How to Write a Research Paper

A few years back I wrote about how to write a research proposal and how to read a research paper. I wrote about these, respectively, because I was busy in writing research proposals and my students were approaching me with questions about how to read research papers. This time around, quite a few of my students are busy in writing research papers. And quite often than not, I am contacted by someone to give advice on as to how to write a research paper. So, often I have to repeat the same details again and again to different people. To save myself from the tedium of repetition, I am writing this article for prospective research writers. The idea is, of course, to educate them to develop a decent writing expertise.

So, if you are adamant about writing a research paper, and you shall not abandon your ambition at any cost, please read on. I hope that this article will make the process of writing easy and fun for you. For your convenience, I have divided this article into small sections. Please read the section that best suits your needs.

Working With Latex

First of all, I would like to emphasize that it is always best to typeset your article in Latex. If you are not sure about what Latex is, please read this article and set it up on your machine. In short, if you are a windows user, install Miktex on your machine and a nice Latex editor such as LED. If you are on Linux, please install TeX Live on your machine. A nice editor for Linux is Texmaker.

Once this installation is done, you are ready to typeset in Latex. All you need now is a good style file to typeset your document. A style file dictates the appearance and formatting of your document. It makes the document appear the way you or your audience want it to appear.

However, you don’t have to worry about that. You can get all sorts of style files from the web. Since you want to write a technical research paper, as you are reading this article, you definitely need a style file that can contain technical content. For technical or scientific articles the best thing is to get one of the following style files.

Sometimes, it can be better to have a specific conference or journal in mind when you decide to write a research paper. Almost all the good journals and conferences have links to their preferred versions of style files or templates. Here are examples of templates for GECCO, Sigcomm, EuroGP, Infocom and IEEE Transactions on Multimedia. As for most of the things in this world, the beauty of a research paper also lies in the eyes of the beholder. And in the eyes of the editorial staff of the most of the world’s top ranking conferences and journals, only those articles are beautiful that are typeset using the templates provided by them. And indeed, the articles look beautiful to others as well. Moreover, the best thing about using a template provided by someone else is that you are relieved from the onerous burden of submitting a good looking article. The only thing you have to care about after that is to fill the template with the right kind of content that your reviewers cannot refuse to accept. Rest of the article is just about these ideas. I am going to try to keep it brief and succinct.

Before you read on, I shall try to draw your attention to a nice research paper. Giving it an eyeball will be helpful for you in many ways. For one thing, you will get a detailed overview of what kind of content goes in a technical research paper. You will get a clear idea about the kind of figures, graphs, charts, results, and analyses go in a good research paper. So here you go. You can definitely download it and use it for future reference.

On Losses, Pauses, Jumps, and the Wideband E-Model – IEEE Journals & Magazine

There is an increasing interest in upgrading the E-Model, a parametric tool for speech quality estimation, to the wideband and super-wideband contexts. The

I am writing the following advice from the point of view of writing about empirical research. However, my ideas would be useful even if you belong to a totally theoretical feat.

Remember that laboratory work, fieldwork and experimentation are at the heart of empirical research. Talking about them does not fall under the scope of this article. If you are wondering how to design and conduct your experiments, please go on and read a different article on how to do research.

Normally you start writing when you have done all the experiments you want to write about. However, you can defy this golden rule and start writing while you are actually doing your experiments, and even before you have started your experiments. So you can have your experiments in mind, that great idea that you wanted to work on, and even before you actually start working on it, you can start writing about it. Actually, in my point of view, it would indeed be really smart if you start writing about it before you have worked on it. The reason as that as you conceive an idea, your mind generates certain peripheral thoughts about it. Writing down your thoughts is nice as by the passage of time, you will simply forget what you were actually thinking about. I don’t remember what I was thinking about six months ago. However, I do remember I had some really exciting ideas in my mind. I archived some of them here on this blog. And I simply forgot some really good ideas for sure. So why not start writing your ideas about your research even before you have started working on it? And why not start writing them in the form of a research paper?

Typical Structure of a Research Paper

Think about what goes in a research paper. A typical research paper has a number of sections. These are basically the building blocks of a research paper. To name but a few, a research paper has the following sections: abstract, introduction, previous research (literature review), methodology, results, conclusion.

What I do to make my life easier is to create a separate file for each of these sections. So for abstract, I make an abstract.tex. For the introductory section, I make an introduction.tex, and so on. Doing this makes the process of writing easier and manageable. Please remember that it will be really nice for you to name your files with one or two meaningful words. Similarly, You should also name all the relevant folders with one or two words. This will be helpful in a painless migration of the article on various computing platforms. There should be no blank spaces between the words of a file’s name. So prefer file-name.tex, FileName.tex or fileName.tex over file name.tex (notice the undesirable space between file and name).

Include all of these tex files in your main latex file. You can use the \input command of latex in the main body of your main latex file. Something like \input{abstract.tex}, \input{introduction.tex} etc. would do. Compile the main file and you will see on your screen a pdf version of your empty article ready in a few seconds.

So what do you do now when you have made all these tex files? Well you have to write content for them. Figures and tables would be required at some stage too. But don’t worry about them right now. They will sneak in as your experiments proceed. Just think about writing the content now.

Writing the Content

Normally I start writing these sections in any order. I can start the abstract before the introduction and conclusion even before I have gathered results. Order does not matter to me. Later on, when the whole article is complete, I adapt my previous writings to match the truths, facts and bitter realities. This works for me.

But how do I write all of this? Or how should you write all of this? First of all, your writing should be inspired by your reading. if you are at a stage of performing your laboratory experiments, you have probably done a lot of literature review. In a literature review, you come across the works of many scientists or engineers in your field. You must definitely have read a lot of research articles thoroughly. Did you not notice their writing styles? How they write abstracts, and how they begin the introductory paragraph by bringing to the attention of the reader the gravity of the problem they are addressing? Every writer has his/her own peculiar style. And it is worth noticing the peculiarities in the style of an author. There would definitely be top notch researchers in your field whose writeups stir up your blood. Such can be the impact of writing if you are a passionate researcher. The idea now is to leverage these styles and techniques.

What I used to do a long time ago before starting writing a research paper was to leave everything and go out for a long walk. I would then think about the section that I had in my mind. Initially, you can be perplexed about what you should write. The thing you really need is a starting point. Where to begin writing from. And honestly, it could be from anywhere. Once you start writing, words follow. In one sitting, you can write a whole lot of content for a section.

What I do for kicking myself off is to think about how other people wrote about their work. I can read and re-read their sentences. This provides me a way to think about scientific thinking. And as you walk you can think about a lot of ideas.

I remember that once I dedicated ten days for writing a research paper. I divided my time and dedicated the slots to the bits and pieces of sections that I had to write. Giving a certain target to yourself to write can help a lot. I remember that I set a target of 500 words per day to write for myself. This helped me a lot. And believe me, 500 words per day is not a big deal. If you develop a decent expertise in writing, you can write 500 words in 15 minutes. Yes, 15 minutes, I reiterate! So like this, you can write 5,000 words in 10 days. And since you began writing much earlier, you may as well dedicate months to your writing.

Try to Have a Quiet Mind

We tend to panic a lot about many things. We presume in our minds that certain things are difficult to do. And we create a big hurdle in our minds about accomplishing our goals. No matter how big your goals are, they might as well not be so much difficult to achieve. So just try to relax. I require solitude as I indulge in a serious write-up. This is absolutely necessary and quite helpful. As you go in solitude and engage seriously with writing, ideas tend to flow through your mind. It is also nice to have a quiet and a calm mind as you write. Sitting back and relaxing can help you achieve that. Moreover, the quietness of the mind also naturally follows in a calm and a quiet environment. You can think more rationally and objectively in a good environment.

Then as I write, my favorite writers literally start speaking to as I write. This is a literally a crucial point that I would like to share and I am not joking or bragging about it. There are a bunch of writers I like too much. George Friedman is one of them. He is the CEO of Stratfor and writes about geopolitics. I love his writing style. He reflects on geopolitics in a bleak, objective and yet a thrilling manner. I find his articles more thrilling than many Hollywood action blockbusters. He is also very authoritative about his opinion and argument and supports it with logic and reason. Malcolm Gladwell is another writer that I like. I once had a chance to read his quite popular book blink. I remember that as I started reading it, I couldn’t put it down until I finished it. My eyes slid through the lines really fast. Honestly, as opposed to struggling to read it, each and every word in the book pulled and steered my eyes through it. And it was as if the author was speaking to me through the sentences in a really fast and a highly engaging way. Sam Harris is another writer that I like. Although I do not agree with some of his ideas, I really like his writing style among other things. And there are loads of other writers that I like and quite a few of them are technical writers. It is nice to be inspired by a spectrum of writing personalities.

So as I sit down to write, where ever I have to pause or I get stuck for a lack of an idea or a thought to write, an idea crosses my mind that I came across while I was reading the works of one of my favorite writers. Consider the previous sentence, for instance. As I was writing that, it tended to become very long. I was running for a lack of an idea about how to finish it beautifully without shortening it. Suddenly Sam Harris crossed my mind and I thought he would have finished it. So I just tried to think like him to finish it. Now I really do not know how he would have himself finished it. I also do not know how he mus have thought. I really do not know his person. But I just tried to emulate him according to the mental picture I had in my mind about his personality. That mental picture may be wrong, but it helped me complete my job. This explanation is also partly inspired by him. Again, the inspiration might be wrong but the explanation is fine. The point that I want to make is that being well read helps in writing. Rest is easy, so long as you sit back and relax as you write.

Keep Your Sentences Short

One good tip is to write short sentences. Short sentences would make it easier for you to write. Shorter sentences are also easier to read and understand. It is also easier to convey your thought process with shorter sentences. As soon as your sentences tend to become large, you tend to get baffled about what you were writing about. Wasn’t the previous sentence too large? Honestly, by its end, I was myself confused about its fate. But it worked out well. The sentence is grammatically correct with no typos. But how long or short the sentences should be? I would suggest that sentences should not be too short either. Is this sentence too short? I presume it wasn’t. But the next sentence is a bit too short. It is. Avoid too short sentences with only a couple of words. Your write-up should also be grammatically correct. They should also convey your idea in a sequence. Short sentences should help you accomplish this.

I use Grammarly now for everything I write. Even right now, Grammarly is faithfully working in my browser like a tacit pet and suggesting me all the corrections that I should make to this text.

Do You Have Enough Words?

Improve your vocabulary. A good vocabulary will help you write your ideas more vividly. You will also have a much higher degree of freedom with a large choice of words to explain a particular theme or an idea. The best way to increase vocabulary is to indulge in a rigorous reading practice. You normally have to read a lot while reviewing literature. Keeping a dictionary and a thesaurus while reading will help you in understanding the content in a better way. A dictionary and a thesaurus also help you a lot while writing. Moreover, if you have ever prepared for a test like GRE, you must surely have developed a decent vocabulary. Tests like GRE compel you to memorize synonyms, antonyms, connotations, and denotations of thousands of words. They force you to maintain flash cards that have words on top and related information behind them. By going through such a practice for a prolonged period of time you will become a word giant.

Of Figures, Tables And Mathematical Equations

All empirical researchers would inevitably have to embellish their manuscripts with figures and tables at some stage. That stage will naturally arrive when you are done and dusted with your experiments and you have gathered all the necessary data for your article.

I normally use Matlab for my data analysis. One of the cool things about Matlab is that you can easily plot all sorts of graphs, charts, and figures with it. What type of figures and plots and tables you would exactly need can be learned from relevant research papers. In Matlab, you can also add axis labels and legends to your plots. Matlab plots are also editable. This is a great benefit as you can save a plot and then reopen it afterward to do necessary changes to it. Changes are inevitably suggested by reviewers. To address this issue, I always save the Matlab figure in four formats. I initially create a Figs folder in my working directory. In that, I create four sub-folders named fig, jpg, pdf, and eps. Each one of these is for a specific format of a figure as the names suggest. In the fig sub-folder, I save the Matlab editable figure. In the jpg, pdf, and eps sub-folders I save Jpeg, pdf and eps versions of the same figure. Like this, I create an archive of all the figures that will go in the article. I do not add captions to figures in Matlab, as I normally add captions through Latex. Moreover, I include the eps figures in the Latex document. Eps figures are best to be used with Latex. They embed seamlessly with the rest of the content and look wonderful.

Tables are easy to create in latex and you should kindly look up elsewhere on the web about this information. One of the great benefits of latex is that it handles mathematical equations pretty well. Please look on the web about how to add mathematical equations to your article. Similarly, writing equations in Latex is also very easy. Actually, Latex is well known for the way it handles mathematical equation. Look up on the web about how equations in latex and you will find loads of information.

Seek Professional Help and Work With a Co-Author

It is often said in research circles that writing your first article is the most difficult thing to do. Once you have done that, writing rest of the articles is really easy. This is quite true. However, you can make the process of writing your first article easier as well. You can do this by seeking professional help and write your article with the help of a co-author.

You can contact someone near you who is an experienced writer. It could be a senior person from your lab or research group, such as a postdoc, a faculty member, a friend or a colleague. Normally it is better to co-write with a postdoctoral researcher if you are yourself a Ph.D. student. I would personally never contact my supervisor to help me in writing a research article. I would try my best to present before him my finished work. So as a Ph.D. student, I would try to seek help from a postdoctoral researcher and co-write with him. Later on, I will acknowledge his/her help by having his name written in the article as a co-author. But here is how I would co-write with him/her.

Make an arrangement with your coauthor that he would help you with the writing process. Then write the article completely. Once you have done that, share it with your co-author and seek his reflections on it. Once you get his commentary, refine your article accordingly and share the refined version of the article with him/her again. Repeat this process many times so that you get a really professional manuscript shaped up. Normally, you may have to have to get the revisions a dozen times.

This can be a really nice way of writing your first article. And even in your subsequent attempts, you can follow this procedure. Having a co-author proofread and revise your work can do wonders to your research output.

I hope all of this helps. If you have any suggestions that I have missed, please share them in the comments below. I would love to get more advice about being able to write well easily.

Photo by matsuyuki