Rules for writing review articles

September 16, 2013

There should be a rule that you can’t write a review unless you’ve published at least three original research papers in that topic/area of focus.

Also a rule that your total number of review articles cannot surpass your original research articles.

57 Responses to “Rules for writing review articles”

  1. eeke Says:

    I disagree. Anyone should be able to write a review on any subject and should not have to be a member of some “club” to publish, as you suggest. If it gets peer-reviewed by members of the field and they like it, why shouldn’t it be published? If anything, if it is well done, it would be an unbiased synopsis of what is current in the field. Also, I don’t think it’s uncommon for grad students to be first- or co-author on review articles – sometimes they are asked to co-author these articles along with the senior author, and it forces them to familiarize themselves with relevant literature. I don’t have a problem with that.


  2. dr24hours Says:

    Ah, the inner elitist comes out, DM.


  3. drugmonkey Says:

    Why is that “elitist”? If anything, research articles are prole work and reviews are from those who see themselves as something above such nitty-gritty.


  4. dr24hours Says:

    For exactly the reasons described by eeke. “If you haven’t contributed X to the field, you aren’t worthy to review it.” That’s an elitist attitude. If a person can write a decent review, then there’s no reason not to publish it.


  5. drugmonkey Says:

    Dude this is about review articles, not the people writing them.


  6. dr24hours Says:

    No it’s not. At least not as you wrote the post. The post is clearly about the person, not the article. It’s about who you think should be allowed to write review articles.


  7. drugmonkey Says:

    Oh and if it helps, there are at least three specific situations on the top of my list and in all cases the participants are most assuredly in the “club”. Two are newish/smallish topic domains. One related to a specific individual who is in the club due to nearly annual reviews on the same old topic with little new to say.


  8. drugmonkey Says:

    eeke- did you read my last post on citation of review articles? That’s why.


  9. dr24hours Says:

    This strikes me as a topic of almost vanishing import.


  10. eeke Says:

    DM – yes. Your other post is irrelevant here. Anyone should be able to write a review article. If it’s universally informative (for readers both within and outside the field), it should be published. Even if your dog wrote it. Your pit bull. It’s not the reviewer’s fault if other people lack scholarship and don’t appropriately cite original research articles. If anything, a good reviewer might highlight research papers that for some reason (such as not published by Glamourmag) have been overlooked or under-cited.


  11. kevin. Says:

    SNAP, eeke.


  12. qaz Says:

    Are you writing this as qaz-bait? This is total BS. A paper is a paper is a paper. It doesn’t matter who wrote it. That’s the beauty of science. This is the difference between science and pseudo-science. Pseudo-science says that only some people have the right to be right, that there are sacred books and sacred people, and only those priests have the right to tell you the truth. Science says that if you can provide a contribution to our understanding, then it doesn’t matter who you are – only that you are right, and can prove it. It says “prove it”, but that the only thing that matters is the text (and the data behind the text).

    I will agree to your stipulation on one condition. Let’s define “review” articles as those articles that only review and do not provide any new insight into a field. (By that definition, I know very few review articles in existence. If you are going to provide nothing more than a list of experimental articles, then you shouldn’t be wasting anyone’s time writing a “review” article, anyway.)

    Theory articles, articles that contribute new perspectives on integrating data, articles that show that two experiments that are thought to be incompatible are actually different perspectives on the same thing (think blind men and an elephant), articles that show that if an experimental hypothesis is true then some other experimental hypothesis must follow, none of these should be counted as review articles, because they contain contributions in their own right. Of course, as you well know, all of these are counted as “review” articles by elitist experimental neuroscientists who do not seem to appreciate the importance or complexity of theory or theoretical contributions.


  13. Dr. Noncoding Arenay Says:

    Agree with Eeke. A good analyst can write a good review paper. The analyst need not have published original articles in the said field to be able to do so. Also, +100 to the “unbiased synopsis” aspect of such an undertaking.


  14. drugmonkey Says:

    Let’s define “review” articles as those articles that only review and do not provide any new insight into a field.

    Sure. Many of them do not and serve mostly as repositories for collecting topically related literature in one References list. As per the last post on this topic, the reviews are then cited for that collection value. Not for any unique theoretical advance or insight.

    One way to go about reducing the lazy citation (and robbing of the original authors) is to reduce the number of these dang reviews.


  15. Grumble Says:

    “Anyone should be able to write a review article”

    Just like anyone should be able to send 65,000 tweets.


  16. DJMH Says:

    DM, I agree! Also let us stipulate that only scientists who are willing to reveal their real names should be allowed to have science blogs, because CREDENTIALS.


  17. DJMH Says:

    P.S.. one of my favorite reviews ever was written by a grad student with no prior publications–it covers a broad territory instead of being so narrow-minded the way a lot of PI-driven reviews are, where it’s all about jamming in as many self-cites as possible.


  18. Mike Says:

    Agree with all the commenters saying this post is elitist nonsense. The message seems to be “Don’t release your Greatest Hits until you have at least three albums”. But a review is not some sort of victory lap, or summary of one’s own work. And writing a useful review is a totally different skill from organizing and presenting one’s own research. There is absolutely no reason to think that someone who runs a major lab tries to write a review, he will come up with the good “synthesis” type of review instead of the bad “just a list of stuff” type of review.

    And of course, the review has a good chance of being written by a trainee anyway.


  19. Walter Says:

    “Also let us stipulate that only scientists who are willing to reveal their real names should be allowed to have science blogs, because CREDENTIALS.”

    Excellent point. The anonymity of science bloggers and twitter handles who claim some expertise and hide behind a screen name to complain about colleagues and the system is quite annoying.


  20. drugmonkey Says:

    This strikes me as a topic of almost vanishing import.

    Why does someone inevitably say this only *after* making several comments on the topic? Wouldn’t you be more believable if you simply ignored it altogether?

    the way a lot of PI-driven reviews are, where it’s all about jamming in as many self-cites as possible.

    tch, tch, DJMH. Hard to hold your seat on your charging high horse when making such scurrilous and uncited charges.


  21. dr24hours Says:

    Maybe I just like to argue about unimportant things, ever think of that?


  22. drugmonkey Says:

    one of my favorite reviews ever was written by a grad student with no prior publications–it covers a broad territory

    Oh, and we can make an exception for a grad student publishing excellent and extensive scholarship that goes into a traditional (real, non3paperstaple) dissertation. I agree with you there.


  23. drugmonkey Says:

    Maybe I just like to argue about unimportant things, ever think of that?

    The thought never crossed my mind.


  24. dr24hours Says:

    Given your twitter shenanigans, I’m unsurprised. You seem addled. 😉


  25. Eli Rabett Says:

    Review articles are not telephone books, but evaluative documents (at least the good ones) that tell you who to believe and who to laugh at. That requires being ass deep into the area with a good sense of the players and the clowns.

    DM is right on this one.


  26. DJMH Says:

    Yeah you have pretty much exploded your own point there, DM. And I never said all reviews were driven by self promotion, but no one can deny that some are. I have also read good reviews from people outside a subfield and starting to get into it, again nice works because they synthesize from the outsider’s perspective rather than that of someone with an agenda to push.


  27. qaz Says:

    See, DM, the reason this drives me crazy is that the much bigger problem in neuroscience is that experimentalists don’t respect things they call reviews (anything that’s not an experiment) because they all think good reviews (theory, integration, etc) are easy (and unimportant). I sent one of the first papers I wrote to an experimentalist colleague who said “you don’t have the right to write this paper.” (That’s a direct quote.). Luckily, I ignored him. That paper turned out to be very important to a redefinition of the field, and dozens of experiments have confirmed that our surprising reinterpretation was right. We need to figure out how to differentiate good and important contributions from crap. Saying only experimentalists can write reviews doesn’t help.


  28. The point of writing a review article is to teach people about a field. Just like how the best researchers are not necessarily the best teachers (or vice versa), I don’t see the point of limiting review authorship to active researchers in a field, nor do I really have a problem with the sort of person who seems to write more reviews than original work — if they are good at it, what’s the harm?


  29. Alex Says:

    It seems to me that there are two types of good review articles: The ones that are useful for a n00b, by clarifying things that are known but not always well-articulated, and the ones that provide a thoughtful and provocative synthesis and try to push the reader to consider some ways of thinking (and even hypotheses?) that haven’t really been explored much as of yet. The first is relatively safe and useful, but if you try to write it and fail then you’re just cluttering the literature. The second is going to face a harder time from reviewers, but might be a very valuable service to the community if you can get it out there.

    I confess that my reviews have been more of the first category, though I do try to highlight issues that others mostly neglect.


  30. @qaz
    As a computational biologist myself I’m certainly annoyed with how (some) experimentalists view non-bench work as somehow trivial or not really science (although I wonder if it isn’t really insecurity on their own part for realizing their limits in mathematics and computational data analysis), but that’s really a separate issue, no?


  31. Ola Says:

    Granted, there are the reviews that actually take 2-3 years to write, with 300 references, beautiful diagrams, lots of synthesized ideas/paradigms. I’ve done a couple of these and they were absolute hell, came in over budget (too long, too many co-authors, late in time), but very worthwhile and have stood up in the long run (lots of citations). But, there are many exceptions to your proposed “rule”, resulting in less than stellar reviews:

    As a mid career PI, you might get 3-10 review invitations per year, and at any given time there are 1-2 students or trainees in the lab who have recently prepared a review type document – thesis, thesis proposal, fellowship application, whatever. What to do with all that lovely prose? My lab does at least one of these reviews a year, and the people who write them have little experimental experience or publications in the given field.

    Then there’s the classic “oh shit we’re about to send in a grant on topic x, but have no publications on it, so quick, write something, anything,” Yep, done a few of those.

    There are also reviews in the category of “we’re submitting some kinda collaborative grant thing, but we have no actual pub’s together, so we’d better co-author something, quick”. You see a lot of this with program project proposals.

    And don’t forget the “holy crap this person will burn down my lab if I dont find a way for them to stay occupied and out of trouble!” situation. A timely review invitation saved my ass in just such an occasion. People can do a lot less damage chained to a laptop than in a lab (YMMV).

    Or there’s the classic “shit shit shit my resident is hitting on my undergrad assistant, better keep them separate for a while” scenario. Reviews are great for “removing someone from the gene pool” for a while.

    Or the play for the wannabe climber – “hey, you really want to portray independence from me, write a review article as a single author on a peripheral topic”.

    So yeah, people write reviews for a whole bunch of reasons that have nothing to do with actually addressing a need in a given sub-field. A lot of reviews out there are nothing more than training exercises in scientific writing, or political plays. Are we filling the literature with clutter? Maybe, but the editors are incessant with their requests, because reviews drive up impact factor. I would guess that the above practices are routine for many PIs, and so a lot of the reviews you see out there are complete cack.

    If only there was a way to filter out the good ones?


  32. drugmonkey Says:

    Yeah, ignore all of them Ola.


  33. drugmonkey Says:

    qaz- I don’t think good reviews are “easy”. I think me-too, categorize-the-lit reviews are stealing the credit from the original authors. Whether they are easy or hard to construct makes no never mind.


  34. drugmonkey Says:

    JB- the harm is to the citation rates of the people who did the original work.


  35. Aisling Says:

    DM: I generally agree that there should be no requirements for being a review author. Your point of giving credit where credit is due by citing original work instead of /in addition to a review where relevant is a much better way of addressing the “citation drain” issue.

    On a side note, thanks Ola for your tales of HR-101 from the bench – that was a great comment 🙂


  36. qaz Says:

    I think we’re all in agreement about the problem, just not the solution.

    So rather than saying people should not write reviews, lets (1) remove all citation length requirements and (2) make people cite original literature. Those solve the problem at hand (citing reviews instead of original literature) without interfering with (as JB points out) other issues.

    As a practical suggestion, I suggest that we commit as editors and reviewers referees to demanding that authors cite original literature and always say that “Due to the importance of these citations, journal citation limits should be relaxed.” (Of course, what we mean is “Due to the fact that they are really stupid, journal citation limits should be relaxed.”)


  37. The Other Dave Says:

    Isn’t blogging about science the same as writing reviews? What about PBS Science documentaries? Do the producers of Nova need to stop summarizing science until they’ve contributed to the experimental literature? Science journalists? Textbook writers?

    Who is allowed to communicate science?


  38. drugmonkey Says:

    No. Blogging is not like writing reviews. At all.


  39. The Other Dave Says:

    I am not so against citation limits. I think some people have a terrible habit of OVER citing. I think some people reference everything that shows up in their PubMed search. Which is not helpful at all, and can be quite misleading. Citation limits make people think carefully about what they’re claiming to have read, and where they want to point readers for more information. Citations are supposed to be an aid to readers, right? They’re not Facebook ‘likes’.


  40. drugmonkey Says:

    Good idea, qaz. I agree.


  41. The Other Dave Says:

    DM: Blogging is not like writing reviews? Aside from the target audience, how are they different?


  42. drugmonkey Says:

    I cannot remember the last time I saw a paper or manuscript that “overcited” the literature. This includes many journals without citation limits. I conclude from this that authors generally know how to do it right. Maybe your field is full of glad handing suckups?


  43. drugmonkey Says:


    It is rare that a blog entry covers more than 1-3 papers.


  44. Evelyn Says:

    As a grant writer, who has to learn a whole new field every 2 weeks, a good review is a life-saver! It gives me an accurate, hopefully an up-to-date snapshot of the field, leads me to the original research that I can then pull, read, and cite. A bad review is a an awful waste of my time but at this point, I can tell a bad one by about third paragraph and I don’t bother reading the rest of it. When I can’t find a good review, my life gets a lot harder since I don’t have the time to read all of the junk research on every topic before I get to the good stuff. At that point, I hate to say it, but I search through glam-journals and usually, those original papers will have enough background to lead me to the important papers in the field.
    So I don’t care if the review authors are in the field or are first-year grand students, as long as they do a good job. But in my experience, the ones that are in the field usually give a better overview of the topic.


  45. Tal Says:

    Oh, and we can make an exception for a grad student publishing excellent and extensive scholarship that goes into a traditional (real, non3paperstaple) dissertation. I agree with you there.

    Right, so now we have our new rules about who can write review articles: no n00bs! Oh, except whenever we have valid exceptions to our rules–like a really exceptional n00b who’s clearly part of the elite. Then it’s okay to break our rules. Oh, except now we need experts to determine when it’s okay to break our rules and when it’s not okay to break our rules. So it’s a bit more complicated. But still, I like this idea. I propose we create some kind of dynamic review board, made up of maybe 2 or 3 invited experts in the field, who get to review each review paper and determine whether it’s good enough to publish, regardless of whether the person who wrote it is a n00b grad student with an excellent and extensive dissertation, or a friend of DM’s with an extensive track record in an area. Maybe we could call this process “peer review”.


  46. drugmonkey Says:

    Oooo, Evelyn. Really?


  47. Pinko Punko Says:

    DM- I think what you mean is that if one is a writer of excess or superfluous review articles these review articles should not be published- the rule shouldn’t be restriction on the author side, it should be restrictions on the publishing side. Of course since reviews are citation bait, there is a race to the bottom by journals to publish superfluous or competing reviews. Funny, journals don’t want to publish papers that are redundant, but they’re happy to publish redudant reviews.

    The post here I understand the types of people that DM wants to restrict but it comes out as restricting one way for younger scientists to have an avenue for visibility.

    The restriction should come from if one is invited to write a review, to make certain there is some extra effort to make the review worthwhile- discussing and adding ideas or synthesizing- not just listing citations or results unquestioningly or uncritically. Be part of the solution not the problem.


  48. Lee Says:

    I agree with DJMH. Also, in my subfield, reviews are often a way to promote the author’s hypothesis for a given phenomenon. I have a feeling DM’s reviews would be about the scourge of pit bull denalistas.


  49. drugmonkey Says:

    It is my dream to someday work the dangers of pitbulls into one of my papers, yes.


  50. The Other Dave Says:

    How about a rule that someone has to publish a review for every half dozen mindless ‘we can measure that too!’ experimentalist papers?


  51. Grumble Says:

    No, TOD, we don’t want to hear from them. In fact, those are the kinds of scientists who end up with enough time on their hands to spam PubMed with not just “me too!” research articles, but reviews that say the same thing repetitively, redundantly, and over and over again, focusing mostly on their own “work,” such as it is.


  52. dsks Says:

    TOD is entirely off his rocker with the blog talk. I’ve been tempted to just post a review on the topic area of my work to a website under my name and simply update the bloody thing as and when new significant advances require it.

    A mechanism along those lines by which editors could invite folk to update an organic and ongoing review for a given topic would certainly cut down a lot of the redundancy that people object to.


  53. Evelyn Says:

    I know – I am Captain Obvious.


  54. drugmonkey Says:


    It is not that.

    I am just considering the whole notion that review articles make it easy for a professional grant writer, who has no prior knowledge of a field, to simulate the expertise of the PI. And considering the notion that study section members presume the PI did most of the writing.

    Which leads to the notion that some PIs can afford to hire a professional grant writer to, presumably, submit more grants to out-compete those of us who cannot afford such luxuries.

    None of these are happy considerations.


  55. Evelyn Says:

    I work for a department – although I do know of some PI’s who have their own grant writers.


  56. The Other dave Says:

    TOD is entirely off his rocker with the blog talk. I’ve been tempted to just post a review on the topic area of my work to a website under my name and simply update the bloody thing as and when new significant advances require it.”

    I DID that. It was well received. No one reads those weird journals that spam your email box with ‘opportunities’ to publish in some Indian or Chinese online journal anyway.


  57. anonymous postdoc Says:

    U mad, bro?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: