Logo IMG
HOME > PAST ISSUE > Article Detail


Imperfect Peer Review

To the Editors:

We add two comments to David Resnik’s essay “A Troubled Tradition” (January–February). First, it seems that journal editors increasingly are reluctant to make judgments as opposed to decisions. In the old days, editors who faced conflicting reviews would read the paper themselves. These days, the trend—at least for some editors—seems to be to ask for more reviews to break apparent ties, perhaps a reflection of the increased specialization of research. As a result, editors have become more like boards of elections. (It would be good to hear if others have experienced this because we understand that our experiences are not statistically significant.) Second, the proliferation of journals and the demand for more reviews (see above) has placed an increased burden on reviewers, particularly the conscientious ones. Academic institutions, although supportive of their faculty’s peer-review activities, are not developing reward criteria that take into account the time needed to execute those activities properly.

Edward J. Behrman
Venkat Gopalan
The Ohio State University

To the Editors:

Reading David Resnik’s essay, I remembered that as long as 60 years ago, when we submitted a paper to a journal for review, it was referred to a competitor in the scientific community. That introduced the possibility that he or she would either delay the paper, in order to have his or her own paper published first, or would totally reject the work. At that time there was no means by which an author could select the referee.

Peer review is a two-edged sword, with ethical reviewers making the needed suggestions for change and expediting the publication process. Unfortunately, a few bad apples will always cast doubt on the validity of the review process.

Nelson Marans
Silver Spring, MD

To the Editors:

David Resnik’s thoughtful and useful discussion of some of the problems associated with peer review did not discuss one important problem for journal editors and readers. A review of any scientific paper is likely to require several different types of expertise. Research teams are often large and complicated, with multiple disciplines, including statistics, contributing to the result. Some authors may not have a deep understanding of the underpinnings of the work contributed by some colleagues. This is not a criticism of the practice of science; it is recognition of the complexity of what we do. No one can be expert in everything.

Editors should choose reviewers with a range of perspectives on the research and with sensitivity to different types of problems. Most experienced editors do just that. A divergence in expertise should often produce a divergence in reviewers’ comments and in their judgment about whether publication is appropriate or how it could be improved. Nearly complete agreement among reviewers is far from ideal. It may be a sign that all but one reviewer is redundant.

John C. Bailar III
Washington, D.C.

comments powered by Disqus


Subscribe to American Scientist