2

I work on both in the area of pure Computer Science and in the area of Computational Chemistry/Biology.

Till now I have submitted papers to well known venues(journals) of both these fields. I don't understand the following situations after paper submission in both these fields:

  • Submitted papers in non-CS journals follow a quick review procedure i.e. submission - editor assignment - reviewer invited - under review - review completed - [...] Turnaround : 3 months

  • But, for pure CS Journals, every step is late except the first step which is obvious. Turnaround: 10-12 months

What can we infer given such situations in peer review world?

Note: I am talking about majority of journals in both these mentioned fields.

11
  • 5
    You can infer, that peer review in CS takes about three to four times longer than in some other fields. What kind of answer do you expect?
    – Dirk
    Sep 22, 2016 at 14:19
  • 1
    Thank you for your enlightening comment. I didn't know basic multiplication. @Dirk
    – Coder
    Sep 22, 2016 at 16:29
  • 1
    Not of answer quality: I was recently chatting with two chemical biologists about this difference. They told me that their reviewers normally only have to check for factual correctness, and do not have to tell whether the result is novel or interesting enough. This is the job of the editor. In a CS review, this "novel/interesting enough" statement is the most important part of the review (unless the paper is plain wrong)! Testing for errors is the easy part. Also, journ. articles tend to go into much detail (compared to conf. articles), so reviewers also often need longer reviewing times.
    – DCTLib
    Sep 22, 2016 at 16:33
  • This somewhat makes sense. But, are these guys saying that the biophysical journals do not check the novelty and originality? - @DCTLib
    – Coder
    Sep 22, 2016 at 16:36
  • 1
    @Coder No, they are saying that determining this is the job of the editor. So the reviewer have a lighter workload. And if you have eager editors or full-time editors (Nature, Science, ...), this speeds up the overall process.
    – DCTLib
    Sep 22, 2016 at 16:38

3 Answers 3

4

In computer science, unlike in the majority of other academic fields, most research is published as articles in conference proceedings, and conferences do follow a quick review process. I imagine that the pressure on journals to provide a quick review process is thus less strong.

4
  • If this is the case then why number of journals in CS are increasing day by day. I am talking about journals from reputed publishers such as IEEE, ACM, Springer, Elsevier...
    – Coder
    Sep 22, 2016 at 16:31
  • There are lots of publication venues, of varying quality, whether they are conferences or journals. Journals have their place (for surveys, for extended versions of a paper with proofs or detailed experiments, for some papers that may not fit well with the conference format...), but conferences remain the main way to publish a paper in computer science. It can be infuriating to have to wait more than a year for a journal review, but at least you know that you won't have to wait more than a couple of months when you submit a conference paper, which is much more frequent. Sep 22, 2016 at 16:55
  • Thanks for the comment. But, the problem with the conference is the systematic way of registration, presentation etc. which is large cost bearing for a research student. However, there are funding opportunities but still are less frequent.
    – Coder
    Sep 22, 2016 at 16:59
  • "why number of journals in CS are increasing day by day." to make lots of money from research councils. It is a simple as that. The problem this causes is that the quantity of competent peer-reviewers has not increased very much, so it becomes difficult for journals to find suitable peer-reviewers for good papers. Jun 3, 2022 at 15:56
1

The time to a decision in Computer Science depends on:

  • The editor assigning an associate editors (at an IEEE journal, this is about 1-3 days).
  • The associate editor reading the article and starting to search for reviewers. The first is about 1-3 days. Finding the reviewers is the difficult thing and it is becoming more difficult. (1) Potential reviewers do not want to review. Sometimes this is because they have reviewed a lot, sometimes they only review for the top conference in their field, sometimes they are just free riders (that benefit from peer reviewing but do not review themselves). (2) Potential reviewers have left the place where they graduated and are not forwarding email. Universities and companies do not see a need to maintain accounts for graduates and ex-employees and even if they allow people to forward, the forwarding becomes complicated after a few moves. These reviewers never answer. (3) Potential reviewers actually give a good answer. Most of these will accept the review.
  • The associate editor goes through a few rounds of inviting reviewers. You cannot really invite more than ten people to review because then more than three might accept. But for a IEEE journal, maybe one in ten reviewers asked will accept if they do not know the associate editor. Journal rules prohibit asking people at the same place as the reviewer or asking the same people for more than a few reviews. They now also do not allow the associate editor or sometimes another associate editor to review in order to strengthen the quality of the peer review process.
  • The paper is under review once the first reviewer has agreed.
  • Associate editor informs senior editor that they are doing their best with finding reviewers to the paper.
  • Finally, enough potential reviewers have accepted to review.
  • After the deadline passes, associate editor gives extensions to reviewers.
  • Finally all reviews are in. Associate editor now does the meta-review for the senior editor. This can take a week as the associate editor needs to reread the paper and read the reviews.
  • Senior editor comes to a decision. Very often a revision with additional reviews.

Summary: Peer-reviewing is broken. The bottleneck is in finding reviewers. Too many academics and industry people are free-riders, largely because there is no good reward structure. There are some people trying to change it, but for example, industry does in general NOT reward reviewing. Contact information is difficult to obtain.

0

The time for peer review is a highly complex variable and it depends on many factors: the ideas explored in the paper by the authors, the time for review and the focus of the referee, the dynamics of the journal editorial board. For example in physics(APS) the time for a review round may vary from 1 week to 2 3 months, or even longer. Hence, the time for publishing vary from one month to one year, or more. In mathematics, for example, the review time is even longer, and it is natural that the referee reports are received in a time frame from 1 month to one year. From all these, we see that the review time is a complex variable where the authors have minor effects and one cannot change much. The best thing an author can do is to write a clear manuscript that would be natural to read/referee. Hence, the review time is not an important aspect to consider, what is more important is the idea and the novelty of the paper, and the quality of the review process, which is linked to the prestige of the journal. For conference proceedings, the review time is shorter that the journals, where a strong review is made. For sure, the strength of the review highly depends on the editorial board and the referees.

2
  • Consider this example: some paper P1 is submitted to a CS Journal J on September 1. The Editor is assigned on September 17. The status remains the same for 1-2 months. Then it goes for review.
    – Coder
    Sep 22, 2016 at 16:34
  • @Coder in some fields it is easier to get competent reviewers than others. If a sequence of peer-review requests are declined (or get no response) then there can be a lengthy gap between submitting the paper and it going out for review. Jun 3, 2022 at 15:59

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .