Omid Mahian and colleagues suggest that scientific celebrity does not seem to influence the editors' choice of Correspondence items for publication (Nature 519, 414; 2015), but their informal analysis has caveats beyond those they mention.

Famous authors may be more disposed to write letters to the editor than is the rest of Nature's readership, for example. Any analysis for editorial bias should include data from rejected contributions as well (which were obviously unavailable to the authors). Then, applying a logistic regression or calculating an index of niche breadth, such as the proportional similarity index (P. Feinsinger et al. Ecology 62, 27–32; 1981), would clinch the matter.

Perhaps a simpler demonstration of a lack of editorial bias might be the publication of this note, signed by a student who has no publishing record at all.