Academic guidelines
Researchor exists for research support. It does not exist to produce work that someone else submits as their own. This document states where that line falls, with worked examples, so both sides of a project know before it starts.
Academic integrity
Every engagement here is bounded by one rule: the person who submits the work must be the person who did the thinking it represents. Expertise can be bought: methodological advice, analysis, editing, training. An assessed submission cannot.
This applies regardless of what a client asks for, what an expert is willing to do, or what either party is paid. Breaching it ends the engagement and the account, and we cooperate with institutions investigating misconduct.
Allowed research support
These are legitimate professional services. They are what the marketplace is for.
- Research consultation and advisory work
- Literature review support and search-strategy design
- Data analysis and statistical analysis
- Research methodology consultation
- Research design
- Questionnaire and instrument development
- Data cleaning, coding and transformation
- Data visualisation
- Systematic review support
- Meta-analysis support
- Research editing and proofreading
- Referencing and citation management
- Research training and skills coaching
- Manuscript preparation support
- Research proposal and grant consultation
Prohibited academic work
These are not negotiable, and no amount of framing makes them acceptable. Requesting or accepting any of them is grounds for removal.
- Academic cheating of any kind
- Sitting or completing an exam, test or timed assessment for someone else
- Completing assignments, coursework, theses or dissertations intended for submission as the requester's own work
- Fabricating research data, participants or samples
- Falsifying results or analytical output
- Fabricating references, or citing sources not consulted
- Plagiarism, including undisclosed reuse of your own prior work
- Manipulating findings to reach a predetermined conclusion
- Creating fake participants or fabricated datasets
- Buying or selling authorship
- Misrepresenting qualifications, affiliation or experience
- Any activity that breaches institutional, funder, publisher or ethics-committee rules
Where the line falls
The same skill appears on both sides of every pair below. What changes is not the difficulty of the work. It is who ends up having done it.
Statistics
Allowed
“Run a multilevel model on my dataset, produce the output tables, and write a methods note explaining the specification so I can interpret and write it up.”
Not allowed
“Run the analysis and write my results and discussion chapters so I can submit them.”
The first buys expertise you then own and can defend. The second buys a submission you did not write.
Literature
Allowed
“Build me a search strategy for this question, screen against these criteria, and give me a summary matrix of the included studies.”
Not allowed
“Write my literature review chapter.”
Search strategy and screening are methodological work. The chapter is the assessed output.
Writing
Allowed
“Proofread my draft for language, consistency and reference formatting, and flag where the argument does not follow.”
Not allowed
“Rewrite this so it reads like my supervisor expects, and fill in the sections I have not written.”
Editing improves work that exists. Writing the missing sections creates work that is not yours.
Teaching
Allowed
“Teach me propensity score matching using my own data, so I can run it myself.”
Not allowed
“Do my assignment on propensity score matching.”
The first is tutoring. The second is the assignment.
Data
Allowed
“Clean and code this dataset, document every transformation, and tell me what the data cannot support.”
Not allowed
“I need 200 more responses. Generate them to match the pattern in what I have.”
Cleaning makes real data usable. Generating responses is fabrication, and it is research fraud in any jurisdiction.
Publication
Allowed
“Format my manuscript to this journal's requirements and check the reporting against the CONSORT checklist.”
Not allowed
“Add me as an author on a paper I did not contribute to.”
Formatting and reporting checks are preparation. Authorship is not something that can be bought.
Research ethics
Where a project involves human subjects, animal subjects or identifiable personal data, the client is responsible for holding the necessary ethical approval before work begins, and for saying so in the brief. An expert who believes approval is required and not held should decline the project and report it.
Responsible AI use
AI tools are part of research practice, and using them is not misconduct. Concealing their use, or letting them replace the intellectual contribution being claimed, is. Disclose AI assistance where the client, institution, funder or publisher requires it; never present generated text as an expert's own analysis; and verify every citation. AI systems produce plausible references that do not exist, and submitting one is fabrication regardless of where it came from.
Data integrity
Data must reflect what was actually collected or observed. Do not invent, duplicate, selectively delete, or alter data to change a result. Analysis choices should be defensible and disclosed. Where a result is inconclusive, say so. An honest null finding is a research output, not a failure to be engineered away.
Plagiarism
Submitting another person's words, ideas, figures or code as your own is plagiarism, and so is reusing your own previously published work without disclosure. Every deliverable on Researchor is screened before the client sees it, and the result is shown to both parties.
Research misconduct
Fabrication, falsification and plagiarism are the three recognised forms of research misconduct, and all three are prohibited here. So is any attempt to engineer a predetermined conclusion: selecting analyses until one gives the desired result, dropping inconvenient cases, or presenting exploratory findings as confirmatory.
Human subjects and ethical approval
Do not share participant data you are not authorised to share, and anonymise wherever you can. State in the brief which approvals you hold and which conditions they carry. An expert receiving participant data inherits the obligations attached to it.
Confidential research data
Unpublished data, pre-publication manuscripts and proprietary datasets are treated as confidential by both parties. Use material only for the project it was shared for. Do not publish, redistribute or reuse it, and do not paste it into a third-party AI tool without authorisation.
Reporting a concern
If you are asked for something on this page's prohibited list, decline it and report it. Reports are reviewed by a person, and reporting in good faith never counts against your account.
Report a breach of these guidelines
We aim to reply within 2 business days.
Related: Safety & trust articles · Terms of Service