Frontloading selectivity: A third way in scientific publishing?

Prestigious scientific journals traditionally decide which articles to accept at least partially based on the results of research. This backloaded selectivity enforces publication bias and encourages authors to selectively report their most persuasive findings, even when they are misleading, biased, and unreliable. One answer to backloaded selectivity is to curtail editorial selectivity altogether, deciding publication on the basis of technical merit alone. However, this strategy is unlikely to win appeal among highly selective journals. A third way is to frontload selectivity—reaching editorial decisions based on rigorous evaluation of the research question and methodology but before the research is conducted and thus regardless of the eventual results. This model, now offered at PLOS Biology in the form of “Preregistered Research Articles” (or Registered Reports), allows a scientific journal to maintain high selectivity for the importance and rigor of research while simultaneously eliminating outcome bias by editors, reviewers, and authors. I believe the rise of Registered Reports among selective journals will change how research is evaluated and may trigger the realization that frontloaded selectivity is the most secure way of advancing knowledge.This Perspective article introduces the concept of "frontloaded" selectivity as a means for journals to remain highly discerning while also eliminating publication bias. The newly launched "Preregistered Research Article" format at PLOS Biology (AKA Registered Reports) is a working example of this principle.