The NIH is cracking down on bad data. The US agency announced in Nature that the research field is plagued by data that cannot be replicated, and points the finger not at fraud, but at poor training and the wrong publishing incentives, among other factors.
To correct this problem, the NIH said it is developing training modules that should increase “reproducibility and transparency of research findings.” The NIH also wants to put pressure on academic institutions to demand more of research, and says the current focus on big-name journals and budgetary pressures “encourages rapid submission of research findings to the detriment of careful replication” and says it may rework its NIH grant applications to better define the role each applicant played in the study, among other changes.
