Reaction to "A few things that would reduce stress around reproducibility/replicability in science"

There is a major issue regarding reproducibility and bias in science. I believe the primary causes of this issue are the limited funding for researchers and the lack of science knowledge by the general public. Together, these cause strong incentives for scientists to publish “high-impact” papers containing novel, important results and oversell the implications of their work, leading to decreased research quality and a lower likelihood of reproducibility. Science is supposed to be a noble pursuit of truth and knowledge for the good of humankind. However, science is performed by people who have personal career ambitions. Naturally, this is a source of trouble for many researchers. For example, it is very easy to publish underpowered studies, particularly when the statistical tests used are complex and might confuse a peer reviewer. Also, there is much more incentive to perform novel experiments rather than reproduce prior experiments, even though reproducing prior results is extremely important.  

            I enjoyed reading the article A few things that would reduce stress aroundreproducibility/replicability in science, by Jeff Leek. In particular, I strongly agreed with the author’s emphasis on acknowledging the difference between exploratory and confirmatory research. A more structured and organized way of confirming exploratory scientific results would be more useful than the current system. I also agreed with the author’s point that “data is valuable but in science you don’t own it.” Very often, scientists become attached to specific results or theories and consciously or unconsciously bias their work towards confirming these beliefs. A good scientist should be judged by his or her ability to reason and interpret results. However, it is extremely difficult to assess researchers’ abilities in these areas, so most researchers are judged on their publication history, again incentivizing less reproducible work. I strongly disagree with the author’s suggestion that we should be more private about our work and data. I believe that a push towards more open data sharing and experimental collaboration is crucial for solving the reproducibility crisis.

            Ultimately, I think improved funding for basic science research, as well as a strong, conscious effort to address these reproducibility issues, would go a long way in improving the quality of science being performed nowadays.