Before you start writing a research manuscript
The most common mistake when learning how to write a research manuscript is starting to write before you know what you are writing. A research manuscript is not a record of everything you did. It is an argument structured to make a specific contribution legible to a specific audience.
Two things determine the shape of your research manuscript: your target journal and your data. Get both clearly defined before you write a word.
Know your target journal
Different journals have different structural requirements, length limits, section headings, and citation styles. Before writing, read the author guidelines for your target journal and note:
- Word count limits for the abstract and full manuscript
- Required section headings (some journals use non-standard structures)
- Citation style (APA, Vancouver, Chicago, journal-specific)
- Figure and table formatting requirements
- Whether supplementary materials are permitted
Writing a manuscript for the wrong format and then reformatting costs significant time. Knowing the constraints early shapes your decisions throughout.
Gather your materials
Before writing, assemble:
- Your data, analysis outputs, and figures in their final form
- All papers you plan to cite, imported into a citation manager
- Notes from your literature review identifying the relevant prior work
- Any co-author agreements about division of writing tasks
If your data is still changing, your manuscript will change with it. Finalize your core findings before writing the results and discussion.
The structure of a research manuscript
Most empirical manuscripts follow the IMRaD structure: Introduction, Methods, Results, and Discussion. Some fields and journals use variations, but the logic underlying IMRaD applies broadly.
Title
The title should be specific enough that a researcher in your field understands the scope and contribution immediately. Avoid vague titles that could describe any paper in your subfield. Include the population, intervention or variable of interest, and outcome or finding if space allows.
Abstract
The abstract is not a preview of the paper. It is a standalone summary that must contain the motivation, methods, key results, and conclusion. Most journals specify structured abstracts with labeled sections (Background, Methods, Results, Conclusion). Even when structure is not required, covering all four elements is expected.
Introduction
The introduction does three things: establish the significance of the research question, identify the gap in existing knowledge that your work addresses, and state what your study contributes. It is not a comprehensive literature review. It is a focused argument for why your specific study was worth doing.
Structure the introduction as a narrowing funnel: broad context, specific gap, your contribution.
Methods
The methods section must be reproducible. Another researcher in your field should be able to replicate your study from your methods description alone. Be specific about: study design, participants or materials, procedures, variables and how they were measured or operationalized, and analysis approach.
This is the section most likely to be scrutinized in peer review. Ambiguity here is a common rejection reason.
Results
Present your findings without interpretation. Use tables and figures for data that would be harder to follow in prose. For statistical results, report effect sizes and confidence intervals alongside p-values.
Order results to follow the logic of your research questions, not the chronological order in which you ran the analyses.
Discussion
The discussion interprets your results in the context of prior work. Address your hypotheses or research questions directly. Discuss how your findings agree with or differ from existing literature and why. Acknowledge the limitations of your study honestly. Conclude with the theoretical and practical implications.
The discussion is where you earn the claim your introduction made. Stay within what your data supports.
Conclusion
Some journals include a separate conclusion section; others end with a concluding paragraph in the discussion. Keep it brief. Restate the main finding and contribution without simply repeating the abstract.
References
Generate your reference list from your citation manager, not by hand. Format them according to the target journal's style. Check for completeness: every in-text citation should have a corresponding reference, and every reference should be cited at least once.
Writing strategies that work for a research manuscript
Write the methods first
The methods section is the most straightforward part of the manuscript to write because it describes what you actually did. Starting here builds momentum and forces you to be precise about your study before you start interpreting results.
Use your data to structure the results
Do not try to write results prose and then figure out where to put the tables. Create your tables and figures first, in the order that matches your research questions. Then write the prose to narrate what they show.
Write the introduction last
This sounds counterintuitive, but the introduction is actually easier to write once you know what your results and discussion say. You need to set up the gap that your paper fills, and you only fully understand what you filled once the paper is complete.
Many experienced researchers write a draft introduction first to orient themselves, then rewrite it last to ensure it accurately frames the paper as it actually came out.
Managing citations while you write
Whether you use a standalone word processor or a built-in editor, you should never type references manually into a manuscript. Every citation should be inserted from your citation manager.
In Alfred Scholar's manuscript editor, citations are inserted inline using a shortcut that searches your workspace library. The reference list is generated automatically in your chosen style. Switching styles (for example, from APA to Vancouver when you target a different journal) takes seconds.
If you are writing in Word or Google Docs, use Zotero or Mendeley's word processor plugins to insert citations and generate the bibliography. The workflow is similar: you insert a citation placeholder that the plugin resolves.
Avoid the habit of using footnotes or notes as placeholder references with the intention of formatting them properly later. Placeholders get lost, and formatting a list of incomplete references under deadline is a reliable source of errors.
Checking submission guidelines
Most manuscript rejections based on formatting can be avoided by reading the submission guidelines before writing. Before you submit, verify:
- Word count is within the stated limit
- Abstract is correctly structured and within the word limit
- Figures and tables are in the required format and resolution
- Citation style matches the journal's requirements
- Author information is formatted correctly
- Any required statements (conflicts of interest, data availability, ethics approval) are included
Alfred Scholar includes a submission guideline check in the manuscript editor that flags common formatting issues against configurable requirements.
Getting feedback from co-authors
Most research manuscripts involve co-authors, which means feedback cycles that can span weeks if not managed deliberately. Common issues:
- Multiple versions in circulation simultaneously
- Conflicting edits that overwrite each other
- Reviewers commenting on different drafts
- Citation inconsistencies introduced when different authors add references
Using a shared writing environment where all collaborators work on the same document eliminates most of these problems. Alfred Scholar's collaboration feature allows co-authors to edit the same manuscript with real-time sync. Supervisors and reviewers can be added with viewer or editor roles.
If you cannot coordinate on a single platform, at minimum designate one person as the version-control owner and enforce a naming convention for draft files that includes the date and version number.
Common manuscript mistakes
Writing for a general audience instead of your field. Research manuscripts assume domain knowledge. Do not over-explain standard methods or concepts that every reader in your field knows.
Citing to support claims that do not need support. Citing "the Earth is approximately 4.5 billion years old" wastes space and signals unfamiliarity with the genre. Reserve citations for claims that are contested, field-specific, or taken directly from another work.
Over-claiming in the discussion. The discussion should be constrained by your actual findings. "These results suggest" is accurate. "These results prove" almost never is.
Burying the contribution. The specific claim your paper makes should appear explicitly in the introduction and conclusion. Reviewers who cannot identify your contribution will reject the paper.
Ignoring word limits. Word limits are enforced. Submissions that exceed them are often returned without review.
Tools for manuscript writing
Most researchers write manuscripts in one of three environments: Microsoft Word (ubiquitous, works with citation manager plugins), LaTeX (standard in mathematics, physics, and computer science; use JabRef or Zotero for citations), or Google Docs (convenient for collaboration; citation manager integration is more limited).
Alfred Scholar's built-in manuscript editor keeps citations, research documents, and writing in a single workspace, which is useful if you want to refer back to source papers while writing without switching applications.
The right tool depends on your field's norms and your co-authors' preferences. The most important thing is that whatever you use integrates with your citation manager so that references are always inserted programmatically rather than manually.
For more on managing references throughout the writing process, see How to Manage Citations Across Multiple Research Papers. For checking your manuscript for plagiarism before submission, see How to Check Your Thesis for Plagiarism Before Submission.