On this page
The data is solid, the analysis is correct, and the first full draft still comes back from a co-author or supervisor marked up almost entirely. Not because the research is weak — because the manuscript reads like a lab notebook: methods details surface in the results, the discussion repeats the results almost verbatim, and the conclusion claims more than the data actually supports.
This is a structure problem, not a science problem, and it's an unusually common one. Physicians are trained to present cases narratively — what happened, in order. A scientific paper doesn't follow that logic. Each section of the IMRAD structure (Introduction, Methods, Results, And Discussion) answers one specific question, and mixing those questions across sections is what makes a technically sound paper read as confused.
Why This Happens
Case presentation habits don't transfer directly
A case presentation is chronological: history, findings, course, outcome. A scientific paper is structured by function, not sequence — and that shift doesn't happen automatically just because both involve medical writing.
Each section feels like it should contain "everything relevant"
Without a clear sense of what belongs where, it's natural to include background context in the results, or restate raw numbers in the discussion — both of which blur sections that should stay sharply distinct.
The conclusion is written under the most fatigue
By the time a draft reaches its final paragraph, it's easy to summarize with more confidence than the data supports, simply because the finish line is close.
Common Mistakes We See
- An Introduction that reads as a literature review. The introduction should build a case for why the study matters and end with a clear objective — not summarize everything ever published on the topic.
- A Methods section that can't be replicated. Missing details — exact inclusion/exclusion criteria, the statistical test used, how missing data was handled — are among the most common peer review comments.
- Results that include interpretation. "This suggests that..." belongs in the discussion. The results section reports what was found, not what it means.
- A Discussion that restates the Results. The discussion should contextualize findings against existing literature and explain their significance — not repeat the numbers already given.
- A Conclusion that overreaches. Claiming a treatment "works" based on a single small study, rather than framing the finding proportionate to its actual strength of evidence.
Practical Recommendations
Each section exists to answer one specific question — writing to that question keeps the section focused.
| Section | Core question it answers | Typical length |
|---|---|---|
| Introduction | Why does this question matter, and what exactly are we testing? | 400–600 words |
| Methods | What did we do, precisely enough for someone else to repeat it? | 600–900 words |
| Results | What did we find? (Facts only — no interpretation) | 500–800 words |
| Discussion | What does this mean, how does it fit existing evidence, and what are the limitations? | 800–1200 words |
| Conclusion | What can be said, proportionate to the evidence? | 2–4 sentences |
Quick Checklist
- Introduction ends with a single, clear objective or hypothesis
- Methods include enough detail for another researcher to replicate the study
- Results contain findings only, with no interpretive language
- Discussion opens by directly addressing the primary objective stated in the introduction
- Limitations are stated explicitly, not omitted or minimized
- Conclusion is proportionate to the strength and size of the evidence
- The reporting checklist relevant to the study design (CONSORT, STROBE, PRISMA) has been followed
Need Expert Guidance?
Book a free call with PureMed and let our research specialists help you save valuable time, avoid costly research mistakes, and move confidently toward publication.
Frequently Asked Questions
Should I write the sections in order — Introduction first?
Not necessarily. Many experienced authors write Methods and Results first, since those are grounded in what was actually done, then write the Introduction and Discussion once the findings are clear.
How much literature should the Introduction cite?
Enough to establish the gap the study addresses — typically a handful of key, recent, directly relevant citations rather than a comprehensive review of the field.
What's the difference between limitations in the Discussion and weaknesses a reviewer might raise?
Ideally there's significant overlap. Proactively stating limitations — sample size, single-center design, potential confounders — signals rigor and often pre-empts the exact comments a reviewer would otherwise raise.
Can the Conclusion introduce a new point not covered in the Discussion?
No. The conclusion should summarize what has already been established in the discussion, not introduce new claims or data at the very end of the paper.