The guide
How to write a research paper
A research paper has a shape reviewers expect: introduction, methods, results and discussion, an abstract that says it all in a paragraph, and a reference list in the journal's style. Getting the science right is the hard part; getting it onto the page in that shape takes longer than it should.
General chatbots write fluent paragraphs with no idea what your study found, and they cite papers from memory, which is how made-up references end up in manuscripts.
The shape of a research paper
Most empirical papers follow IMRaD: an introduction that says why the question matters, methods that let others repeat the work, results that report what you found, and a discussion that says what it means. The abstract compresses all four into one paragraph.
Each section answers a different question for the reader, which is why mixing them weakens a paper. Interpretation in the results, or new results in the discussion, are among the first things reviewers notice.
Three common paper types
The type sets the structure and the length your journal expects.
- Empirical articleReports new data or experiments. Usually IMRaD.
- Review articleSynthesises published work on a question. Structured by theme.
- Short communicationA brief, focused finding. The same parts, much shorter.
A common writing order
- Methods
- Results
- Discussion
- Introduction
- Abstract
Many researchers write in this order rather than front to back. Methods and results are the most concrete, so they are easier to start with, and writing them first fixes what the paper actually shows.
The discussion follows from the results, and the introduction can then set up exactly the question the paper answers. The abstract comes last, when you know what the paper says.
Common mistakes
Avoid: Interpreting results in the results section
Do: Report first, interpret in the discussion
Avoid: An introduction that reviews everything
Do: Review only what leads to your question
Avoid: Writing the abstract first
Do: Write it last, from the finished paper
A worked example
Background: one sentence on why the question matters. Methods: one sentence on who was studied and how. Results: one or two sentences with the main finding. Conclusion: one sentence on what it means and for whom.