ProWrite ResearchAcademyExplore coursesCredible work makes its reasoning traceable.
Practical principles for using evidence, sources and AI tools without surrendering accuracy, authorship or accountability.
A baseline for trustworthy research and writing.
Adapt these principles to the discipline, study design, audience and rules governing each project.
- 01
Keep the question visible
Let the research question—not a convenient dataset or preferred answer—drive the design and interpretation.
- 02
Record the method
Document searches, inclusion decisions, instruments, transformations and analytical choices well enough to audit the work.
- 03
Represent sources fairly
Read the original source, preserve its context and cite it for the exact claim it supports.
- 04
Separate findings from inference
Distinguish what the data show from interpretation, explanation and recommendation.
- 05
Report uncertainty
Describe limitations, missing data, alternative explanations and the boundaries of generalization.
- 06
Respect authorship
Credit substantive contributions and do not assign authorship to people or tools that cannot take responsibility for the work.
- 07
Use AI transparently
Follow relevant disclosure rules, verify every output and never invent or rely on unchecked citations.
- 08
Protect confidential material
Do not submit private data, unpublished manuscripts or restricted material to an unapproved system.
- 09
Correct the record
Respond openly to errors and make material corrections visible to readers.
- 10
Follow the applicable standard
Use the reporting, ethics, citation and institutional requirements that govern the specific work.
Pause before publishing when…
- A central claim is not supported by a source you have checked.
- The method cannot be reconstructed from the available record.
- Results and interpretation are blended in a way that hides uncertainty.
- Confidential material has entered an unapproved system.
- AI contributions, conflicts or material limitations are undisclosed.