In a previous blog, I talked about the limitations of peer review in catching cases of scientific fraud. Science has traditionally been considered a self-policing field, but if our peer review process cannot determine if a paper is fraudulent, how can we catch these scientific criminals? The general answer to this question is that results that cannot be replicated will eventually be proven false because of the nature of the scientific method. However, in the real world, how often are studies replicated? Outside of a few high-profile research areas, replication seems unlikely. This is a consequence of the funding system, which emphasizes novel research. Imagine submitting a proposal to replicate previously published work – what is the likelihood that proposal would be funded? As a consequence, it is likely that these fraudulent results will remain in the system, leading other researchers down unproductive paths. As one researcher put it, “it’s like an epidemic, in the sense that they’re infected with these wrong ideas, and they’re spreading it to other researchers through journals.” How can we fix our system? When I talked about peer review, I suggested that more transparency in the scientific process was needed. Another possible answer might be setting aside funds for proposals designed to replicate research, although I’m not sure how these studies would be selected.
What do you think? Do you think science is self-policing? What are your suggestions for catching scientific criminals?
I agree with you that sometimes it is harder and sometimes not as beneficial to replicate the results, as you may be interested only in somedata, but a good validation is very important. One way to go about it would be to stop using the results based on single study. One may question that since the results have not been replicated its not statistically singnificant to be used and thus forces researcher to replicate elsewhere or be certified from another place to establish authenticity.