IMRaD architecture
Use the introduction to define the gap, methods to explain the test, results to present evidence and discussion to interpret it.
WORKFLOW STAGE 04
Build a clear, traceable manuscript in which every claim is supported by evidence, every method is reproducible and every revision is documented.
Scientific writing is not decoration: it is the documented chain from research question and method to result, limitation and conclusion.
Use the introduction to define the gap, methods to explain the test, results to present evidence and discussion to interpret it.
Align terminology, objectives, primary results and conclusions across the title, abstract, main text, tables and supplementary files.
Report data source, sampling, instruments, preprocessing, parameters, software versions, ethics, analysis and stopping rules.
Present findings without speculation in Results; interpret mechanisms, prior literature and implications in Discussion.
State bias, uncertainty, validity threats and generalization limits without weakening claims through vague or defensive language.
Answer every comment, identify exact manuscript changes, justify declined requests with evidence and maintain a revision log.
Complete the technical audit before journal upload or thesis review.
| Claims & evidence | Each major claim maps to a cited source, result, figure or analysis; causal language matches the study design. |
|---|---|
| Figures & tables | Readable labels, units, sample sizes, uncertainty, legends and standalone captions; no duplicated information. |
| References | Accurate metadata, consistent style, verified DOI or URL, balanced literature and no unsupported or irrelevant citations. |
| Reporting standards | Apply relevant guidance such as CONSORT, PRISMA, STROBE, CARE, ARRIVE, TRIPOD or discipline-specific checklists. |
| Research integrity | Disclose ethics approval, consent, funding, conflicts, data/code availability, author contributions and use of AI tools where required. |
| Journal compliance | Confirm scope, article type, word limits, anonymization, file formats, declarations, supplementary files and cover-letter requirements. |
Avoid: overstating conclusions, mixing results with interpretation, hiding limitations, using unverifiable references, changing data after review without disclosure, or submitting before checking journal-specific instructions.
Share the manuscript, journal guidelines or reviewer comments for structured support.