference. Display items, such as figures and tables, often containing the main results, are an important source of P values. In this article, published in PlosONE, Ioana A. Cristea and John P.A. Ioannidis conducted a survey comparing the overall use of P values and the occurrence of significant P values in display items of a sample of articles in the three top multidisciplinary journals (Nature, Science, PNAS) in 2017 and 1997, respectively. The authors found that display items in these journals rely heavily on P values and that the use increased dramatically over the last 20 years. Overwhelmingly and almost universally, P values in figures and tables point to statistically significant results. Importantly, significance testing and P values can only inform the existence of associations, but as spurious associations are almost ubiquitous, the real scientific challenge is to distinguish those that are most relevant for causal inference. Here, the use of multiplicity corrections is one factor that could help to attenuate the ubiquitous prevalence of statistically significant P values. Hence, the authors conclude, that a wider appreciation of the need for multiplicity corrections would be a welcoming evolution, but the rapid growth of reliance on P values and implausibly high rates of reported statistical significance are worrisome.

Leave a Comment

Required fields are marked *.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Scroll Up