Occasionally, but too often, I see a paper that has been really badly refereed. The referees (and of course the author(s)) could have had no knowledge of the earlier literature. The authors — regrettably there are few single-author papers published these days — are allowed to foist their wrong interpretation of their results on unsuspecting readers. In turn, the justification for spending time energy and money on the work in the first place is also deeply flawed.
The reputation of the journal as a repository of good science is no guarantee of good refereeing. I saw a shocker in Science recently while the worst I ever recall was in Proceedings of the National Academy of Sciences of the USA (PNAS).
Poor referees allow myths to be perpetuated and refuted hypotheses to be regarded as extant. Peer review is only effective if referees are chosen who are able to do the job properly. A major problem I see is that with the horizontalisation of the biological sciences, fewer of those working at the lower levels of biological organisation have knowledge of the higher levels. Molecular ...............ists (you add the field) seem to me to be the worst offenders. Ignorance of tissues, organs and whole organisms seems to be bliss.
I smile (with an accompanying degree of annoyance) at the efforts of global warming evangelists to promote their cause by writing to newspapers pointing out that certain findings cannot be ignored by governments and opinion formers because, 'it is peer-reviewed science'. They do the cause they espouse no good and open it to derisory attack. Peer review only lessens the possibility that research findings and their interpretation are wrong. Uncertainty is the name of the game. Good peer review can lessen the degree of uncertainty but review by poor referees can perpetuate ignorance and uncertainty.
The problem for a readers of research papers and those making evidence-based policy is how does one know the quality of the referees who reviewed the original paper and recommended its publication?
