Peer review is designed to act as the protector of science.
The idea is quite simple; experts anonymously review a submitted body of
scientific evidence and evaluate its suitability for publication. However, in
reality this process fails. This flaw in science’s quality control mechanism
was recently discussed in The Economist's “Trouble in the Lab,” in which the
author writes that the peer reviewers who evaluate papers for journals aren’t
very good at catching errors.
Many journals are aware of this issue and are starting to
take action to address the failings of peer review. For instance, the European
Journal of Neuroscience is now publishing peer reviewer’s comments along with
their names online. This removes the anonymity from the process, and places
considerable pressure on reviewers to critically evaluate manuscripts. I like
this idea and I think that it will improve the quality of reviews and the peer
review process. Another option to address the peer review bias, but maintain
the sacred anonymity is to use a blinded approach in which the reviewers are
not provided the authors’ names and institutions. This would remove inherent
bias in which a reviewer may alter their evaluation of an article due to an
author’s celebrity or affiliations.
What else can be done? As PhD’s in training we should be
taught how to critically evaluate an article’s experimental methods, and it’s statistical techniques. In reality, most scientists are not
statisticians; thus, it is difficult to evaluate a paper’s methods without
formal training in the use of statistical methods in biomedical research. Including
more applied statistical training, in addition to the classical mathematical
statistical training, within the biomedical PhD curriculum is needed. Ultimately,
we must be taught how to look at science with a skeptic’s eye, or else the peer
review process will continue to fail. This includes training in statistics,
along with the classic training in experimental techniques. The concept of peer
review is great; however, its time to improve the status quo.
No comments:
Post a Comment