Back to all Advice

The process of writing a paper 

Created 10/18/2020, last revised 7/18/2024

A research paper is a document that organizes complex experiments and findings into a structured format that explains your ideas to others. Given the complexity of the ideas and document, trying to write from beginning to end is a poor strategy for this task. Instead, there is a non-obvious but efficient sequence of steps that can take you from a paper idea to a fully-developed manuscript. Additionally, it is important to build in review and feedback steps during the process. 

I originally developed this document for my PhD advisees, to communicate how I coordinate the review process with them. I think you will find this sequence of steps useful, even if your advisor don’t use this process.

Stage 1: Paper scope and conclusions

When we are talking over research, we will get to the point where we say, “This looks good, let’s write it up.” Questions to answer at this point are:

• What is the topic?

• Who is your audience? At first, figure out the domain of the audience (structural engineers, city planners, etc.) Next, list names of specific people you will want to read the paper. Visualizing specific people will help you think about whether your conclusions are relevant and valuable to them. It will also help you calibrate how much background information you need to provide. Related to this, pick a target journal to publish in.

• What do you want your audience to learn?

This is also the time to discuss authorship and (if appropriate) invite the other authors to join the effort. By establishing the authorship and roles of each author now, the later steps will be much smoother. It may feel tempting to wait on discussing this with potential coauthors, out of uncertainty, or a hope that it will work out naturally, but delaying usually creates complications. Your colleague may feel surprised or left out if they didn’t know you were working on a paper, and you may feel uncomfortable adding them if they haven’t contributed to the paper preparation. Conversely, you don’t get the benefit of feedback and help from your coauthors if they haven’t yet committed to helping coauthor the document.

Stage 2: Outline

This stage is like putting together a puzzle. There are lots of complicated ideas that need to be interwoven, and it can be a fun challenge to piece it all together. There is usually no single right way to put everything together, so try a few ideas and don’t get caught up in careful writing at this point.

Some ideas to keep in mind at this stage:

• Focus on the conclusions first. The rest of the paper builds towards your conclusions, so knowing the ‘finish line’ will guide what else you need to present. I will look carefully for your conclusions and use them when commenting upon the appropriateness of other material in the paper.

• Identify the figures needed to explain your ideas and conclusions. Create rough-draft figures and place them in the outline. You will add, subtract and revise figures as you proceed, so don’t spend much time at this point polishing your figures. Often, the figures you can use here are ones you have used in PowerPoint presentations.

• The best structure for your paper is not to describe the chronological order in which you performed your work. The chronological order may be ingrained in your head, but there are usually clearer ways to explain your analysis and findings.

• Develop bullet points of critical concepts and observations to discuss in each section of the outline.

• Look at good papers by others on similar topics. How did they organize their material? How much space did they devote to introducing key concepts? Exemplar papers are great resources to help you structure your outline.

• You will have more results than you have room to discuss, and presenting everything will obscure your primary results. The same goes for background information. There are lots of things that might be interesting to discuss, but are they necessary? You can always save extra results for other papers, appendices, etc., so it’s not as if the material is lost forever.

• Don’t worry much about the literature review at this point. It will be easier when writing your first draft in the next stage.

Your tasks at this point

• Rank the top two or three conclusions that you hope the reader takes away after reading your paper. Now look at the text and figures: have you focused sufficiently on these main points? Have you spent an excessive amount of time on anything that isn’t a major point of emphasis?

My reviewing focus

1. Are you focusing on the most important contributions?

2. Are the concepts in the paper presented in a logical order?

3. Are the choices of figures and examples appropriate?

Stage 3: First draft

Write the text of the paper. After you have the outline, key concepts and figures prepared, it will be easier to decide what text is needed in each section. Write quickly, and don’t worry about making your writing beautiful or grammatically perfect; if you are thinking about grammar, then you won’t be able to concentrate on the technical content and structure. And you may delete big chunks of text later, so don’t waste time making it perfect yet.

Draft a title, and an abstract that summarizes the topic of the paper, the key approaches used in the analysis, and key findings of the work. You may re-write the abstract later, but at this point, it is an excellent test to see if you have those concepts identified.

Your tasks at this point

• Review the structure of the paper. If you just read the headings without the text, is the outline of your paper clear and logical? Does all of the text beneath each heading fit within the stated topic?

• Check each paragraph—does it have a coherent topic? Is that topic apparent in the first sentence? Are the paragraph topics arranged in a logical manner? Do the transitions between paragraphs make sense? If you are answering no, that is a sign of incoherent writing.

• Show a copy of your draft to some fellow students—see if they find it interesting or understandable.

• Use my review comments as an opportunity to learn. How does the revision improve the paper? Are you making consistent mistakes that you can learn to avoid in future writing? Do not take corrections personally. There might be many corrections on the pages, but that is a natural part of the collaborative writing process and a result of an outside perspective bringing new ideas.

My reviewing focus

1. Are you providing appropriate detail in each section? Do any sections need substantial lengthening or shortening?

2. Are the conclusions supported by evidence?

Stage 4: Refined drafts

Once the paper’s structure is in place, you can turn your attention to details in the paper. Below are items to check as your writing is getting more polished. Don’t try to do these while you are writing your first draft; they will distract you from writing down your ideas. Instead, finish your draft and then go back later to see how you can “tighten up” what you’ve written. See my paper formatting advice article for additional items to check at this stage.  

1. Focus extra attention on the abstract, conclusions, and introduction in that order. Many readers will only read those sections, so they should clearly explain what you have done and why it is important.

2. Check that you are getting to the point quickly. For example, if you have a section called “Objectives,” do you state an objective in the first sentence or two of the section? Taking a long time to get to the point frustrates readers and obscures your key points.

3. Avoid emotional or judging words in technical writing. Saying that an issue is “of utmost importance,” is “obvious,” or is “trivial,” may seem reasonable to you, but not everyone will agree with your conclusions. By using those words, you tell readers how they should react, instead of guiding them to draw their own conclusions. It is more professional to be dispassionate in your writing and let the results speak for themselves.

4. When you use “this,” “it,” or similar words, check that the meaning is clear. A common example is a sentence starting with “This means that…” Is it crystal clear what “this” refers to, and would the sentence still work if you substituted in the name of the thing you are referring to?

5. Check that you have appropriately used qualifying words such as “might,” “could,” “may,” and “seem.” Are you using them because you lack support to make a more definitive statement (if so, they’re fine) or out of a general lack of confidence or desire to be polite (which can be confusing for readers)? Be definitive whenever you can, and hesitant whenever it’s called for.

6. Use fewer words and simple words, when possible. Consise writing is pleasant to read and easier to understand.

   -Bad: It can be seen that in Figure 1, the dependence between the X variable and the Y variable can be approximated by a linear relationship.

   -Good: Figure 1 shows that X and Y have an approximately linear relationship.

7. When you are proofreading, focus on a single issue during each of your passes through the document. If you multitask, you will miss problematic items. Some items to check during individual passes:

   1. Your numerical data and equations are correct.

   2. You have defined all of your acronyms and symbols.

   3. You have used consistent terminology to refer to key ideas.

   4. You have included accurate references for all of your citations (use a reference manager like Zotero).

8. Consider using a text-to-speech tool to read your paper out loud to you, as that will help you notice awkward or incorrect phrases that you don't notice when reading. Consider also using tools like Grammarly to review your text and suggest grammatical corrections.

My reviewing focus (I will take multiple passes on versions at this stage)

1. Can sections or paragraphs or sentences be shortened?

2. Are figures clear? Can anything be added or subtracted to make them better serve their purpose?

3. Can phrases be re-written to be clearer?

4. Did you follow all of the above rules?

Stage 5: Submit the paper for review

See this article for details.