Topic: Errors in Science
Comments by Vaibhav Pai
Tufts University
5/29/2015
Proposed Actions
Misplaced incentives in the publishing business is/are the major hurdles for improving reproducibility. There is absolutely no room in the current publishing business for 2nd or 3rd in line of discovery, repeat results and negative results. Not only does this encourage scooping of studies to be the first to publish, push for cutting corners and publish fraudulent results but also actively discourages validation experiments and repeat studies, ultimately leading to one hit wonders that are irreproducible.
NIH has a vested interest in publishing the research conducted using its funds even if it validates or refutes (negative result) previous findings. NIH should invest in creating a centralized (online) publishing outlet (free for NIH funded research/researchers) that accepts individual experiments, negative results and validation results with a post-publication review model. Such a model has already been tried in the field of Physics and is enormously successful in cutting down the publication costs (which is ultimately borne by NIH anyways through their grant money) scooping of research projects, publishing fraudulent results and increasing transparency of research, reproducibility and collaborations among researchers. Such a publishing outlet by NIH would serve as a primer go to source for evaluation of research projects and their feasibility for the entire research community