Lessons from a failed research project: what’s in a name

Abstract

This essay discusses and reflects on a project that was based on a simple clinical question. Despite easy access to routinely collected data we were not able to answer our question. Studies in which one concludes that the research question cannot be answered are probably common, but are rarely published. We believe that there are also valuable lessons to be learned from such research. In our essay both the benefits and general and some more specific pitfalls are highlighted using an example from our own practice, followed by a reflection with recommendations. Potential pitfalls of being able to test different hypotheses quickly and easily are discussed including an increased chance of random findings and the risk that only positive findings are shared. We emphasize the importance of a priori hypothesis generation, pre-registration and preprints and the sharing of the research process and findings, even in the case of negative findings.