When experiments don’t go as expected, negative results and unsupported hypotheses often remain hidden in lab notebooks or forgotten in file drawers rather than becoming part of the published scientific record. However, this information is still an integral part of the scientific process. Should these findings be published? If so, why? The Scientist asked its readers to weigh in on the value of publishing negative data.
Ashutosh Tiwari, Protein Chemist, Michigan Technological University
Yes! The publication of negative results is very important. Such findings offer a critical perspective by demonstrating that a scientifically plausible hypothesis or research direction may not yield the anticipated outcomes under rigorous testing. Disseminating these results contributes to a more comprehensive and accurate scientific record, reducing bias toward positive outcomes.
Publishing negative results conserves valuable time and resources by minimizing unnecessary repetition of unsuccessful approaches. When researchers are informed that a specific hypothesis has already been tested without success, they can refine their research questions or explore more promising directions, rather than duplicating efforts that have proven unproductive. This practice is especially significant in the context of substantial federal investment in research, as it promotes more efficient use of taxpayer dollars.
Nora Stevens, Biologist, Portland Community College
If negative results are not published, others do not know that a particular pathway doesn’t work and may repeat pointless work. Alternatively, others could offer a new perspective on an apparently failed experimental strategy that could lead to a positive outcome.
Riekelt Houtkooper, Biochemist, Amsterdam University Medical Center
It can be quite difficult to publish a negative story. I'm not even sure you should call it negative data because it's actually confirmation that something doesn't change: In a way it's not negative; it’s also positive. Ultimately, methodological robustness matters. There should be more emphasis on reproducibility and robustness of data.
Ana María Cetto, Physicist, National Autonomous University of Mexico
Absolutely, negative results should be given their place too. Often, they tell us more than positive results do. They can falsify a hypothesis or theory. Failed procedures can teach us valuable lessons too. They help others avoid making the same mistakes. In short, sharing our mistakes and negative results helps science progress.
Elena Okon, Biologist, University of British Columbia
Yes, especially if the results were expected. Otherwise, several other laboratories would lose time testing the hypothesis again. Also, it is necessary to publish it if the negative result does not confirm the positive result previously published.
Lara Campana, Senior Vice President of Head of Research and Translational Science, Resolution Therapeutics
They should be published for several reasons. Good scientific methods often bring about a rejection of a hypothesis, and this should be shared with the scientific community. Trial and error are at the basis of good assay and experimental setup development, and sharing this could streamline other people's work. Learning from others’ work should be encouraged as it promotes better utilization of the resources and compresses research timelines. Lastly, a PhD or postdoctoral researcher who has worked diligently for years on a project ending in rejection of a hypothesis should be able to share the good science they have developed in a publication, which could be crucial for their career.
Qiaoxiang Dong, Toxicologist, California Department of Pesticide Regulation
Absolutely. In my opinion, negative results are more important than positive results. Scientific publications nowadays are biased with the preference to publish positive findings. By giving negative results equal opportunities for publication, or even placing greater emphasis on them, science in any field can be advanced in a truthful and objective manner.
Donna Thompson, Psychologist, Niagara University
Yes, otherwise many people will try to collect similar data and also fail. Preregistration is also helpful as scientists can check there to see what failed to get published.
Kevin McCully, Professor Emeritus of Kinesiology, University of Georgia
Yes, and papers with negative results should be held to the same data and writing standards as “positive” result papers. This provides a more accurate evaluation of methods, conditions, and treatments.
Jon Richfield, Entomologist, Biochemist, and Computer Consultant, Retired
Certainly. The function of science is to improve our ability to select the most effective working hypotheses concerning the empirical world. Knowing the outcome of experiments is one of the major components of that ability. Not knowing which experiments gave negative results impairs that ability.
Paul Scinto, Molecular Biologist, Retired
Published? Yes, but not in the way a successful paper is published. Negative results are important to determine what does not work. This information is important in eliminating variables. The less the number of variables in any given experimental process will improve the chance for a more successful experiment. A listing of negative results will also save resources and reduce environmental contamination by reducing the number of unnecessary experiments which are known to fail. Perhaps some type of database can be established to help researchers know what not to try.
Thomas Young, Anthropologist, Retired
I was taught that negative results are just as important. Your initial hypothesis did not explain your observations and results. One must explain your first hypothesis and outline how you formulate a new one; changes in your methods of data collection and measurements are necessary to better explain the initial or observed phenomenon. I think publishing negative results that don't fit into your hypothesis may be of greater import in archaeology as one does not expect to uncover or detect all the information you would desire. It is also why a multidisciplinary approach makes so much sense if you interpret your data or derive new data with other methods.
Responses have been edited for length and clarity.
