(Nature386;319 (1997)) Copyright Macmillan Journals Ltd
Robin P Clarke
Sir---Simon Wain-Hobson's correspondence item (385, 384) is by no means the first expression of dissatisfaction with the refereeing of papers, and won't be the last. I second his proposal for reviews to be of a sufficient length. Say 10% of the paper's length for a rejection, 5% otherwise?
There is a change that could not only improve the quality of reviews and consequent decisions, but also make the editorial decisionmaking process more efficient. The idea is for editors not to examine reviews initially but rather to send them immediately to authors for response, and then if, and only if, the author feels he or she has a credible defence against the reviews will she or he then submit his or her response to the editor for consideration. Reviewers would then know that they cannot get away with sloppy or malicious reviews. And editors would be saved the work of trying to weed out poor papers that can slip through the less controlled review process that is the current usual norm. This could go some way towards reducing the massive volume of poor-quality papers being published (other than in Nature of course!).
Meanwhile authors could copy the trick I once used, of compiling a collection of previous reviews of the paper along with my rebuttals, and sending that with the submitted manuscript to the next journal. It worked for me. The paper involved was described by a reviewer for Personality and Individual Differences as well-written, well-argued, and well-documented, whereas a British Journal of Psychiatry reviewer reckoned it was of lowest grade in all three respects! It is difficult to see how this gross discrepancy can be accounted for except in terms of maliciousness on the part of the latter reviewer, such as I encountered with many previous reviews.
Robin P. Clarke
Back to INII Homepage