Does the removal of results from a submitted paper reduce publication bias?

Abstract

A recent paper by Findley et al. (2016) in Comparative Political Studies suggests that the removal of results from a paper may decrease publication bias. However, in the biomedical sciences, a paper without results is not interpretable; therefore, such a solution is not viable for addressing the reproducibility crisis. Instead, Findley et al.’s proposal should form a pre-submission step that enables colleagues and peers to evaluate a paper's experimental design and protocol prior to submission to a journal for regular peer review. Introducing a new, data-free model for peer review would only dilute the efficacy of current models and weaken efforts to improve existing publication practices.