How to Write a Scientific Paper

Your research is completed by publishing a paper in a scientific journal.

The paper should be written in Scientific Writing (SW), which is neither Japanese nor English.

SW is a written expression for reporting the results of scientific research and should be thought of as a language rather than a writing technique. Following SW, your writing will not only be understandable to other researchers but will also be more likely to be accepted for publication.

First, read the textbook on SW recommended below by Naota Hanasaki.

Lewis, N. Whitby, and E. Whitby: How to Write Scientific Papers in English for Scientists and Engineers - Secrets of Writing Internationally Accepted Papers, Tokyo Kagaku Doujin, pp219, 2004

Tetsuya Kono: Introduction to Writing Reports and Papers, Keio University Press, pp116, 2012

★Things to do & know before you start writing your paper!

• Search/collect all relevant literature
• Know the journal to which you are submitting, the readership, and the areas of the field
• Your first draft does not have to be perfect

★Where do I start writing?

• Make a list of all figures and tables and create an outline (separate blog post)
• Organize those figures and tables in presentation order (results and discussion)
• Do not write the introduction first...
• Start with the methods, e.g. experimental methods
• Next, write the results and discussion
• In the case of model studies, the results are often abstract, and it is easier to understand if you also include a discussion
• However, be sure to separate the results and the interpretation paragraphs.
• The standard for the amount of text (not including figures and tables) is: Introduction:Methods:Results+Discussion=1:2:4

★Tips for Writing Your Paper

• The first sentence of each paragraph should be the gist of the paragraph.
• Therefore, you should be able to follow the outline of the research and the logical structure of the paper just by reading the first sentence of each paragraph.

★Abstract Tips

• An abstract is a "summary," but as the name implies, it is easier to organize well if you "extract" specific information from the text.
• Extract the following key information from the paper according to the number of words specified.
  • The Objective section of the Introduction
    • 【Important】 Background, Earlier works, and Shortcomings of earlier works in the Introduction are not mentioned in the Abstract, but one sentence for each is included.
  • The most important point of the Method.
  • Most important points in Results and Discussion (include numerical results).
  • The Answers section of Conclusions.
• In the case of a model study, it is relatively easy to summarize the results in the following way.
  • To clarify the effect of XXX on XXX, we conducted a simulation study on XXX [Objective: 1 sentence].
  • The study was conducted under the condition of ____ [description of model: 1 sentence].
  • The simulation was conducted under the condition of ____ [description of simulation: 1 sentence].
  • For ____, it was found that ____ under the condition of ____ → In contrast, it was found that ____ under the condition of ____ → Other results were found to be ____ for ____ [Explanation of results: 3~4 sentences].
  • From these results, we found that the effect of ____ on ____ is ____.
• Only information that can be easily and quickly spoken - Never copy and paste sentences - Practice many times so that the Abstract can be read in a flowing manner.

★Introduction Tips

• Construct with four paragraphs: Background, Previous Research, Issues of Previous Research, and Objectives.
• Background: Direct the reader's attention to the subject of your research by:
  • Importance of the very broad research topic
    • Global warming is a threat to global development.
  • Importance of the broad research theme (only address the issue addressed in this paper).
    • Global warming alters precipitation across the globe.
  • Importance of the specific research theme that the paper addresses.
    • This is expected to change the most basic water resources of humankind and river flow throughout the world.
  • Earlier works: Summarize the findings of the research topic as identified by previous studies.
    • A summary of important related studies in order from oldest to newest.
  • A shortcoming of earlier works: The purpose here is to summarize the key points of the issues that we are now going to deal with in the previous studies.
    • Point out logically the problems of previous research while paying respect to the achievements of those who came before us.
    • Avoiding negative critical expressions, as it is often the authors of the previous studies who review the papers.
  • Objective
    • Objectives of the study: Indicate what the study aims to solve among the problems in the previous studies.
    • Methods of the research: Show why the problem can be solved where previous studies have failed.
      • This is a very important part of the introduction, so show the originality of your research straightforwardly.
    • Research question: Define the question to be answered through the research.
      • This is the most important part of the introduction - a thesis is a statement that shows the answer to the question you set yourself - it is not a thesis unless you set the question!
    • Structure of the paper: the subsequent chapters show the structure of the paper, but it is often not necessary if it is IMRAD structured!

★Method Tips

• Method's great example is a cooking recipe. The part that shows how we have moved the frontier forward and how the reader can reproduce it - anyone can make the same dish by following the recipe.
• In the case of a study in which simulations were conducted while the model (computer simulation) was developed and refined, it is easier to explain if it is organized in the following four sections.
  • Model section: Overview of the model.
  • Data section: input data to the model and possibly validation data.
  • Simulation method section: simulation conditions and settings & a list of simulations performed, etc.
  • Analysis Method section: If statistical processing is performed on the simulation results, it should be described as well.

★Results and Discussion Tips

• Make a clear distinction between Results and Discussion.
• Describe the frontiers of your study objectively (Results) and provide an interpretation (Discussion).
• In the case of modeling studies, the results are often abstract (they depend entirely on the assumptions of the model and the simulation setup), so it is easier to include the interpretation as well.
• However, if results and interpretations are mixed up, it will be difficult to tell where the results are obtained according to the Methods and where they are the authors' interpretations.
• Results should be objective and straightforward, describing the results obtained according to the simulation and analysis methods.
  • Example: "Variable X is largest in simulation A and smallest in simulation C."
  • Like in a recipe example, the result is a picture of the finished dish.
• The discussion states the interpretation of the results - Going one step further, write to convince the reader that the results you have shown are trustworthy.
• What to write varies widely depending on the results presented, but probably the minimum to include is "why the results were obtained (=mechanism)" and "the validity and consistency of the results.
• Mechanisms should describe how the model, conditions, and settings lead to the results.
• In the case of modeling studies, the results often reveal weaknesses or shortcomings of the model, in which a brief explanation of why such weaknesses are apparent should be provided.
• Validity and consistency describe how the results and mechanisms fit with theory and facts.
• If there are shortcomings in the model, show that they do not affect the conclusions of the study and are outweighed by the strengths of the model.
• Discussion is a recipe example (although a bit off the subject) A sentence such as "XXX and XXX go great together" next to a picture of a dish - shows the mechanism of why it tastes good and the interpretation of the deliciousness of the dish.

★Tips for Conclusions

• Conclusions, not summaries or closing.
• Describe how important your research is now and what it may lead to in the future.
• Conclusions should be based on a three-paragraph structure: Answers, Significance, and Implications.
  • In the Answers section, provide answers to the "research questions" listed in the Introduction.
    • 【Note】: This paragraph should be written as an answer to the questions, not a summary of the paper.
  • In the Significance section, write how Answers has contributed to the development of the research frontier.
    • 【Note】: This paragraph cannot be written well without a thorough review of previous research in the Introduction... Write the contribution focusing on Answers, not on the paper or research contribution. If you write the contribution of the paper or research, it will be a repetition of the Introduction!
  • In the Implication, you should write about your expectations and wishful thinking (if you don't have any, don't write about them).
    • For example, "This research will lead to the solution of XX problem" or "Future research issues are..." If you have something that has neither been implemented nor realized, but you want to make a point, write it in this paragraph.
    • 【Note】: Write only about the direct implication of the method used or the results obtained - For example, if you use simulation and obtain the result that the number of years of drought recurrence is halved in the future due to global warming, write about the implication of the simulation method or the result of halving the number of years of drought recurrence. Many people write, "Therefore, it is important to prevent global warming," but the connection to the method and results is too small to make a convincing argument.

★Tips for References

• Reference is a mirror of the author's attitude toward research.
• The part where you present a list of references you have cited.
• All research builds on prior research, so don't neglect it.
• There should be no misvalues or deviations from the specified format.
• Use literature data management software such as EndNote to create complete References.
• Download bibliographic information from the article's website.
• Never enter author names, titles, etc. manually to avoid typographical errors.
• Download Endnote Styles and other formats whenever possible.
• Reviewers read References carefully, so missing references, mistakes, or formatting errors can make them feel like they are being asked to rework an incomplete paper that has not been well-researched.
• Picture the reviewers reviewing your paper, and spend enough time on the References.

★Read the manuscript steadily, translate and revise

• Read it out loud - you can take a slightly more objective view of your writing.
• Reading aloud allows you to make simple grammatical mistakes and find alternative expressions for parts of your writing that you find difficult.
• Tips for writing in English: For important sections such as the Title or Abstract, or for sections that you are particularly struggling with, translate every word directly into your native language.
• If the meaning or logic doesn't make sense in your native language, first correct the translation in your native language, then re-translate it and correct the English text.
• However, it is not recommended to write the entire paper in your native language from the beginning and then translate it into English at the end.
• In particular, the way of constructing logic in Japanese is completely different from that in English, so extensive polishing is required, which is inefficient.
• Standard Japanese and English differ in the way they arrange sentences.

★Submit your manuscript to the journal that uses the most literature.

• First, identify several candidates.
• The journal with the most relevant previous research published should be selected as a candidate.
• If your research theme is relatively new and you cannot narrow down your choices to a single journal, you can choose the journal that has published the "classics" (the most important papers that served as the starting point of your research theme) as your candidate.
• Next, check the basic information of the journal -> If the paper is written in English, check whether the journal is registered in the Science Citation Index (SCI), Science Citation Index Expanded (SCIE) (whether the journal has a so-called Impact Factor IF or not).
• Journals with IF have more subscribing institutions and are easier to search, so they have more chances to be read, but they also tend to be more rigorously peer-reviewed.
• Next, find out as much as possible about the journal's main readership.
• Finally, after careful consultation with your co-authors, choose the journal that will make you the happiest when your paper is published.

★Written responses to peer review should be addressed to the Editor-in-Chief, not to the reviewers.

• The reader of the response letter: The first person to read the response letter is the editor of the paper (Editor or Associate Editor).
• They will decide whether to send the manuscript back to peer reviewers or to accept it for publication.
• How the response letter is read: Editors are busy - The response letter should be written in such a way that the editor can understand the communication even if he/she only reads the response letter without reading the revised manuscript.
• Advanced Techniques: The response letter is more powerful and interesting when it includes a clear statement of the author's intention and intent for the research, rather than simply responding to the reviewer's comments.
• For positive comments, the response should further address the merits of the research.
• Critical comments should be responded to with a rebuttal and an emphasis on the advantages that outweigh the disadvantages.

★Co-authors are the destiny of the research - Include those who have helped you in the acknowledgments.

• A paper is essentially an irrevocable statement for the future.
• You can withdraw the paper, but it will remain on record, including the fact that you withdrew it.
• The custom in some fields in Japan is to say "I co-authored the paper because I was indebted to you for your research," but if you are indebted to someone, you should say so in the acknowledgments.
• The first author sends a manuscript to co-authors one month before paper submission - time to exchange comments.
• Allow at least two weeks to revise in response to comments.
• Co-authors should revise the manuscript thoroughly multiple times - failure to do so is imposing this process on reviewers and readers, which is embarrassing.
• Only submit manuscripts that all authors believe will be "published without revision" or "published with minor revisions.
• The very best research groups always submit near-perfect manuscripts.

★How to make a diagram

• Figures should be skillfully produced! If the figures in your manuscript contain breathtaking quality, your paper will not be rejected!
• Always look at the figures in Nature and Science papers as they are showcases of the highest level - catch technology trends and get inspired!
• Use MS Excel as little as possible - Does it look a little old-fashioned, made up quickly with a handful of software?
• Better not to use the software's default extreme color palette.
• Better not to create diagrams with Microsoft Office's default color palette (as of 2020) or GrADS' default rainbow colors.
• Colorblind people may not be able to distinguish rainbow colors well, so Color Brewer should be used - it looks more sophisticated (as of 2020).

LOC: ...Public/★FINALIZED★_Blog