How to write a scientific paper

16 minute read


This post is dedicated for students that wish to publicize their own research work by means of a scientific paper.

If this is your case and you want to get some tips on the basis of my personal experience on doing it, read further. I will assume that you are familiar or have already read How to read a scientific paper effectively and How to select a research topic.

The recipe

A paper is generally written following the same “recipe” by scientists, and this recipe has been developed through the centuries. It is not always the same, of course there are exceptions, but bare in mind that the majority of research work published in the engineering sciences follows the same pattern we will devise next.

Very often a manuscript is divided into sections we describe below:

  • Introduction

The introduction ideally should contain the (i) contextualization, the (ii) motivation and justification, the (iii) state of the art critical review, the (iv) contributions, and the (v) manuscript organization. I advise you to divide these topics in at least one paragraph each. You can use more paragraphs, for example if you want to review the state of the art of two topics related to your work which deserve their own attention. It is completely fine to divide in more paragraphs. Remember: each paragraph ideally should contain one main idea to be conveyed. Always start a paragraph knowing which idea it is. In the following we discuss each of the topics (i)-(v) that are generally given in the introduction.

(i) Contextualization: this is the first paragraph of your paper. The reader is just starting to read it, and you have to make sure your paper is engaging so one is encouraged to move forward. Write about the general context in which the problem you are solving is inserted. It is fine to give statistical and general information but remember: you are writing a scientific article, be specific.

(ii) Motivation and justification: describe why anyone reading your article should care about the problem you are trying to solve. Is it a 1 trillion dollar industry? Can you help saving lifes by increasing safety standards? Can you improve the error of a state of the art model by using a new concept from another area? If you don’t know why is it important, it is fine. Take the opportunity to figure it out. Items (i) and (ii) are very important for feeling confident when you will present your research work, as you generally start with them in a talk.

Again, the more compeling you are, the more likely you are to engage your readers (and your peer reviewers too) and make impact in your field by being cited in the future once the paper gets published.

(iii) State of the art critical review: the literature review should cite in a concise manner the most relevant research work published, preferably more recently but not only, about the topics related to your paper. You should cite the papers accordingly so that the literature review gives a big picture of what has been done in the field. Lastly, it is a good idea, if not mandatory, that you write critically about what has been done. You can point out what eventually was overlooked and is still unsolved. Otherwise, if everything is perfect, why should you bother working on this topic? Positioning your review like this puts you in a good position to motivate the need to publish your own work/contributions, see next.

(iv) Contributions: as far as you are concerned when preparing your manuscript, this is the most important part of your introduction section. If you don’t know what to write here, it is ok. Take one step back, read more, re-define your goals, and refine your experiments/results/developments accordingly so you can write convincing contributions. The more contributions you have, the greater are the chances that your paper gets accepted.

Be careful to clearly state the novelties of your work and its improvements compared to the literature. It is better that you think about the contribution from the very beggining of your research work, so you don’t spend time re-doing things. Be sure that the time will come when you will have to state your contributions, be ready to do so. That’s why (iii) is so important. Don’t hesitate to take one step back and refine your literature search and your goals so as to assure your contributions are relevant. It is advisable that you discuss what you plan as contributions beforehand with your advisor and collaborators.

(v) Manuscript organization: it is not mandatory, but in general a good practice, to inform the reader the structure of the document. As the paper has no list of contents, the reader may find what he’s looking for in this paragraph.

  • Methods and development

Insert here everything is needed to reproduce your results. You can also add briefly the theory needed to understand the methods. Include also the description of your contributions. I repeat, make sure that everything you need to implement the methods you used and proposed are stated here. Make references to other papers when you need to be brief, give room for your own developments. Please note that in general the content here described might take more than one section in total.

  • Results

It is not so easy to write a concise results section. Bare in mind that you need to support your conclusions with the results you expose.

So give all the results needed for you to reach to a conclusion and that support your claims with respect to the contributions. Always start this section by enumerating the experiments that you will depict, and then detail each one of them. You have to inform the reader about what you are going to expose before doing so. The reason is that the results you show might be long and intricate to read, depending on the subject, and you want to make sure your paper is easily skimmable.

If you insert a figure, cite and analyse it in the text. Use the captions to make strong statements, which you describe in the text. The same applies to tables.

  • Conclusion and future work

Tell the reader what the goals were and how successfull you were in achieving them. Briefly re-state with different words what you have alluded in the results section.

Good research should always point new directions and possibilites. State them here, and don’t hesitate to tell what you are hypothesizing for your next research work. You might even get contacted by readers interested in collaborating.

Are all the papers organized like that?

Not all the papers are written in this monolithic way. Nonetheless, if you are starting, I suggest you to not to complicate things and use this template. You will benefit in many ways. For example, to mention a few, it will help you to:

  • Organize your ideas

Once you know what you have to write, and where you are going to write, it is much easier to finish your paper because then it is a matter of having the time and patience to fill in the blanks. Another important thing is to be disciplined to always write something everyday, even if it is only a paragraph. There will be times when you will strugle to write a sentence, and others when you will write many pages. You always have to try, otherwise you will miss the prolific days. You will see that it is much more pleasant to write little by little than being obliged to write everything at once, what can be extenuating at times, mainly if you have a close deadline.

  • Avoid that you spend time on unecessary stuff

It is very common that students waste time writing unecessary material for their monographs. For instance, going into lengthy details of a well-known method which is irrelevant for corroborating to the claims you made in the contributions. Remember that the contributions are the most important thing to spend your time on.

Before you start writing, it is important to have the big picture set in your head. Then you can start materializing it by writing the structure of the paper. Otherwise you will get lost during the process, possibly writing lengthy texts which will take you nowhere close to what is expected in a research paper. You will thus loose efficiency and you don’t want to waste your time.

  • Make sure you insert all the information needed for a sound manuscript

If you follow the steps above, even if the text is intricate, you will still be sure that everything that is important for the paper is there. If there will be anything to change, it will be a matter of form, not content. Many papers get rejected because they fail to address any of the points mentioned that are expected to be devised in a manuscript.

  • Comply to requirements made by peer reviewers

Once you get reviews for your submission, eventually you will be critized and will need to make amendments. In general they will be allusive to one of the points mentioned before. If you follow the template, you will know exactly what and where to change in the manuscript, and so you gain efficiency in the process as a whole.

Getting things done

So now you know which is the general structure of a paper. It is time to practice. Start a paper defining the sections which you are going to develop and specify all the paragraphs you are going to write in each section. You might change that in the future if you wish, this is only the first attempt to organize your work. At this point it is important that you begin.

Be careful to balance the length of each section you are going to write. In general, most conference papers have page limits (2, 4, 6 or 8, or even more) and journal papers have word limits (between 4 and 6k). Always check the author guidelines before starting so you don’t waste any time.

For example: let’s say we want to write a 6 pp. paper for a conference. A suggestion could be

  • 1.5 pp. for the title, abstract and introduction;
  • 2 pp. for the methods;
  • 1 p. for the results;
  • 0.5 p. for the conclusion and future works;
  • 1 p. for the references;

what adds up to 6 pp. Note that if you start writing your methods in, say, 3 or 4 pp., you are very likely to need some re-work to summarize it otherwise the paper will be unbalanced. Do plan, so you don’t have to do things twice.

Should I use Word, Open office, Latex, or whatever other tool?

It is up to you. Find out the tool that you are most productive with. I personally prefer Latex when any of the following are true: writing longer texts, with equations, with cross-references (figures, tables, citations). Recently I came across Overleaf, which is a great tool for collaborating online with Latex that requires no installation. You just login, start writting and collaborate effectively. I strongly advise you to use it, if you are into Latex.

A colleague told me that Word now is much better on typing equations, I’m not up to date on this. Mendeley has developed a very interesting citing add-on to Word that makes it very easy to organize your references. I used it and liked very much. Check the following on this matter: add references to Mendely and citing in Word your Mendeley library.

After submission

If you submitted a journal paper, the Chief Editor checks the adherence to the journal scope and, if he finds it interesting, he handles the paper to an Associate Editor which will find the reviewers to evaluate your work.

If you submitted to a conference, the schedule will be known beforehand. Make sure you make an appointment in your calendar so you remember to check the submission system and so avoid any problem with emails that went to spam folders for example.

There is nothing you can do but wait. After submission, it is a good idea to start working on developing your next paper. It may take very long to get your paper published in a respectful journal. This is because it takes 1-4 months to get the first round of reviews (or even more) and then, most of the times, you will need to reply to the comments made by the reviewers, maybe more than once, and each round may take a couple of months more until you know the outcome.

Once you get a decision letter, there are roughly three possibilities. Your work may be accepted and in this case it will be published after small corrections. It may be conditionally accepted after you make a major review, which in this case you will have to reply to the comments made by the reviewers. And you might get your paper rejected.

It is common that your work gets rejected. There is no problem with that, you will fail many times now and in the future. It becomes a problem if you don’t accept it and don’t try to improve on the basis of the decision letter. Take the comments from the reviewers and try to improve your work. Be humble and accept the comments from the reviewers, they are invaluable to you. If you have submmitted your work to a respectful journal, the comments have been made by an expert of the field which has been selected by the Editor and freely given to you. Treasure the decision letter and comply with its recommendations as you progress your knowledge on the topic. Read it more than once so you understand the concerns the reviewers might have. Think of them as someone helping you not to publish something inconsistent, incomplete, or incoherent. You can check in many acknowledgements, which are often put before the references, authors thanking the reviewers saying something like “The reviewers helped to improve the quality of the manuscript”.

It is also possible, if you are going to reply to the comments by the reviewers, that you eventually don’t agree with a recommendation given by them. If that’s the case, first of all, bare in mind that this person you are talking to is taking his/her precious time to review your work voluntarily. Then write to them politely and carefully state why you think otherwise, supporting your statements accordingly.

Why should I care about writting papers?

Because that’s the way scientists report their results and so impact society. If you are a student developing a research topic, very likely you receive funding from the Government. Tax payers should benefit thus from your work and one way to do it, even if indirectly, is the impact that is made by publications. Getting published allows your colleagues around the globe that study the same subject to get to know your ideas and the outcomes of their applications. When you publish you get the chance to influence and help shape your research area. Before you, many scientists did the same, and so you stand upon the shoulder of giants.

Not less significant for you, the research you publish will be important for your career. Once you start applying for grants or projects, one of the most important metrics is the quality of your publications. It is logic as, if you are applying for money to develop a topic, people want to know what is your competence and experience on it. Be thoughtful, then, on taking time to plan for your research output and where to publish your so hard-worked research.

Further reading

Below I give some indications for extending your reading in the topic of scientific writing: