Can and should peer reviewers catch fake data or questionable information in a scientific research paper?
While I would rather use “fraud in science” than “scientific fraud” to refer to this issue of quackery in science that seems far more prevalent than is reported in media, the problem is serious and demands urgent attention.
A recent blog in Discovery Magazine titled “Should Peer Review Catch Fraud?” brought up the issue, thanks to the publication and the writer. It correctly pointed to the sad fact that reviewers are not much interested in ensuring accuracy but more in interested in publishing the paper due to the incentives involved.
While the blog cites the case of a retracted cancer biology paper in the well-known journal Nature Cell Biology, firsthand observation and careful reading can surely reveal a lot of this fake data issue in published papers, many in high-profile journals. The case cited in the blog in question is one where something faulty was included and did not get caught.
However, there are instances when researchers omit data or manipulate it to make things look good on paper, thus easily getting away with it. Reviewers, even if they are careful in their reviewing work, are unable to catch such fraud. Only whistleblowing can reveal that it was fraud ad whistleblowing is rare – everybody wants to stay safe and keep their job.
The fraud of national and international level in vaccine research as exposed in the film Vaxxed, showing CDC’s manipulation of data pre-publication of the research, is a sore example of this problem.
So what is a good solution?
I think every research needs independent regulation/monitoring from a non-affiliated person or group of people from the same discipline who are present and recording for public viewing all steps – a to z – in any research project. Data needs to be gathered, processed, and published in their presence. This can likely reduce – not totally eliminate – the practice of manipulating data. And surely it will ease the work of reviewers.