This guide is far more than a list of instructions on what to include in each section of your research paper. In fact, we will:
- Use a research paper I wrote specifically as an example to illustrate the key ideas in this guide (link to the full-text PDF of the research paper).
- Use real-world data (on 100,000 PubMed research papers) to show you how professional scientists write in practice, instead of presenting my own opinion on the subject.
- Provide practical tips on how to: improve your writing, find the right journal, and submit your article.
Let’s get started!
- Structure of a research paper
- Writing the Introduction section
- Writing the Methods section
- Writing the Results section
- Writing the Discussion section
- Writing the Abstract
- Writing the Title
- Writing optional sections
- Refining and improving your article
- Managing and formatting your References
- Submitting your article
1. Structure of a research paper
Most research papers follow the IMRaD structure that consists of 4 main sections:
The paper also has some essential elements–Title, Abstract, and References–and may contain other optional sections–Conclusion, Acknowledgements, Funding, Conflicts of interest, and Appendix.
These sections often appear in the following order:
The advantages of following the IMRaD structure are:
- To make the paper easily scannable by readers (since most won’t read the entire manuscript.
- To avoid repeating the same information in different places.
To follow the IMRaD structure, you must learn what information goes where.
So, here’s an overview of what each of the main sections represents:
|Introduction||Why you chose this topic and what is your objective|
|Methods||What you did and how you did it|
|Results||What you found|
|Discussion||What your results mean|
Together, these 4 sections start with the main topic of the paper and end up with a conclusion regarding that topic:
1.1. Where to start?
When writing a research paper, some people prefer to start with the Results section—since it comes out right from the data they just analyzed. Others start with the Methods section—since information about how they designed the study and analyzed the data is still fresh in their mind. Personally, I prefer to start with the Introduction section for 2 reasons:
- While doing a literature review for the introduction, sometimes I discover a problem in my approach or an interesting secondary objective that I did not think about, which as you can imagine, changes a lot of things in other sections of the article.
- I want to formulate the hypothesis before analyzing the data in order to avoid HARKing (Hypothesizing after the results are known) which is a major problem in statistics (see: 7 Tricks to Get Statistically Significant p-Values).
2. Writing the Introduction section
The Introduction targets a non-specialized audience, so when writing it, make sure to use simple and beginner-friendly terms.
2.1. Length of the Introduction section
The introduction section should be:
- 400 to 760 words long (3 to 5 paragraphs).
- The shortest section of the article (half the length of the other sections: Methods, Results, and Discussion).
(These data are based on an analysis I made on 61,518 articles from PubMed)
2.2. Structure of the Introduction section
Here’s what you should include in the Introduction:
- Step #1: Describe the general context of your work (your aim should be to convince the reader that the topic of your research is interesting).
- Step #2: Summarize the results of previous studies on the topic (report what others have found and provide references. But don’t do an in-depth literature review, a short summary of these findings is enough).
- Step #3: Identify the gap, problem, or limitations of previous studies (find the missing pieces of the puzzle).
- Step #4: State your objective, hypothesis, question that you want to answer, or problem that you want to solve (make sure that the purpose of your study is clear and understandable, otherwise people won’t care about your results).
- Step #5: Present your solution: explain the approach you used to achieve the objective, explain what is different about it and what makes it special. Here you have to sell your approach. But keep it short (leave the details to the methods section).
2.3. Verb tense and voice in the Introduction section
Use the past tense for things that were already done and the present tense for things that continue to be true today.
“Previous studies found that the rate of heart disease is increasing“.
“The goal of this study is to explore why the rate of heart disease increased in the past 10 years”.
You should write the Introduction using mainly the active voice.
“A recent study found conflicting results”.
Should be favored over:
“Conflicting results were recently found“.
Writing is not just about following a series of rules: you should keep an eye on the flow of your story that ties your paragraphs together.
Here’s an overview of the story of our Introduction section:
3. Writing the Methods section
The Methods section is the recipe for the study: it should provide enough information to replicate the study without looking elsewhere (although most of those who read the Methods section will not be interested in replicating your study, instead they just want to make sure that your study is credible).
The Methods is the most technical section of the article. So, unlike the Introduction, don’t shy away from technical terms, since those who are not interested in such details will most likely skip this section.
3.1. Length of the Methods section
The Methods section should be:
- 760 to 1,620 words long (6 to 14 paragraphs).
- The same length as the Results or the Discussion, and about double the length of the Introduction.
(These data are based on an analysis I did on 61,514 articles from PubMed)
3.2. Structure of the Methods section
Here’s what you should include in the Methods section:
- Step #1: Mention the study design, including:
- The date and duration of the study.
- The sampling procedure.
- The assignment to different study groups.
- Step#2: Describe the materials and equipment used, including:
- The source of the data.
- Any approval needed to conduct the study.
- Step#3: List the inclusion and exclusion criteria (i.e., the characteristics that participants must have to be included in the study).
- Step#4: Explain the procedure you followed to collect and clean the data. Make sure to clarify:
- The reason behind choosing such procedure.
- The order in which things were done (a flow diagram can simplify the description of complex procedures).
- Step #5: Justify the statistical methods used, including:
- The calculation of the minimum sample size needed.
- The role of each variable (dependent, independent, or control variable).
- The methods used to address bias in the study.
- The methods used to handle missing data.
- The measures used to summarize the data.
- The type of statistical test or model you used to test your hypothesis and the threshold for statistical significance (don’t go into detail about obvious statistical tests or models, but advanced methods should be either described or referenced).
- The statistical software used [optional].
3.3. Verb tense and voice in the Methods section
Use the past tense (because the things you did took place in the past).
“The data were downloaded“.
“A linear regression model was used“.
Use the passive voice (to avoid repeating the pronouns: “I” or “We”).
“Variables were summarized using the mean and standard deviation”.
“I summarized the variables using the mean and standard deviation”.
Note that in some cases, you will be forced to include some results in the Methods section. Although the research paper has a separate Results section (which we will discuss next), sometimes we include some results in the Methods section to justify the use of a certain material or method.
For example, in the Methods section above, in order to defend the use of the variable Journal impact factor instead of Journal, I ended up reporting the number of journals in the study (which is a number calculated from the data, so it normally belongs to the Results section):
“Since the data contained articles from 1,040 different journals (and to avoid complicating the analysis by creating 1,039 dummy variables), I ended up adjusting for the Journal impact factor, a direct descendent of the deconfounding variable Journal, thus representing most of its effect.”
4. Writing the Results section
In the Results section, you should describe and summarize your findings without explaining them (the interpretation should be left for the Discussion section).
4.1. Length of the Results section
The Results section should be:
- 610 to 1,660 words long (5 to 11 paragraphs).
- The same length as the Methods or the Discussion, and about double the length of the Introduction.
(These data are based on an analysis I did on 61,458 articles from PubMed)
4.2. Structure of the Results section
Here’s what you should include in the Results section:
- Step #1: Describe your sample data:
- At each stage and for each group of the study, report the number of participants (if some were lost to follow-up, provide the reasons).
- Use figures and tables to:
- Describe participants’ characteristics.
- Compare participants in different groups.
- Describe the main variables in the study.
- Step #2: Report the outcomes of statistical tests or models. For each, report:
- The statistical significance (the p-value).
- The precision (the 95% confidence interval).
- The practical significance (the effect size).
4.3. Using figures and tables
A table or a figure are useful to highlight important results or to represent a lot of numbers that, if reported in the text, can be unpleasant for the reader.
Here are a few rules regarding figures and tables:
- The supporting text should complement the table or figure but not repeat the same content.
- The table or figure should stand alone (i.e., the reader can understand it without referring to the text).
- The format of the table should be as follows:
- No vertical lines.
- Three horizontal lines:
- A line above the header row.
- A line below the header row.
- A line at the bottom of the table.
- No horizontal lines to separate data rows.
(Refer to the example below to see how your tables should look like)
4.4. Verb tense and voice in the Results section
Use the past tense for completed actions.
“In our sample of 9,830 articles, the median title length composed of 16 words (IQR = 6), had 2.2 yearly citations (IQR = 3.33), and was published in a journal with an impact factor of 2.74 (IQR = 1.67).”
Use the present tense for things that continue to be true today.
“The Poisson model shows a significant negative effect of longer titles on citation count.”
Use the active voice when possible.
5. Writing the Discussion section
In the Discussion section, you should explain the meaning of your results, their importance, and implications.
5.1. Length of the Discussion section
The Discussion section should be:
- 820 to 1,480 words long (5 to 9 paragraphs).
- The same length as the Methods or the Results, and about double the length of the Introduction.
(These data are based on an analysis I did on 61,517 articles from PubMed)
5.2. Structure of the Discussion section
Here’s what you should include in the Discussion section:
- Step #1: Answer the study objective (i.e., where the Introduction ended). Your first sentence can be: “We/I found that”, “This study shows/proves that”, etc.
- Step #2: Discuss your findings:
- Start with your main finding:
- Explain its consequences.
- Comment on whether it supports or refutes your initial hypothesis (i.e., was this result expected or unexpected?).
- Compare it with the results of other studies (if they contradict each other: explain why, and suggest a way for further studies to resolve this contradiction).
- Then discuss your secondary finding (if you have any) by following the same steps as you did for the main finding.
- Start with your main finding:
- Step #3: Point out the strengths of your study (e.g., the use of a new and superior method, a larger sample size, etc.).
- Step #4: Point out the limitations in the study design, data collection, and data analysis (these can be: possible sources of bias or confounding, or problems that affect the internal or external validity of your results). Make sure you discuss:
- How you addressed these limitations in your design and analysis (i.e., justify
the methods used in your study).
- What future studies should do to address these limitations.
- How you addressed these limitations in your design and analysis (i.e., justify
- Step #5: Conclude with a takeaway message that reminds the reader of your most important finding and its implications (this Conclusion paragraph is sometimes put in a separate section after the Discussion [for more information, see: Length of a Conclusion Section: Analysis of 47,810 Examples]).
5.3. Verb tense and voice in the Discussion section
Use the past tense for completed actions. For instance:
“I found that…”.
Use the present tense for things that continue to be true today. For instance:
“This study shows that…”.
Use the active voice when possible.
6. Writing the Abstract
The Abstract is a summary of the article.
6.1. Length of the Abstract
The Abstract should be 220 to 320 words long (1 to 4 paragraphs).
(These data are based on an analysis I did on 61,429 articles from PubMed)
6.2. Structure of the Abstract
In the Abstract, you should provide a summary of each section of your paper (It can be divided into subheadings, if the journal allows it):
- Step #1: Start with a one sentence introduction to the subject.
- Step #2: Mention the study objective.
- Step #3: Summarize the Methods section.
- Step #4: Highlight key results in numbers (including data is important for researchers who want to cite your article based only on the Abstract).
- Step #5: End with a one sentence conclusion (i.e., skip the detailed discussion of the results and go straight to the takeaway message).
7. Writing the Title
The last thing one discovers in composing a work is what to put first.Blaise Pascal
The Title’s role is to describe the content of the article and attract people to read it. Remember that only 18% of those who read the title proceed to read the Abstract [Source: Mabe and Amin, 2002].
7.1. Length of the Title
The Title should be 11 to 18 words long (80 to 129 characters).
Keep your Title as short as possible, since:
- Google shows only the first 60 characters of titles in their results page, so longer titles will be truncated when they appear in Google search.
- High-impact journals tend to publish articles with short titles.
(These data are based on an analysis I did on 104,161 titles from PubMed)
7.2. Structure of the Title
The Title should:
- Mention the central question or the purpose of the study (including important variables).
- Be front loaded: this means that the keywords should be close to the beginning of the title (remember that readers are scanning the title and they want to determine as fast as possible if they are interested in your article).
- Have a meaningful short version. For those searching online, Google will show them only the first 60 characters of your title and the rest is truncated. So, make sure to pack enough information in this part for users to be able to judge whether they want to click it.
- Mention the study design [optional].
- Avoid abbreviations and jargon. For instance: “The effects of having CVD on the psychological status“ should be replaced by “Psychological effects of cardiovascular disease”.
7.3. Example: writing a Title
The following figure shows how the Title of our example article follows the structure discussed above:
8. Writing optional sections
8.1. Writing the Acknowledgement section
In this section, you should acknowledge any significant technical contribution, permission, advice, suggestion, or comment you received.
“I would like to thank Prof. John for assistance with choosing an appropriate study design”.
“Thanks are due to all the hospital crew members who contributed their time and effort to make the data collection feasible in the shortest time possible”.
8.2. Writing the Funding section
In this section, you should provide the sources of funding, or the sources of the equipment and materials used in the study, and the role of funders.
“The authors received no financial support for the research, authorship, or publication of this article”.
“This work was supported by [name of the funder, and grant number]”.
8.3. Writing the Conflicts of Interest section
In this section, you should state if you have any direct or indirect competing interests that may have influenced the outcome of the study, such as: financial, work, personal, or religious interests.
“The authors declare that they have no conflicts of interest”.
“The corresponding author was a former employee in company X that sells the main product used in this study”.
8.4. Writing the Appendix
In this section, you should provide supplementary information that was too large to be included in the main text, such as: data, questionnaires, and additional details on the materials and methods used.
9. Refining and improving your article
The following is a list of useful tips to improve your writing:
- Avoid jargon, be concise, and focus on saving your readers’ time. The truth is that nobody enjoys reading, if readers can download information into their brain, they would!
- Assume that your readers are beginners: so, use terms that are easy to understand.
- Avoid acronyms when possible.
- Use active instead of passive voice, i.e., subject-verb-object. But, you can use the passive voice when:
- You don’t know the subject.
- You don’t want to repeat the pronouns ”I” or ”We” in many places in the same paragraph (although it would be fine to use them sparingly, see: ”I” & ”We” in Academic Writing: Examples from 9,830 Studies).
- You want to emphasize what was done instead of who did it (especially in the Methods section).
- To maintain the flow of ideas (for more information, see the video lecture by Steven Pinker below).
- Write short sentences and paragraphs: each paragraph should be between 2 and 6 sentences long (65 to 167 words), and should cover a single topic. (For more information, see: Paragraph Length: Data from 9,830 Research Papers)
- Get rid of hedge words: e.g. ”These results might suggest that a fair amount of x is suspected to have a meaningful impact on y”. These make you sound hesitant or unsure about what you are talking about.
- Avoid using “They” or “Their” when the subject is singular. For a gender-neutral language, revise the sentence to make the subject plural. For instance, use: “Participants were assigned according to their choosing” instead of “Each participant was assigned according to their choosing”.
For more writing tips, I highly recommend this lecture by Steven Pinker:
10. Managing and formatting your References
When it comes to references, you should:
- Cite between 25 and 56 references overall (approximately 1 reference for every 95 words or 4 sentences) [Source: How Many References Should a Research Paper Have? Study of 96,685 Articles].
- Aim to find those published within the past 13 years [Source: How Old Should Your Article References Be? Based on 3,823,919 Examples].
- Cite the original source, not secondary sources.
- Cite research papers and books instead of websites and videos (unless these contained original data not available elsewhere).
- Use a citation management software to collect and organize your references. I recommend Zotero® since it is free, easy to learn, and has a lot of tutorials online.
11. Submitting your article
Here’s a step-by-step description of how to find a journal and submit your article:
- Choose a journal (ideally, before writing the article):
- Go to: The Directory of Open Access Journals (This is a database of 17,614 journals that publish open-access articles–i.e., if you publish in these journals, your article’s full-text will be available for free to your readers).
- On the left sidebar, select the following options:
- Under SEE JOURNALS, select: Without article processing charges in order to exclude journal where you have to pay to publish your article.
- Under SUBJECTS, choose: the domain that is closest to the topic of your article.
- Under LANGUAGES, select: English.
- Select a journal from the suggested list.
- Go to the journal’s website, look for their “Instructions for authors”, and format your article accordingly.
- Sign-up to their website and submit your article.
Once your article is submitted, the editor takes a look at it and may:
- Reject your manuscript right away (in the matter of hours or a couple of days). This happens if:
- The topic of your article is not interesting for the journal’s audience.
- Your work is not important enough to be published in that journal.
- Send you paper for peer review. After a few weeks, you will get their decision. It will be one of these 4 options:
- Rejected: In this case, you have to send your article to another journal (don’t get discouraged by rejection, sometimes important articles get rejected).
- Rejected, but can be resubmitted after making some major changes suggested by the reviewers (for instance, expanding, deleting, or re-writing major parts of the article): in this case, you can either revise and resubmit, or look for another journal.
- Accepted, but needs minor changes.
- Accepted (without the need for changes).
When you want to revise and resubmit your article, you should prepare 2 things:
- A revised manuscript with all the modifications you made highlighted (to make it easy for the reviewers to see what you changed).
- A response for the reviewers where you address their comments point by point: you can either agree or disagree with their recommendations (but, in case you disagree, you should explain the reason).
Once your paper is accepted, you will get a final version formatted in the journal’s style. Be careful to look for errors before you accept this final version.