科学网经常能看到关于写文章、投稿、审稿、以及与这些有关的事情的讨论。我还只是个博士后，没有特别丰富的研究经验，但觉得在此分享我积累的一些经验是个不错的主意。您若对某些观点有异议，希望批评指正。由于英语是当前科技写作的国际通用语言，我下面的内容用英语书写。
Tips for scientific writing and related issues
1 The elements of a scientific paper
1.1 The Title
The Title should be informative, specific, and concise (but not too concise, as it should be informative).
1.2 The Abstract
The Abstract should contain all the critical information of the paper. It should answer (explicitly or implicitly) the following questions:
--What is the motivation or the broad context of the paper?
--What are the objectives (goals)?
--How do we achieve the objectives?
--What are the findings and (optionally) what do the findings mean?
The Abstract should be concise. Therefore, the abstract should provide maximum information with minimum words. We can organize the Abstract in the following way:
--One or two sentences: State the broad context and/or the motivation.
--One or two sentences: State the objective(s) and the method(s).
--One to several sentences: State the major findings and (optionally) their significance.
Why it is optional to state the significance of our findings? Because a good work will speak for itself.
1.3 The Introduction
The Introduction is a very important part of a paper. A natural way to structure the Introduction is to go from the more general aspects to the specific research problem(s) of the paper. So, the Introduction can be structured as follows:
--One paragraph: Settle down the big context of the paper.
--One paragraph: Go from the large context to the specific field that the paper concerns.
--One to two paragraphs: Motivate to the subject of the paper and briefly state the major findings.
-- One paragraph: (optional) State that how the rest of the paper is organized.
It's a good idea to finalize the Introduction after finishing all the other parts, and check if all the acronyms have been properly defined.
1.4 The main part of the paper
Different journals have different requirements for this part, but there are some common elements:
1) About the Equations.
Make the equations clean and precise. Be consistent with the use of mathematical symbols. Do not simply
copy equations from other papers; derive/check the equations carefully. There are many errors in the
literature. Do not spread the existing errors.
2) About the Figures.
Figures are usually the best way to present numerical results. However, do not present too many
figures; be selective. Make the legend and figure caption clear and self-contained.
3) About the Tables.
Tables are used for presenting data that are not suitable for figures, such as the data
whose exact values are important.
4) About the data/results
Check the data/results over and over again and make sure that they are reproducible.
Everyone could make mistakes, but we should admit/correct our mistakes as soon as we discover them.
1.5 The Conclusions
The Conclusions should be brief. One to three small paragraphs are enough. A typical Conclusions section can be organized in the following way:
--Summarize the major results. Only summarize; do not motivate again. The motivation has already been presented in the Introduction. One paragraph.
-- (Optional) State the significance of the findings, but avoid exaggeration in describing the findings.
One paragraph.
--(Optional) Point out some possible future directions. It would be great if we can also point out some shortcomings of our work. This might motivate other people to further study the topic. One paragraph.
1.6 The Acknowledgement
Firstly thank the people who helped with the research, and then acknowledge the funding sources.
[We can also thank the anonymous referee if that person discovered major flaws, or made significant suggestions to improve the scientific quality of the paper. I have done this in one of my early papers, where I wrote "We also thank the referees of EPJB who pointed out a mistake of our original manuscript and helped improve this paper a lot.".]
1.7 The Appendix
Not every paper should have an Appendix, but the Appendix is a good place for lengthy/tedious mathematical derivations and/or additional figures/tables.
We should regard the Appendix as a separate piece of work (like any paper we cite), to which we refer properly. Nowadays, it is also popular to have a Supplementary Material (or other similar names), but I really don't like it.
1.8 The References
--Cite the most important and relevant papers. Do not cite a paper just because we think the cited author is a potential referee. Do not cite our own or our friends' papers if they are irrelevant.
--Only give references to papers that we have really read.
--Cite the original paper, not secondary ones. For example, if A developed a method and B used it, and you learned the method from B, then you should cite A, or cite both A and B, but not just cite B.
--There are many improper citations in the literature. We should not follow other’s mistake. We should follow our own judgment.
2 Language aspects of scientific writing
There are many good suggestions for the language aspects of scientific writing. I only collect a few below:
--Be clear and concise, not ambiguous and convoluted. Write short and simple sentences; avoid long and complicated ones.
--Be accurate and objective.
--Avoid pretentious language. Use common words and expressions.
--Avoid using too many acronyms, because using too many acronyms can heavily reduce the readability of the paper.
--It's good to only have one main idea in one pragraph; long paragraphs are difficult to read.
--The connection between one paragraph and the next should be obvious and natural.
--Keep a balance between passive and active voices.
--Use of “we” is valuable to distinguish between our new contributions and existing results.
3 Publishing, referencing and other ethical aspects
3.1 Why do scientists write and publish?
--Why we write? We write to be read. So we should write with the readers in mind.
--Why we publish? We publish our papers to report new scientific results and get the credit. We do not publish just because we need a job, a promotion, or a grant.
3.2 The process of writing a scientific paper
--Start to draft as early as possible.
--Once we are satisfied with our draft, it's a good idea to show it to others, at least to our collaborators.
Modify the draft in the light of their comments and suggestions and repeat the process several times.
--When all the co-authors are very satisfied with the manuscript, it is ready to submit.
--It is very important that all the co-authors must read the final manuscript carefully before submitting.
3.3 About the authorship
Discuss with all the potential co-authors about the following aspects of the authorship:
--Which names should be in the author list?
--What should be the order of the authors?
--Who is the corresponding author?
Important:
Do not add non-contributing authors. Paticularly, never include a person as a co-author without her awareness.
[I have had a very unpleasant experience. I have had a collaboration with someone who had only shown me the manuscript once and hadn't notified me of the submitting/resubmitting. My name was mistakenly written as Zhezhong Fan (should be Zheyong Fan) in the published paper! Actually, I do not dare to include that paper in my publication list!]
3.4 About plagiarization
--Do not plagiarize. That is, do not copy others. We should cite, not copy.
--Do not self-plagiarize. That is, do not copy ourselves. We should cite ourselves, not duplicate.
3.5 How to reply to referees' reports?
--First, send the reports to all the co-authors and ask for their opinions.
--Then, read the reports very carefully for a few times. If the reports are not very positive, be calm, not angry.
--If there is a need to do more calculations/experiments, do them as soon as possible.
--Reply point-by-point to the referees' comments politely, carefully, and clearly.
--Finish a draft of the response letter and show it to all the co-authors and revised it many times.
--Resubmit.
--If we are not lucky to get the paper accepted, do not be too sad. We can revise it further and try some other journal.
3.6 How to review papers?
Referencing is important, perhaps as important as doing research. Refereeing other scientists’ papers is not a waste of time. So, accept to referee if we are able to report and review it carefully and timely. However,
--pass a manuscript under review to others (e.g., our students)
--contact the authors of the paper
--accept to review a manuscript if we are not able to do it (the correct way is to recommend someone else to the editor)
转自立文悉通微信公众号，仅作学习交流，如有侵权，请联系本站删除！
