Statistical Guidance to Authors at Top-Ranked Journals across Scientific Disciplines

Scientific journals may counter the misuse, misreporting, and misinterpretation of statistics by providing guidance to authors. We described nature prevalence statistical at 15 (top-ranked Impact Factor) in each 22 scientific disciplines across five high-level domains (N = 330 journals). The frequency varied (Health & Life Sciences: 122/165 journals, 74%; Multidisciplinary: 9/15 60%; Social 8/30 27%; Physical 21/90 23%; Formal 0/30 0%). In one discipline (Clinical Medicine), was provided all examined two (Mathematics Computer Science) no guidance. Of 160 guidance, 93 had a dedicated section their author instructions. most frequently mentioned topics were confidence intervals (90 journals) p-values (88 For six ‘hotly debated’ (statistical significance, p-values, Bayesian statistics, effect sizes, intervals, sample size planning/justification) typically offered implicit or explicit endorsement rarely opposition. heterogeneity top-ranked within between highlights need for further research debate about role can play improving practice.