Negative Results Strengthen Science
By Rachel Lane, PhD
Science has a truly terrific storyline: daring discoveries are made, and truth is triumphant. Like every great story, science has protagonists and antagonists. Despite equal investments of time and money, studies with positive results are much more likely to be published than those with negative results. Thus, in the current publishing environment, positive results are seen as protagonists, promoting science progression, and negative results are seen as antagonists.
The problem
This publication environment, in which the majority of research goes unpublished, stunts scientific advancement. Failed experiments one tweak away from success and valuable insights from negative results remain buried in lab notebooks, hidden from peer review. Furthermore, many studies have likely been re-attempted, resulting in wasted time and resources, because results from the first attempt were not publicly accessible. Restricting publication to positive results impedes our understanding of the surrounding world.
The drive to attain positive results also influences study design. To avoid negative results, many studies compare high- and low-responder populations, creating polarized results that do not represent the average cell, animal, or human response. While results from these studies are insightful, their clinical relevance and potential for application are limited. Brilliant, progressive experiments are often tabled and replaced with less risky, seemingly more efficient pursuits that improve the likelihood of success. These practices reduce the maximum impact of science and slow scientific advancement.
When predominately positive results are reported, confidence in research is inflated. On any given topic, the few negative results that slip into publications are overwhelmed by the plenitude of positive results. Even if a negative result is more representative or common in laboratories, a literature review reflects the opposite scenario. This publication practice and the consequential overconfidence may have contributed to the current “reproducibility crisis,” in which a significant portion of published research cannot be reproduced. An outlet for reporting negative findings may have averted this crisis.
How preprints could help
Publishing negative results would enhance our understanding of the world, would potentially encourage more risky, insightful studies and would strengthen confidence in science. Preprint services offer an avenue for sharing negative results: instead of determining the impact value or novelty of the science, this type of platform focuses on quickly disseminating information and does not differentiate between positive and negative findings.
Several preprint services currently exist, but most do not carefully scrutinize manuscripts prior to online posting. Integrating the preprint platform with a basic screening procedure that verifies author institutional affiliation and evaluates manuscript and figure integrity (such as Research Square’s Review Ready Precheck) would increase reader trust in preprinted manuscripts. Such screening may be critical for the sharing of negative results, which will otherwise never undergo peer review.
The prompt dissemination of research studies may avoid further negative results, enabling scientists to pivot their focus instead of belaboring defunct aims. By changing the publishing environment, studies with positive and negative results will function as protagonists of science, supporting and strengthening its great story.
Dr. Lane is an Academic Editor at AJE/Research Square and is passionate about improving science communication. She received her BS in Nutritional Sciences at the University of Oklahoma Health Sciences Center and worked as a clinical dietitian before returning for her MS and PhD in Biochemistry and Molecular Biology. Dr. Lane’s MS work focused on investigating the connection between metabolism and redox balance in chondrocytes, and she created novel sugar polymers and polymer-mediated drug delivery systems during her PhD.
