Monday, December 17, 2012

Ground Rules for Sending Scientific Papers for Publication

A veteran and distinguished Computer Science researcher with decades of experience in academia and industry in India, UK & USA shared with me the ground rules he follows while sending scientific papers for publication. Further, he kindly permitted me to share that below for the benefit of interested readers.

Let me tell you the ground rules under which I send papers for publication.

a. I choose a journal or conference which I think is most appropriate for the subject matter (e.g. because of related publications that have appeared there and/or I trust the editor/program chairperson to be fair).

b. When I receive the referees reports (usually 6-12 months later for a typical journal, 3 months for a conference), I take a deep breath and read them.

c. Usually, they start encouragingly and end by saying things I don't want to hear. I put them aside for a day or two and then re-read them.

d. I make a summary of the changes they have asked for and which the editor agrees with in a general way.

e. I revise the paper to meet the objections and resubmit it to the editor, making a list of the changes asked for and made. If I don't agree with any objection I explain why I have not made corresponding change.

The rule I follow is simple: if I want a paper published in a journal, I must satisfy the editor that my article follows most of the recommendations of the referees. I must do this whether I like it or not. It does not make sense to pick a fight with a referee or to tell the editor he or she does not know their job.

If the paper is accepted and appears I invariably see that the new form of the paper is definitely better than the original submitted version. That is true for every paper I have ever submitted and however strongly I first objected to the referees' comments.

Many (perhaps most) papers are rejected after refereeing: good journals and conferences may accept only one of 5-10 submissions. So everyone experiences the rejection of a paper. Of course it is dispiriting and one begins to question everything from the sanity of the reviewers and the editor to their objectiveness and knowledge. There may occasionally be an unfair rejection but in the overwhelming number of cases a rejection is justified because:

a. The paper is just not good enough;

b. The paper has been submitted to the wrong journal or conference;

c. There are errors or weaknesses that lead the reviewers to question the author's knowledge of the area; it does not matter what you think the paper is about, it's what someone reading it concludes;

d. The same results have been reported earlier. Saying that this is the first time something has been done in India is not a valid argument for it to be worth consideration. Science is universal and it does not matter where the work was done or what language it was reported in.

I discovered that the results from one paper I published in the UK in 1986 were rediscovered by  a researcher in Argentina and by a PhD student in the UK, neither of whom had read my paper. They both graciously accepted that their work, while done independently, was done later. It's the job of the referees to be aware of all work related to a submitted paper but they are human and will sometimes not be aware of everything that is done.

There's not much one can do when a paper is rejected except to grin and bear it. You can revise it and try resubmitting it to another journal or conference. That works sometimes but the paper may well end up with some of the same referees!

There may be things in the paper that can be used elsewhere if they are made part of a bigger piece of work with something genuinely new. Or there may not.

By the way, the paper I refer to above was rejected by one journal before it was resubmitted to and accepted by another (it has since had over 800 citations).

No comments:

Post a Comment