“Well, we failed to reject the null hypothesis for our experiments. I guess we should just put that submission idea on hold for now since we have no good results to report.” – the Professor says to his advisee in a frustrated tone while flipping through some graphs. Another one bites the dust! Hours of intense work and significant monetary expenditures become a victim of the “file drawer effect”; or to modernize the expression: “the hidden computer folder”.
Unfortunately, this is not an isolated incident. Fanelli (2012) demonstrates that negative results have been gradually disappearing from academic literature over the past two decades. Meanwhile, articles primarily and clearly stating positive results have grown 22% between 1990 and 2007. As positive results are more likely to lead to prestigious publications, discarding odd and unexpected findings is common in the scientific publishing system that privileges these “successful” results. Traditionally, it is expected that successful studies will obtain research findings in alignment with well-established literature or expected outcomes.
However, are negative results indeed meaningless? Or is there potential value in sharing negative results with a broader academic community?
When examining reasons for the low profile of negative results publications, one common assumption is that publishing negative results might harm scientists’ reputations. Along these lines, negative results are believed to indirectly communicate to the scientific community that a study was poorly designed and researchers were either unknowledgeable about the phenomenon or incapable of tailoring more robust research hypotheses. Moreover, the discouragement to submit publications reporting negative results is due to a higher likelihood that these papers will be filtered by the peer-review firewall, given their perceived lack of soundness in comparison to studies with “successful” results.
Popper (left) states research is a “voyage of discovery”, which is subject to unpredictability and fallibility (Popper, 1963, p.33). Science evolves according to testability, which might result in refutations or confirmations. Thus, the absence of anticipated correlations should also be counted as important and publishable results because they advance science.
One recurring misunderstanding among scientists is that negative results are equivalent to bad results and are products of flawed or ill-designed science. This quality argument can be easily challenged, because even positive results are not exempted from having their quality questioned. Furthermore,, some of the most well-known examples of fraud and data fabrication in science (eg. Jon Sudbø, Hwang Woo-suk and Marc Hauser) were associated with studies claiming positive correlations or supporting researchers’ anticipated hypotheses. Instead of suggesting poorly conducted science, negative results can indicate novel findings or unexpected outcomes of rigorous scientific investigations, directly or indirectly contributing to scientific discovery.
A classic example of negative results being recognized by researchers as scientific and ushering in a paradigm shift in science took place in the 17th century. Albert Michelson (right) and Edward Morley conducted sequential experiments in the late 1880’s seeking to enhance the accuracy of the prevalent aether theory. All of their efforts to advance the theory led to a continual rejection of their research hypotheses. A few years later, these null results were published in the American Journal of Science and played an important role in inspiring new experiments, including a well known one that confirmed a major physical theory proposed by Albert Einstein in 1905: the special theory of relativity.
Examples like this one suggest that reporting negative results creates an information sharing opportunity with peer scientists that can potentially inspire new directions for further studies. This example also suggests that as far as reputation, negative results should not be mindlessly connected to unskilled researchers.
The potentially positive contributions of negative results to the scientific community and society at large lie in the possibility of reducing duplication of effort, and inspiring further investigations with new methods and/or variables. There is no doubt that journals devoted exclusively to publishing negative results still have to improve their visibility among scholars if they are to be widely accepted. These journals face two major challenges: overcoming the traditional, outdated, and vicious scientific publishing enterprises that have been prevailing for generations and convincing others of the need to improve the dynamics and transparency of the scientific information flow.
After all, it would be interesting to speculate what the current state of Physics and other sciences would be if more negative results such as Michelson and Morley’s had been published.
Written by Gabriella Anderson (original source)