The First Step is Admitting That We Have a Problem

Many scientists have written about the lack of reproducibility in scientific research. This inability to reproduce, according to an article by Jared Horvath, has been the case since the 17^th century. With this in mind as well as the Bayesian analysis that was presented by Jeremy Berg’s article, the scope and reasoning behind unreproducible results finally makes sense to me. While these articles delve deep into the explanation of this problem in science, they do not seem to suggest any detailed next steps or actions that the scientific community should take to rectify this pervasive issue. They simply suggest that scientists need to work hard to make sure their work is “as reliable as possible” and that there need to be open lines of communication between scientists.

This made me think: what can we do to help improve reproducibility as well as transparency in our research? I think that it would be beneficial for journals to request that researchers publish very detailed supplemental materials and methods sections. As someone who has had experience in the biomedical device field, I have seen the advantage to having EXTREMELY detailed protocols to make sure that the lab technicians and research and development teams can properly test their products. This prevents large variability between those conducting the tests and variability by testing site in the results. Having a protocol for a paper published could also alleviate some of the issues seen when replicating results, as well as allow for reviewers to further scrutinize the publication to avoid any mistakes being published. Overall, I think the first step to addressing this issue of reproducibility in science is being aware that we have a problem, but I think now that this is understood, steps need to be made to begin to rectify it.