In my opinion, the number of revisions of a good academic paper depends on a number of factors. Some of the them are:

1-The language used in the preparation of the paper and the mother tongue of the author.

2- The content, length, complexity and the clarity of the paper.

3- The professional capacity, experience and knowledge of the reviewers.

4- The professionalism of the editor.

5- The publishing rules of the journal.

6- The experience, seriousness and knowledge of the author on the subject of the paper.

7- The time available by the reviewers and the author to carry out the revision process.

Dear Reginald,

very often you need at least one revision, mostly two or even three revisions are required. So you have to wait a long time. Perhaps this is the reason that so many new journals appear and researchers are getting nervous if they have to wait such a long time, especially when they want to finish Ph.D. and need the good journal publications. There is no easy solution to that problem, but it is helpful when you select a good adviser who is well-known in the community, has "good connections and knows the journal business well".

Good luck,

Reiner

I think question was how many times we need to revise before submitting. that depends on everyone own writing skills and how serious someone would like to improve his or her paper. I can say its between 10-15 or more times for end product.

I do agree with Dr. Reiner's comment on [a good adviser who is well-known in the community, has "good connections and knows the journal business well".]. However, in academia, only those who publishes in top tier journal are well known to the editors. and those who have those qualities, they guide for improving the paper and it become publishable. Its not that much connections but it may work sometimes for low category journals

I have colleagues who had to revise their article 40 times before it was finally accepted in Long Range Planning. So my answer is as many times as is necessary to reach the top level of academic writing.

When does revision '1' stops and revision '2' starts? How to define intervals between two revisions?

with skilled and experienced reviewers, in my opinion two revisions must be sufficient

Dear Reginald

I've normally found 2-3 revisions to be sufficient for a well written paper. But those reviews take a long duration. If there are too many revisions, like Mohamed Benmerikhi has mentioned, it could be that the author did not understand the changes required by the reviewer.

Sometimes the period between revisions can be too long. Those of us who may have completed their Ph. D. may find a lot of time lost in the anxious wait and the data we have collected may become outdated. I strongly believe that the time taken between submission and acceptance needs to come down to encourage more people to publish.

Dr. Neetu, I think you and MB are not talking about the same thing. 40 revisions may take place before submissions to a top tier journal. I am not sure which top journals someone can reasonably submit with 2 or 3 revisions, no mattter how much experience someone have. even top authors need several cycles of revisions....

But what is a revision and its scale from a practical point of view? People write in bouts also involving rewriting across bouts [(e.g. Writing for two hours (write bout 1), then stop (pause = inter-bout) and restart writing (write bout 2), etc...)]. How to define a revision when a revision can be defined at different scales of analysis (e.g. number of rewritings of a single sentence or a single paragraph or a single heading...)?

usually it depends upon the seriousness of the author, that how much s/he is serious about his/her publication.

as much the author would be serious s/he will deliberately follow the guidelines of that particular journal.

more over most of the time it requires two to three revision to finally reach at the require standard by thatjournal

In my opinion, the number of revisions of a good academic paper depends on a number of factors. Some of the them are:

1-The language used in the preparation of the paper and the mother tongue of the author.

2- The content, length, complexity and the clarity of the paper.

3- The professional capacity, experience and knowledge of the reviewers.

4- The professionalism of the editor.

5- The publishing rules of the journal.

6- The experience, seriousness and knowledge of the author on the subject of the paper.

7- The time available by the reviewers and the author to carry out the revision process.

Yes, all points raised relevant. But what about the end product - often sometimes turns out to be a 'different' paper from the initial someone intended? This revision, apart from editing, is a tricky business?

Well said Dr. Imbun. In my experience, after revision/modification/changes/correction for about 18 times among 4 authors, we realized the original wording has changed more 80% BUT meaning or objectives of the paper topic become more clear/concise/and fluent to read and visualize the topic what we were describing in the paper. Ofcourse it can differ person to person but revision by peers is must before submission

The only answer I could think is: it depends (on discipline, journal, etc). Second, keep in mind that readership (and the "later existence" of the article in the mind of readers) is quite different than peer-review readership, so no one could predict in advance whether was it was judged as top article in the phase of production will be the judged to be the same by a much wider audience (and circle of readers).

The answer to how many revision it takes to produce a top article depends on which discipline is the paper on, the number of original authors, the native language of the authors (as papers in english by non-natives will probably need more reviews than papers from natives, etc). I would say that in sociology to revisions would be ok in most cases, three if the paper is authored by more than 1 person.

Normally 2-3 revisions and finally get accepted.

I cannot add anything serious to the excellent comments above. So instead, here is a simple equation:

number of revisions = [(10 x number of authors) + (2 x number of submissions)] / cumulative decades of experience

Professor Chatburn,

I like your formula. :-) And, I agree, although diverse, the comments have been very informative.

I would agree with Jorge Morales Pedraza .

the more a writer is acquainted with the publisher's house style for the journal, and the skill of the peer review panel, the more easier one is able to publish a journal article.

Very good equation Chatburn...

Just a thought--I recently published with Frontier's open review process and found the ability to interact directly with reviewers very helpful. If you can I would give the open review process a try, I think it would give you a better sense how much revising you personally need to get the paper to be publishable.

It is difficult to give a number. It depends on the capability of the author and reviewer as finds @Shafig Ibrahim Al-Haddad and also the breadth of the paper as well.

A million dollar question and billion dollar answers.

It all depends upon the type of work done and type of journal it is being submitted for publishing. It will also depend upon,whether the reviewers are able to understand in the first instance, what authors want to convey. Overall, no thumb rules here, it varies from case to case.

You do not want to have so many revisions that you are forced to invite the referees to become co-authors!

Things you can do

Check the scope of the journal ( reviewers will point this out if your article does not fit, even at a later stage in revisions), read instructions for authors, often papers are sent out to reviewers but the format may not have been checked before doing this. Poor English is often the reason for authors undertaking many revisions, editing takes quite a lot of time - on a very small number of occaisions 7 revisions have been needed. Sometimes the authors do not adequately respond to reviewers and so they have to be sent back.

Initially it can feel disheartening revising an article on a number of occasions but in my opinion the article is usually much improved by the process and you need to avoid feeling that it is an attack on you personally. It is worth passing round your article to colleagues within your own institution for comment before you send it to the publisher as this will help eradicating spelling and grammatical errors and also sentences where meaning is not as clear as it could be. It will also get you used to the review process. Making sure you've observed the format and focus of your chosen journal is imperative if you are to avoid unnecessary rejections of your written work. Once you've got your reviewers comments make sure you respond to every point. You don't have to agree with every suggested alteration but if you are rejecting a suggestion you need to make clear your reasoning for this.

Ian, do you (or anybody else in this post) heard of an actual enforced co-authorship? I have, indeed.

I created a topic about it with more details, which you can find here:

https://www.researchgate.net/post/How_often_does_enforced_co-authorship_occur

If you are aiming at top tier journals, many of which have rejection rates of 95% plus, then my experience has been that the best you can hope for realistically is major revisions and probably 2-3 rounds of revisions. It has happened that I've had something rejected at the 3rd round and I have had another paper go through after (if memory serves) 6 rounds of reviews, everyone with experience has their 'horror stories' about reviews. No way should you withdraw an article with major revisions though - assuming that is that you sent it to the appropriate journal in the first place. That is the result you are hoping for if you are aiming high. Your point about delays in the system is well made, but although it's true at the level of one paper, publishing is not 'job production' if I can put it that way, it's 'flow production' ideally you need some under review, some being revised, some drafts, and some on the drawing board. To improve your prospects, try to work with other people - ideally who are more experienced than you :)

I'm not sure if "how to deal with rejection" is exactly on topic, but I'll just put it out there in case it's useful (forget how I found this article, apologies if someone already mentioned it in this thread):

http://www.insidehighered.com/advice/2009/04/27/belcher#sthash.4dvDWeNH.dpbs

These days I am taking more to heart the idea mentioned in this article of not revising a submission until it's been outright rejected three times. I think revising is hard for most of us and can interrupt the momentum of the process. In my opinion, choosing an appropriate journal to submit to next can actually be a better use of my time than revising. Once I know for sure that a journal is interested that's when I really spend time on quality revisions.

Rachel,

What you suggest has been my experience. Thank you.

Colleagues,

All of you who have posted an answer, thank you too, because they have been very informative and helpful.

I agree with Jorge on the fact that plethora of factors determine the number of times to revise articles before it is finally accepted. Two most important déterminants however are the quality of the work itself, and secondly, the pedigree of the publishing agents.

Does anybody know of official statistics? Couldn't find them myself.

Depends on what you mean by revision.  For my first article, we literally did a couple of hundred rounds--this was because my mentor wanted to really teach me how to write.  I often do this with students--not a couple of hundred but many.

Once submitted and asked for a revision it usually takes only one or two.  Rarely three.

If some of the new interactive review processes such as Frontiers, the definition of revision become blurred

.

I agree with dear dr. Jorge that it depends on such factors that he mentioned above, 

I agree with Dr. Jorge, But some time the number of reviewers is not indicator to a quality paper its about the reviewers experience. 

What if more parameters/analytes are requested and there is impossibility of specimen re-use, what is the best way to approach that?

Reiner Creutzburg Marcel M. Lambrechts Domenico Montanari

Dear Kamoru,

Perhaps give a concrete example? You Always can add new data in an existing data file, or not? The alternative is to start a new study on the same model system taking the referee remarks into account?

It is often the case that a high quality paper undergoes two consecutive revisions: A major revision and a minor revision. This is the way prestigious journals operate and the papers which are subjected to such modifications usually enjoy originality.

Dear Dr Reginald L. Bell , For the top academic articles, most important things is from your research design, hypothesis, sample size , Results.....later it may go either one or two revisions.

You have obviously also revisions across different publications to finally get a top academic article using the former publications as background information?

To address the question of: " How many revisions does it take to produce a top academic article?", I think the number can not be predefined, as it depends on many factors. If the scholar makes a rich and meticulous basis for the article with painstaking preliminary revisions, the further modifications may not need more time. Furthermore, the first rejection from a journal can be a good guide for improving the paper either resubmitting to the same journal or submitting as a new paper to other ones. Any way, it requires time, patience, and interest.

Dear Reginald, I suggest that assume a paper as a puzzle that you need after drafting to scramble and re scramble its parts for several times in order to finalize it as a perfect original job. So, number is not important.

Good Luck

From zero to as many as needed. Usually, beyond 3 revisions, the editor start having a rethink about a paper. Consideration is usually given beyond 3 revision/reviews if the idea is novel and worth sharing.

My first article had more than 200. You read right. My mentor was trying to teach me to write. He succeeded. Now probably less than 10

Many experts believe seven revisions should be done by the author before submitting to a journal.

As many times as the researcher is ready to comply with the instructions of the publishers and be honest (original) in his research. Number of times is irrelevant but the ability of the researcher to know that he must obey the instructions of the publishers (ethics), guides to publications, declaration of originality etc.

I always consider a request for major revisions a win. You are told what you need to do to improve the article. Just listen and do it, get a good proof reader if you need to, and in my experience, 99% of the time you will get an acceptance. The only time I did not was when the resubmission took too long - new editors or reviewers can have a different opinion or may change their opinion of the work over time, so prioritize your revisions. And, do not take critiques personally. They generally improve the final product.

My Confirmation document reached 28

This is a good question, which could be ammended: what happens in the review process? It may be that a reviewer in the first round only had one major concern, but in the second discovers a lot more. This can become a dilemma, because other reviewer could be satisfied. So, addressing additional concerns by reviewer A may make the originally satisfied reviewer B less happy. In brief, I am tempted to say that at any one journal, more than two rounds of reviews do not seem to make sense, because there is no way that some people will not feel thoroughly frustrated. If a manuscript is good, it should entertain editors and reviewers for a limited amount of time. Authors can be assured that good manuscripts will find a way through, but perhaps not at the journal they thought it would fit the best. By the way, I agree 100% to what Marian said above.