Last week, NPR’s All Things Considered discussed scientific research methodology, and the losses resulting from such mistakes as unclear instruction, cell misidentification, and undetailed methodologies.
PLOS Biology estimates $28 billion worth of research per year is beset with these issues, however that number is in dispute. Some irreproducible research is still fruitful, plus the magnitude of the errors are not differentiated within that total.
Is our US research model flawed? Making all doctoral candidates train in research is obviously important, but there seems to be a lack of decently paying post-doc research jobs that causes these people to quickly move on to other lines of work after only a few productive years as grad students. Realigning employment incentives just might tighten up the rigor.