To the best of my knowledge, the editor assigns the reviewing process to 2 individual reviewers, after the initial manuscript selection. However, 2 is considerd to be the minimum number acceptable, whereas the maximum number of reviewers is not always defined. So, an editor can choose to adress to many reviewers if there is controversy, to improve the scientific rigor of a publication or just to make sure that at least 2 reviewers finally accept to spend time on this task. In this case, no time is wasted if someone declines the invitation, as those who accepted will deliver their feedback earlier than they would if been invited after the first invited reviewers declined. However, if there are not many declining, it is possoble to receive multiple feedback that, if the article is not controversial, will not be of much help or importance...

Dear Ioannis, thank you very nice suggestion!

In our case, that is the number of real reviewers with their comments!!!

Dear all!!!

Paper is published recently in another journal without any comments from reviewers!!!

"Surgical trauma and ..."

Of course there are many different opinions in any case but sometimes we meet engaged editors independently of journals impact factor!

Dear Ospan,

A lot of differences according to journals.

In general, and depending on the list of reviewers of the journal (specialits in different fields of research to cover the different possible subjects published in the journal), they assigned 2 or 3 reviewer s (best is at minimum 3 if the advices are not concordant in order the editor-in-chief can decide, if is not an expert of your topic). But finally, that's the editor who decides and reviewers have only consultative advices.

The problem is that the editor, who first select your paper on the basis of the journal policy, number of solicitation, to ssend it to reviewers can change his/her first advice after receiving the reviewer comments.

Some reviewers spend time to analyze your data, others few (also according to their notoriety and time thet have).

But you have to remember that you can disagree with reviewer comments and if arguments seemspertinent oto the editor, he can assign another reviewer (it happended for me), especially if you demonstrate that a reviewer has not really deeply red or analyze your paper.

Finally, "problems" occur perhaps with less recognized jjournals, obut also by with very high impact factor where editor send your laconic negatice answer with no comments.

Best regards

Didier

Conference contributions are often evaluated by just a single referee. Positive recommendations are almost never contested, but when the (only) review is negative then, usually, conference organizers want more opinions, which are made by the sub-team of their scientific board.

My personal experience with regular journals is that the highest number of referees was 5; two of them finally declined to produce any evaluation after more than 5 weeks of considerations. As the 3 remaining opinions were "mild", neither positive nor negative, the paper was rejected. The voice saying "this is nothing new", without any pointer to substantiate this claim, appeared thus decisive to editor.

But I also have another experience: the more you discuss with referees, the more citations your paper will have. After it is finally published, of course. Those published "as is", no questions asked, are much less likely to be cited, if at all.

@Marek, interesting comment. I was wondering how you discuss with referees, if there's a double-blind review process.

@Michael: double-blind review is a future. For now authors' names are usually known to the referees. Anyway, there is a "middle-man", i.e. the editor, who is communicating the referees' opinions to the corresponding author, the need will be. And awaits the author's response, often within strictly prescribed time limits.. His role is not limited to sending the final decision only.

I have (seldomly) experienced review with 6 reviewers. In one of these cases, the reviews were between "reject" and "major revisions", so that might have been the reason for the high number of reviewers. Btw: it was a journal with double-blind review, so such systems are existing. Although, if you are common with the topic (as you should be as reviewer), you still might find out (or simply know) who the author is...

Usually 2 Reviewers, but could be 3

Each journal uses their own set of rules, so there are no 'general rules'. That being said there USED TO BE some form of consensus in the old days when manuscripts were still being sent to reviewers via express mail. In those days the editor had a quick look and if he thought the MS wasn't worth sending out to reviewers he/she would do an editor reject in order not to have the authors wait for something that probably was going to be a rejection. He/she would then sent it out by express mail to two potential reviewer who would get 2-4 weeks to write a report and then sent the original manuscript (incl photo's an figures!) back to the editor. The reason there was thus only two reviewers was that the authors needed to sent 3 copies of the text and the original photo's & figures. One for the editorial office and the two other ones for the reviewer.  

Now times have changed and almost all journals use electronic procedures for the process. In the initial period (up to some ten years ago) this went quite OK as most reviewers had experience with the old way of doing things and still felt that not doing a review was not the way to go (compare to the old days - you had to return the express mail envelope back to the editor without even opening it - and this would cost serious money and cause serious delay and reviewers were very much  aware of this!).

Now you receive an invitation to review by email. In my long experience as editor for several journals many potential reviewers (>50%!!)  do not even bother to read this email (or provide a negative reply to it). Those who do reply mostly inform the editor that they do not have time, so only a handfull informs me that they are willing to review the MS but miss the deadline and then fail to answer any follow-up emails of the handling editor. So slowly it is starting to become more and more common to invite 3-4 reviewers in order to meet the demands of the authors who demand a fast return of the verdict on their MS.

If these authors would only realize that reviewing is not going to happen automatically AND it is a voluntary, unpaid job they would perhaps next time read the invites to review a paper and preferably accept this or in minimum reply to the editor that they cannot do it this time..... This system of peer reviewed MS can stay alive. If not the system is going to die and we need no longer worry on how many review reports there is going to be.....

Dear Sir,

            I am agree with you. Same thing happen with me four-five times. 

I have put one GUI to demonstrate the real time real time performance of the proposed approach. Five reviewers was invited by the editor to review my paper. Four reviewer said that the GUI demonstration makes the paper very effective and accept my paper but fifth reviewer said GUI can not be the part of the journal papers and he recommend rejection of the paper. The shocking thing for me was that The editor did not ask me to remove the GUI and reject my paper.  

Eight reviews is a lot. I had a paper get reviewed by three reviewers and one reviewer said a key model had to be removed from the paper, we said the reviewer had a conflict of interest so "fortunately" the editor agreed and then sent it out to three new reviewers and all three came back positive so the paper was accepted after a total of six reviewers. More recently papers have been reviewed by only two reviewers. There are plenty of good journals, I would revise your paper based on the reviewer's "good/useful" comments and then resubmit to a different journal. I recently submitted a paper to five journals (one after the other) and after each journal's review of the paper it got better. In the end I was glad the paper went through so many journals/reviews since it became a much better publication due to the reviewer's comments and the time I had in between reviews to add additional experiments and round out the discussion/conclusions.

Yes , In the first set of review all the reviewers have given comments which we included to the best of our knowledge. In the second set the reviewers totally changed and discarded the comments and answers of the first set of reviews.

I have a little experience to appropriately answer this question, but I feel like giving a feedback based on my perception.

I think the ideal number should be 3 to 4, and strict minimum 2, and friendly maximum should be about 5... Now when a review is done with more than 6, it is unfortunate for this review to have an outcome that will be normally distributed between minor revisions and rejections like I had an experience of been among the 23 reviewers (that was too much) of a Q2 journal article for a manuscript. In this case, most of the editors will just reject the manuscript.

But I highly suspect that if the handling editor wants to get rid of a paper, he/she will invite so many reviewers with previous rejection records so that he/she can justify his first impression...

Earlier in my career (biological sciences/molecular biology) it was effectively always 3 (80's, 90's, early 2000's) and anonymity was strict. An author could recommend a particular editor due to knowledge in a field, or recommend against a specific editor due to conflict of interest, but the reviewers themselves were generally anonymous. More recently it seems that only 2 reviewers are chosen and authors can recommend people to be reviewers. I was very startled by this development. A recent experience as a reviewer of a manuscript for a reputable journal has left me disillusioned.