What Happens to Research Failures?

The Positive Results Bias

Last week I learned about a thing called the Positive Results Bias in research.  

Basically this describes a situation where research that produces a positive result or "significant findings of interest" is considered more worthy of attention in the research marketplace than research producing what is called a negative result. Negative results aren't an indication of finding something bad. The term "negative" is used when the research findings failed to support a new hypothesis, a new theory "doesn't work" or a commonly held belief could not be disproved.   

By characterizing results as 'negative' they become immediately less appealing to research journals and can even reflect negatively on the perceived quality of the research.  This contributes to what is sometimes referred to as the "file drawer effect."  When early research findings show that a hypothesis may not be supported, or the results are negative or inconclusive, the findings may be 'filed away' and never shared with the broader research community.  

This bias is of particular concern in biomedical research where, for instance, the failure to publicize negative results in clinical drug trials could have life and death consequences.  But for any researcher concerned with a balance of rigor and relevance, as I am in the social sciences and research work regarding organizations, this is very concerning.

When research journals and the popular media focus their attention on studies with positive results they do it at the expense of understanding all the knowledge being created through research in any given field.  From my perspective - a scholar-practitioner who uses research as a tool to help organizations be more effective -this is keeping valuable knowledge from the marketplace of ideas.

Change in organizations is constant and influenced by many different factors. A finding of something NOT producing an expected result or findings that support already existing theories and practices can be just as relevant and valuable as, for instance, findings that isolate a statistically significant change in the contribution of one individual factor to a multi-factored result.  

Some of the best ways I can help organizations is to identify barriers to progress and to help eliminate things that DON'T work.  This can often be found in what would be considered negative research results.  When researchers limit the sharing of their work to that which only produced positive results it means that other researchers will continue to invest in needless repeated trips down the same rabbit hole. 

Researchers and research journals should consider a research project to be a success if it is a quality, rigorous, methodology, design, and execution that produces reliable results, regardless of whether those findings produce positive or negative results.