I personally look for novelty. I love seeing people who are great designers of experiments. Lastly, I look for breadth in their work whether it be from Physiology to Pharmacology to Biochemistry.

Since I am not a reviewer, I was thought this would be a great forum for some feedback so as to improve upon.

To all that answer, it is much appreciated.

I appreciate your response. It is ever so helpful. Dr. Gandhi, do you care to share any certain qualities that you specially look for in a manuscript when reviewing?

I've only reviewed once. I also took about one week. The prose was problematic and led me to a misquoted reference. I spent most of the time on materials, methods, looking for controls, and checking that discussion actually matched data results. I also checked with a statistician about the methods used for a small sample number.

I think that for anyone who cannot write in scientific prose, it would pay to have the manuscript edited. The manuscript that I reviewed would have been below journal standards if published as submitted.

Great feedback Dr. Eshel. I especially liked when you thought the data set was too small you went the extra mile to make sure it was accurate and still finished in a week.

I was hoping this question would provide a unique perspective and insight to into peer review process. More on the level to understand the different philosophies and focal points reviewers use as guidelines when reviewing manuscripts.

As already mentioned by Dr. Gandhi, reviewing a paper can be very subjective. For my point of view "novelty" is the first important quality, next the "Presentation". I devote most of the time on "Result & Discussion" and next the "Materials & Methods" part. Generally, I try to complete a review within 7-15 days.

"Novelty" is the first quality I observe, followied by results and discussion. After this, I read carefully meterial and methods and presentation. I usually can revise using 2-4 days. I can expend more time, in order to check similar published papers.

Thank you everyone for you answers. It is much appreciated. If you know of any other reviewers of journals if would be so kind to share this question with them if they have a Reasearch Gate account. Really great and helpful information thus far. I would like to get a little more breadth to have a really nice sample size.

It usually takes me about a week (less to accept it more to reject it). I usually look for novelity, sound science (the data presented is robust and the author's conclusions are sound). What I most hate is when authors do not give due credit to others, misquotation of references (in some cases making it clear that they have not read the reference they quote) and when the conclusions are not supported (in some cases even contradict) the data presented in the manuscript

In addition to what Jose said, I would add that, given a reasonable quality of exposition, I would 'forgive' minor mistakes IF the idea is absolutely novel and sound (a rare case). As I mentioned elsewhere, today we should value particularly highly truly novel, especially radical, ideas.

The concept, originality, quality of data, hypothesis testing, referencing etc. most probably one week is sufficient. I think "Review delayed is Review Denied"

In my fields, the best practice journals provide a template for the abstract, reducing problems there. The best practice journals also provide a rubric (template) for the referee to ease, speed up and standardise the analysis of the paper. A summary of the paper takes the most thought and time. Because refereeing papers is unpaid (but expected) work, the turn-around time (which I believe is the metric being estimated above) increases with the academic load at that time. A fairer metric would be "Time At The Screen" (TATS). My TATS vary from 15 minutes for a perfect paper (and the same for a terrible paper written in incomprehensible English!) to eight sessions averaging an hour each = 8 hours total TATS. This happens when there is a lot of cross-checking or interesting referencing to follow. Papers that are plagiarised may take up to 16 hours (TATS) for me to document fully.

I did a review for a paper 8 times. Took me 1 to 7 days.

What a learned is: In addition to their final evaluation, Reviewers need a "Questionnaire template": answers to give one by one, point by point to questions that would go deep into papers (hypothesis, Methodology, results, conclusions)

In my opion the Review Process is a Science per se -or should go towards it!- and should not (only) be an opinion...

@Ian: I like the TATS(:-)

Reviewing is also an art. You need to be able to break away from the mold and take the initiative. Fairly soon into the paper, I can see which way the paper is going, and I then gather evidence to justify the final decision. It is too easy to get carried away and want to redo or extend the research or writing, when all that is required is an informed decision.

@Alessandro: Was that 7 days of 8 hours continuous work = 56 hours total TATS?

I submit reviews in within a month. I try to make it within 2 weeks. As a writer a 3 month wait seems fairly common especially for the more "prestigious" journals. I have taken only 4-5 hours reviewing really well written articles. I have rejected some poor ones in a similar time frame. The ones that take a lot of time are those that use good data and have something useful to say but are not currently publishable. Perhaps the author did an incomplete review of the literature and is missing key information that helps with the interpretation of the data. I have had reviewers comments that improved my papers. I feel an obligation to extend that service.

Reviewing can be sped up if the reviewer has a soft copy. This lets the reviewer quickly determine key words, search for all occurrences of terms or troublesome words, quickly go to URLs cited or cut and paste some text to search for plagiarism.

Being a speed reader also helps.

Ses my answer on the attachment

I've only reviewed three papers yet for three different journals. I also took about four days to one week. I really see the concept, originality, quality of data, hypothesis testing, referencing etc.of the manuscript, mostly as per questionnaire of the journal provided.

I ususally spend more than two weeks for each article. I appreciate some Journals that make you earn CME for each article, because this is a real hard work from the reviewer's standpoint. Indeed, I believe that Editors should try to acknowledge the peer reviewers someway because of the high quality of the work, that is time-cosuming

Some Editors acknowledge their reviewers with a twenty percent reduction of the price for a futur publication in THE journal within the year....

Reviewing must remain free , to avoid the risk of potential conflict of interest, but should be taken into consideration by universities and our institutions. For our group, reviewing may largely impact our usual activities, and consume some evenings orweek-end days.

Two weeks for each article, I am not sure that we discuss about the same concern. At evidence, when we have read the manuscript, we may have some questions, we think about it , before sending our comments to the Editor and even may discuss, but two weeks is hard to believe, if you are in charge of an intensive care department , Futhremore manuscript central or equivalent calculate the delay to respond ( two weeks and no more), just after your internet acceptation. An hypothesis to explain this discrepency ?

It happens very often to me to review manuscripts, for this reason I must be hurry (maximum 1 hour) I know is not so much but if an article is good you are immediately catched by the problem and by the kind of idea behind, then it is possible there are some imperfect passages but it is a minor problem, otherwise if it is a 'me-tto' manuscript (they are by far the majority) you can understand it very soon.

I think the important thing is for the reviewer to keep in in mind authors put a lot of work, patience, passion and, even if their work is not good for you, you must in any case be usefult tot he authors well explaininh how and why they could ameliorate..

It takes from a day to two and some times more in case of papers with good work but poor presentation. Newer aspect or info. regarding the reported work and its present-ability is what I look for the most.

As retired former editor of an international journal I admire reviewers who only needed a few hours for a fair reviewing. When I got papers from other journals for reviewing I had often problems with sufficient free time to perform this important task to the needed high standard. Thus, I stopped reviewing due to time limitations. The time for writing grant applications and reports was more needed to support a larger working group demanding sufficient time for advice. In this context I wonder what my fellow researchers think about reviewing (evaluating) international research proposals of normally > 100 pages and > 10 international partners and costs of sometimes > 10 million $ within only half a day including a check of the quality of participants and a written report with a quality mark. Are those demands by a funding body leading to a fair evaluation of those trans-national proposals? Up to now it seems that I am the only one who complaints about the lack of time for a high quality evaluation.

I perfectly agree with the fact that reviewing for The Editor of an International Journal is very very time consuming and thus may induce difficulties in the management of its group. In my specaility, Critical Care Medecine, the Editor of a review may change every two ot hree days, whereas others are in charge of an Editorial Board for a decade or more.

I beliveve that the very large discrepency between the reviewers answers is due to the difficulties to evaluate our "thinking time"; When, we accept a review, we look at the abstract and then reed for a first time the manuscript. Then you think about it ,make an update on the topic, and finally evaluate the manuscript for many hours. How do you calculate your time of thinking?

For instance, last two weeks i reviewed a manuscript, unfortunately not suitable for publication in this hight impact factor journal. I consume one hour for the first lecture, think of this manuscript during four days,( time not evaluable), spent nearly six hours for my first draft, and proposed to days after to reject this study during a seven hours saturday work In this case i think that i spent at least 14 hours for the all process. 14 hours taken on my private time. "Thinking time" may be considered by some reviewers, and thus they may declare that i spent at nearly a week!!

We need two metrics: Time at the screen ( TATS ) and meditation time ( MT ). In obvious cases, the meditation time can be zero, but this is the exception.

The main points to ponder should be hypothesis, novelty (basically difference from prior art or presence of new approach), organisation of the contents, the study protocols (avoid unnecessary explanations and repetitions) and proper discussion (not only support from prior art). The review process should not be completed in a "single go", this approach is used by "critics" not "researchers". The first reading should be to understand and in the form of a learner not a learned one. Next reading should be evaluation. Each section should be read separably, say, 1 h a day. The review will not be a burden and will be completed in a week or less. However, initial prior art search is mandatory and a little time consuming.

Dear Sir, you enquired about something, which is a universal question and in the mind of all, but not able to question. Thanks a lot for asking this one.

I have got some inspiration to do quick reviewing. I will certainly overcome my lazy attitude in reviewing.

Thanks & Regards,

Rajeev K Singla

Any research project worth 6 month's effort is worth devoting 1 month's time to write he report. However, scientific reviewing can strengthen scientific thinking, because it can be personally challenging and professionally rewarding. The time and effort are worthwhile. Good scientific reveiwing is as much the result of the referee's knowledge and expertise with work. Dividing the manuscript to the problem statement (general problem, literature review/significance of the study hypothesis), methodology (design, population and sample, data collection techniques, statistical hypothesis and tests), results, conclusion and recommendations (discussion section), data obtained, analysis, including statistical tests, conclusions, interpretations and limitations. Thus, overall, referees are generally recognized to be expert in the subject matter of the paper under consideration (3-5 days are usually enough for a referee to serve as a mechanism for reviewing and criticizing the methodology and conclusions of a paper).

I review more than one manuscript per week. It takes me up to a morning (3.5-4h) unless it is too complicated or too obscure. Low quality manuscripts may take far less and are easy to be picked up.

Important thing is the relevance of research in the field. Is it just a modification of previous research or actually a novel finding. If it is a novel finding, see the materials and methods and how the results are interpreted statistically. What is the claim and how far it is relevant. In case of a modified study, just look at the way it is going to add to the existing knowledge. If it makes a meaningful contribution, its acceptable. Another thing which annoys a reviewer is poor grammer and long sentences. This is my personal opinion. Might be different form others, but I have a look at this aspect also. Also check the compliance to author instructions and referencing in the text and its corresponding appearance in the references section.

Usually review gets completed in a weeks time. I prefer to give a decision promptly if I find the matter interesting and worth publishable.

It takes usually few days (3-5). However, I tend to read the manuscript first one time. I leave it for one day and think about it. I if I would think to accept the paper for publication, I read it one more time and start writing down my notes (major). For the minor comments, I have to go over the manuscript one more time.

You have to take into consideration to write a summary of the paper in your report before adding your comments.

good luck

I do agree with Hilal Zaid that you have to think twice in accepting the article, as I have said that 3 - 5 days is usually enough to evaluate the manuscript and search for some references and other articles related to the manuscript. One major point that the referee has to write a comment or summary on the mansucript before dealing with major and minor comments.

The important thing in reviewing research manuscript will be originality, authenticity, presentation and finally, manuscript usefulness for its concerned field. I reviewed the manuscript in one week.

Very interesting topic William: thank you!

More than " how many time we spend reviewing a paper" it would be interesting and instructive to know " What are your criteria to review a paper?"

One review criterion that has not been mentioned so far is the appropriateness of the manuscript for the journal to which it has been submitted. This covers area, depth of project, type(s) of approach, and level of interest to broader or narrower audience. I have to echo one of the earliest answers, the emphasis on writing quality. If the quality of the writing is poor, it is very difficult to review the content of a manuscript. This can increase the time taken for review, as well as decrease my enthusiasm for a manuscript. The quality of the figures also is VERY important. Overall, if the manuscript is in an area with which I am very familiar, then a day or so is sufficient. If the manuscript covers an area for which I need to increase my background, then a review could take several days. I also agree with those who mentioned that they need time to let the review percolate- I have found that this time can really change my viewpoint, or simply allow ideas that could be of use to the authors to rise to the surface.

Usually it takes couple days (up to a week) if I have a time. The problem is in that how familiar is the reviewer with the topic of manuscript, how professional is he (she). I have declained the manuscript if I am not so familiar. Therefore time and quality are the problem of the reviewer and, in addition, of Editors. Sometimes I don't understand why a long time (1-2 months) is required to review some manuscripts. The best journals pay attention to the quality and to time also.

Worst case scenario, to check some detail in the paper, the reviewer had to order a copy of a referenced thesis from an overseas university. But, just as it comes through the post, all the lecturer's examination papers come through for marking! It is wrong to assume that the *unpaid* referee can always operate at the drop of a hat.

In reviewing a paper, novelity of the research should be key. Logical organization of the ideas is another factor. clarity of prose is an important consideration as well as conclusions from the data-should be objective. A week or two should be enough;depending on how busy one is!! sometimes i can take a day or two on a familiar manuscript

Time: depends on the familiarity with the topic. The MS has to be written so that is is easily readable. As a reviewer your job is not to pick about a few missing commas, that is up to the editorial staff, do not waste your time on that. Your job is to judge if the experiments are sound and the correct conclusions were drawn that are supported by the data. The topic has to fit with the journal and the research has to be novel.

BUT.... it is not your paper, remember that! If you would do things slightly differently ... great write your own paper.

Thank you for the excellent discussion here about reviewing manuscripts. Although I have not yet reviewed a manuscript, I have experience with reviewers, usually they take more than a week or even months, for reviewing. Most of the time there are time constraints for a very busy scientist for reviewing an article.However, it is good to admit the inability to review the manuscripts before hand, also it is good if the journal itself has provision if the authors suggest their reviewer themselves for expedited manuscript processing. As the credibility of the author's hard-work and the scientific quality of the journal is both dependent on the reviewers comments. Reviewer's job is the most challenging one for a quality scientific journal.

CME credit for reviewing manuscripts is on the order of 3-6 hours of CME category 1 credits (eg, for JAMA it's up to 3 hours and for Annals of Internal Medicine it is up to 6 hours), which gives a sense of how much actual time (in hours) is needed for a typical review. The actual turn-around time depends on how quickly one can get to it, and varies greatly.

I recommend organizing the review in decreasing order of priority:

Section 1: Overview of your impression: Does the manuscript have "fatal flaws" -- issues of validity or study design that cannot be reasonably addressed by a revision? How important is the article to move the field forward? Is it the right journal for this type of work? Is it a really important contribution to the literature? Is the writing style excellent or painful to read?

Section 2: Major comments: What would the authors need to change to improve the article globally (better framing of the issue, better statistical methods, better description of the methods to allow for other investigators to repeat the study)?

Section 3: Minor comments / typos: Just jot them down as you go... helpful to authors but least important.

Also: from the Editor standpoint, do NOT sugar-coat your reviews. There's nothing more frustrating as an editor than ending up with scathing commentary in the confidential comments to the editor section and nice comments in the comments to the authors section; that forces the editor to be the bad guy and to paraphrase the scathing comments (if the editor agrees). Better to be honest to the editor and author together. After all, with most journals you are reviewing anonymously. Confidential comments to the editor should be made just to communicate with the editors things that you might otherwise put in an email (eg, "I could not follow the authors' statistical methods fully; if other reviewers feel similarly, you may want to get a statistical reviewer's input.")

As a companion topic to this nice discussion on manuscript review.

What are the tipping points to your final comment, Accept, Accept with revisions, and Reject?

I have rejected a few articles with fatal flaws, and one that was written so poorly I would have spent hours just revising so it is legible. However, if the reasoning is sound and it is not just a rehash of, or lower design of what's already around (retrospective analysis of hypothesis already answered by multiple RCT), I tend to accept. Am i too easy? I certainly have had manuscripts rejected that I would have accepted.

Also, D Brotman mentioned CME for reviewing an article?

Reviewing manuscript is always hard work to be fair. Thus, concentrate on objective, methods, statistics procedures, conclusion and if the results justified the conclusion or not. I mean sometimes, you have to be qualified e.g., in statistics to do the job correctly. Thus, the time for the reviwing the manuscript depends on the paper type. Usually simple manuscripts without too much statistics, needs less time.

Unfortunately, the reviewer has to go deeper than just ticking off the presence or absence of five qualities. If it were that simple, refereeing could be over in minutes.

The time it takes to review a manuscript is also dependent on what type of manuscript you are reviewing...e.g. research, letter to the editor, or review article. I am probably a little on the "easy" side because I am as likely to "accept" a research manuscript with negative findings, as I am one with positive findings.....assuming that the findings of the research need to be known by the scientific community. The actual review process can take a while, depending on what else I'm having to do at the time. I try to make sure that I am consistent with my comments to the editor and my comments to the author - but I think that you can always be constructive in both cases - scathing reviews do little to help anyone - If there are problems, state them in both cases - but there's no need to be insulting or to insult the intelligence or competence of anyone. You can get your point across without cutting someone off at the throat.