如何评审一篇论文？（How to review a paper?）

前言

本文旨在讲述：如何评审一篇论文？

How to review a paper？

[1] Peer Review at Science Publications. http://www.sciencemag.org/authors/peer-review-science-publications

The submitted manuscript is a privileged communication and must be treated as a confidential document. Please destroy all copies of the manuscript after review. Please do not share the manuscript with any colleagues without the explicit permission of the editor. Reviewers should not make personal or professional use of the data or interpretations before publication without the authors’ specific permission (unless you are writing an editorial or commentary to accompany the article).

感觉没什么重要的信息，除了上面这个，do not spread the manuscript casually before its publication or without the explicit permission of the editor.

附件RAinstr13.pdf 我感觉也没什么内容，这个和第二轮审稿没关系，而且都是一些最基本的内容，比如：Novelty: Indicate in your review if the conclusions are novel or are too similar to work already published. 不过了解这些还是对自己写文章有点用处的。

[2] How to review a paper. http://www.sciencemag.org/careers/2016/09/how-review-paper

Writing a good review requires expertise in the field, an intimate knowledge of research methods, a critical mind, the ability to give fair and constructive feedback, and sensitivity to the feelings of authors on the receiving end.

确实，论文的review，comment要写好，这个必须得有一个很专业的知识，我现在还差的远诶…

What do you consider when deciding whether to accept an invitation to review a paper?
I consider four factors: whether I’m sufficiently knowledgeable about the topic to offer an intelligent assessment, how interesting I find the research topic, whether I’m free of any conflict of interest, and whether I have the time. If the answer to all four questions is yes, then I’ll usually agree to review.
- Chris Chambers, professor of cognitive neuroscience at Cardiff University in the United Kingdom

这个挺有意义的。

Once you’ve agreed to complete a review, how do you approach the paper?
Unless it’s for a journal I know well, the first thing I do is check what format the journal prefers the review to be in. Some journals have structured review criteria; others just ask for general and specific comments. Knowing this in advance helps save time later.

这个也很重要的，你得先看看期刊论文的评审是不是有特定的格式。

I almost never print out papers for review; I prefer to work with the electronic version. I always read the paper sequentially, from start to finish, making comments on the PDF as I go along. I look for specific indicators of research quality, asking myself questions such as: Are the background literature and study rationale clearly articulated? Do the hypotheses follow logically from previous work? Are the methods robust and well controlled? Are the reported analyses appropriate? (I usually pay close attention to the use—and misuse—of frequentist statistics.) Is the presentation of results clear and accessible? To what extent does the Discussion place the findings in a wider context and achieve a balance between interpretation and useful speculation versus tedious waffling?

- Chambers

这个很有道理，原来是在pdf里面做笔记，然后不断问自己问题。这个很酷。

I subconsciously follow a checklist. First, is it well written? That usually becomes apparent by the Methods section. (Then, throughout, if what I am reading is only partly comprehensible, I do not spend a lot of energy trying to make sense of it, but in my review I will relay the ambiguities to the author.) I should also have a good idea of the hypothesis and context within the first few pages, and it matters whether the hypothesis makes sense or is interesting. Then I read the Methods section very carefully. I do not focus so much on the statistics—a quality journal should have professional statistics review for any accepted manuscript—but I consider all the other logistics of study design where it’s easy to hide a fatal flaw. Mostly I am concerned with credibility: Could this methodology have answered their question? Then I look at how convincing the results are and how careful the description is. Sloppiness anywhere makes me worry. The parts of the Discussion I focus on most are context and whether the authors make claims that overreach the data. This is done all the time, to varying degrees. I want statements of fact, not opinion or speculation, backed up by data.

- Michael Callaham, emergency care physician and researcher at the University of California, San Francisco

这位学者前半段对我很有用。还是要有一个checklist，或者多问自己问题。

Most journals don’t have special instructions, so I just read the paper, usually starting with the Abstract, looking at the figures, and then reading the paper in a linear fashion. I read the digital version with an open word processing file, keeping a list of “major items” and “minor items” and making notes as I go. There are a few aspects that I make sure to address, though I cover a lot more ground as well. First, I consider how the question being addressed fits into the current status of our knowledge. Second, I ponder how well the work that was conducted actually addresses the central question posed in the paper. (In my field, authors are under pressure to broadly sell their work, and it’s my job as a reviewer to address the validity of such claims.) Third, I make sure that the design of the methods and analyses are appropriate.

- McGlynn

这个很酷，感觉很专业。

First, I read a printed version to get an overall impression. What is the paper about? How is it structured? I also pay attention to the schemes and figures; if they are well designed and organized, then in most cases the entire paper has also been carefully thought out.

When diving in deeper, first I try to assess whether all the important papers are cited in the references, as that also often correlates with the quality of the manuscript itself. Then, right in the Introduction, you can often recognize whether the authors considered the full context of their topic. After that, I check whether all the experiments and data make sense, paying particular attention to whether the authors carefully designed and performed the experiments and whether they analyzed and interpreted the results in a comprehensible way. It is also very important that the authors guide you through the whole article and explain every table, every figure, and every scheme.

As I go along, I use a highlighter and other pens, so the manuscript is usually colorful after I read it. Besides that, I make notes on an extra sheet.

- Melanie Kim Müller, doctoral candidate in organic chemistry at the Technical University of Kaiserslautern in Germany

都很专业啊，都值得学习！

还有很多，我不一一看了，应该也都差不多吧。
不过区别大概就是每个评委在看的时候侧重点不一样，比如：A侧重 the novelty of the article and its impact on the field ， B侧重spend a fair amount of time looking at the figures。

文章还讲了以下方面：

• How do you go about drafting the review? Do you sign it?
• When, and how, do you decide on your recommendation?
• How long does it take you to review a paper?

I make a decision after drafting my review. I usually sit on the review for a day and then reread it to be sure it is balanced and fair before deciding anything. - Boatman-Reich

太精彩了，确实，应该至少看个三天，每天都拿出来审一遍。不然太没有诚意了吧！ 而且书读百遍本就其义自见也，很酷，我觉得这种方式很适合我。

>In my experience, most papers go through several rounds of revisions before I would recommend them for publication. Generally, if I can see originality and novelty in a manuscript and the study was carried out in a solid way, then I give a recommendation for “revise and resubmit,” highlighting the need for the analysis strategy, for example, to be further developed. However, if the mechanism being tested does not really provide new knowledge, or if the method and study design are of insufficient quality, then my hopes for a manuscript are rather low. The length and content of my reviews generally do not relate to the outcome of my decisions. I usually write rather lengthy reviews at the first round of the revision process, and these tend to get shorter as the manuscript then improves in quality.
- Selenko

噢，原来是先一审，给出very lengthy的意见，然后二审的时候可以短一点。

It usually takes me a few hours. Most of the time is spent closely reading the paper and taking notes. Once I have the notes, writing the review itself generally takes less than an hour.

- Walsh

有道理，做好笔记之后，写review就很简单了。

A few hours. I like to use two sittings, even when I am pretty sure of my conclusions. Waiting another day always seems to improve the review.

- Callaham

Normally, a peer review takes me 1 or 2 days, including reading the supporting information.

- Müller

I almost always do it in one sitting, anything from 1 to 5 hours depending on the length of the paper.

- Chambers

太有道理了！！！即便对自己非常自己，还是要坚持第二天再审一下，这个实在太负责了吧！！！很专业，professional

• What further advice do you have for researchers who are new to the peer-review process?

The paper reviewing process can help you form your own scientific opinion and develop critical thinking skills.

有道理，可以帮助自己develop critical thinking skills，这个是我很缺乏的东西。

Remember that a review is not about whether one likes a certain piece of work, but whether the research is valid and tells us something new. Another common mistake is writing an unfocused review that is lost in the details.

这个很客观，有道理，看论文，不能有偏见，也不能太主观，应该客观评价论文是不是有意义，等等。所以，“少量多次”在论文评审中也是有用的，因为可以让你further improve your review