A scientific paper to be published in a journal generally captures the results and conclusions of a specific body of research. Research articles or papers are intended to further human knowledge by reporting new research.

Laboratory reports (‘lab reports’, also called ‘experimental reports’) are internal reports used to document and report on the results of experimental research, usually to other researchers in the group. The structure and approach are the same as for journal articles.

Each journal has instructions for authors (sometimes titled ‘author guidelines’ or ‘advice to authors’). The instructions cover many different areas, from formatting and maximum word limits to procedures for submitting papers and copyright issues. 

Structure

A scientific paper generally uses the following structure:

• Introduction. What was the problem or question?
• Methods. How did we study it?
• Results. What did we find?
• Discussion (which may include a separate Conclusion section). What do the results mean?

This structure helps you, as the author, to organise your information; it will also help the reader to obtain information efficiently. The 4 main parts of a paper link as 2 pairs:

• Introduction links with Discussion.
• Methods links with Results.

These links mean that the questions, challenges or problems posed in the Introduction should be considered and answered in the Discussion, and the structure of activities in Methods should link to the structure of outputs in Results. Having similar structures in the pairs of parts helps your reader to understand the information you are presenting.

Other components of a scientific paper may include:

• abstract
• author names and affiliations
• keywords
• abbreviations
• acknowledgments
• conflict-of-interest statement
• appendixes
• references
• additional online material.

Writing a scientific paper

It is easiest to write the paper in stages.

Stage 1

• Methods. It is generally easiest to start here because you will know what you did, the writing is straightforward and, if the study is based on other papers, you can often simply reference those:
  • Use mainly the past tense and the passive voice (the focus is on what was done rather than on who did it).
  • Give information in sufficient detail for your work to be reproduced. This may include precise details of equipment or software whose performance or capabilities are integral to the results.

  Computational modelling used version 123.45 of program XYZ.

  Data were collected on the B12345 beamline at the ABCD synchrotron.

  • If you conducted scientific research using human or animal subjects, indicate that you had appropriate approvals.
  • If you completed a literature review, discuss the breadth and depth of your research.
  • Use subheadings if the journal allows them; subheadings make this section clearer.
• Results. As with the Methods section, you will already have the results, so this section should be relatively straightforward. Use tables and figures, where appropriate. Check the instructions to authors to ensure that you meet the requirements of the journal.
• Other material. At this stage, you can also assemble some of the additional material, such as references, acknowledgments and conflict-of-interest statements (depending on the relevant instructions to authors).

Stage 2

• Introduction and Discussion. Because the Introduction links to the Discussion, it is helpful to write them together. The Introduction sets the scene by explaining:
  • what is known
  • where the knowledge gaps are
  • what questions you are trying to answer
  • how the current study aims to answer the questions and fill certain knowledge gaps.

  The Discussion should not repeat the results, but should:

  • link back to the Introduction, discussing what the results show (eg relationships between things, principles, generalisations) in relation to the question being asked
  • discuss how the results link to the broader field, with possible explanations for any anomalous findings (eg findings agree with X but disagree with Y, and possible reasons)
  • end with a strong conclusion (some publications will expect a separate Conclusion section).
• Abstract. The abstract must be easy to read and understand. It is the first thing that a journal editor and reviewer will read. If the paper is published, the abstract will be the part that readers use to decide whether to read the rest of the paper. An abstract summarises each of the main sections of the paper in 1 or more sentences. Therefore, it is best to write it once the body of the paper is complete. Depending on the journal, the abstract may contain subheadings, such as Aims, Methods, Results and Conclusion. When writing the abstract:
  • be brief, follow the usual structure for the specific journal, and ensure that the abstract is easy to read and understand on its own
  • do not include citations, references to tables and figures, or abbreviations (unless a term is used several times).
• Title. Although it is good to have a working title, you can leave the final version until you have written the rest of the paper. The title should be informative and fit within any length restrictions.