On 7/9/07 9:53 AM, "stan moore" <[EMAIL PROTECTED]> wrote: > What is the minimal obligation of any peer reviewer for any scientific > publication? For instance -- must the peer reviewer read all the reference > materials? No: the time commitment could be enormous. But presumably the reviewer is someone in the field, and so will be familiar with most of the references. I would say the "minimal obligation" is to read and professionally evaluate the ms, while disclosing any conflicts of interest, within the time period requested by the journal. > How much time and effort should go into any peer review? I expect, on average, to spend 2-5 hours on a ms, including writing the evaluation. > > What should be the ultimate and subultimate goals of the reviewer? ?? > > For instance, is the overall objective of the paper under review subject > to question? Sure. Perhaps it isn't appropriate for the journal, for instance. > > Is peer review more superficial and mostly a spell-checking and grammar > review excercise? Not in my experience. > > On what basis should peer reviewers be chosen -- species expertise, > technical application of scientific method? Either or both. Ideally a ms is read by many reviewers who bring multiple points of view to bear. > > Does reviewer acquaintance, friendship, collaboration with author(s) > constitute bias? Yes, and thus it should be disclosed. But it doesn't necessarily disqualify the reviewer. > > How should obvious errors in peer review after publication be dealt with? Can you give a more concrete example?
> These are questions that enter my mind after perceiving many failures in > peer review while reading the scientific literature over the years. I'd be interested in what for you constitutes a "failure" of peer review. Such failures certainly exist, but the most notorious cases (Henrik Schon, for instance) involve fabrication of data. Peer review can't be expected to catch this: reviewers and editors must assume basic honesty on the author's part. Ben
