Outlier Science
Research Integrity and Responsible Research Support
Research support should strengthen the quality and clarity of the work while preserving the responsibility of the people conducting it. These principles explain the responsibilities and boundaries that should guide contributions, data handling, analysis, writing and publication support.
On this page 8 sections
01
Genuine research and accurate evidence
An engagement must be based on genuine research questions, legitimate materials and an accurate account of the work performed. Data, references, approvals, participant records and results must not be fabricated or altered to create a preferred conclusion.
The client is responsible for providing truthful information about the project and the rights to use its materials. Contributors are responsible for the work they undertake and for making limitations visible within their scope.
If the available material cannot support the requested task, that limitation should be explained. Missing records should not be represented as observed results, and an illustrative example should never be presented as a finding from a real study.
02
Analysis that follows the question and evidence
Analytical decisions should be appropriate to the design, data and research question. The rationale for material exclusions, transformations, missing-data treatment and changes to an agreed analysis plan should be documented.
Support must not promise statistical significance, hide inconvenient findings or repeatedly change analyses merely to obtain a publishable result. Exploratory analysis can be useful when its role is made clear. A finding should be described with its uncertainty and limits.
The researcher or responsible author must be able to review the analysis and understand the conclusions used in the work. Where code, technical notes or further explanation are needed for that purpose, they should be considered in the scope.
03
Contributions and authorship
Authorship reflects qualifying scholarly contributions and responsibility under the applicable journal and discipline rules. Payment for a service does not purchase authorship, and an administrative, editorial or technical contribution should not automatically be represented as an author role.
For biomedical work, the ICMJE authorship recommendations provide a relevant reference. Other fields and journals may apply different criteria. The actual requirements should be checked for the project.
The scope should make intended contributions clear at the outset and revisit them if the work changes. Appropriate acknowledgements, contributor statements and disclosures should describe what was done accurately. No contributor should be named without agreement.
04
Writing support and assessed work
Academic editing can improve language, structure and presentation. Methodological feedback and collaborative writing can make more substantive contributions. The assignment should distinguish these roles and retain the authors' responsibility for the content.
For doctoral or other assessed work, support must fit the institution's rules. Coaching, methods advice and permitted editing can help a researcher develop their own work. Undisclosed completion of assessed work on the researcher's behalf is outside the scope of this support.
Authors should review substantive changes, verify references and confirm that the manuscript represents the research accurately. A polished document does not remove the need to check its evidence.
05
Publication and research assessment
Journal selection should consider the research itself, intended readers, journal scope, policies, costs and current coverage. Editorial decisions belong to the journal. Publication support must not imply guaranteed acceptance, access to an editor or influence over peer review.
Citation manipulation, fabricated reviewer identities and misleading claims about journal indexing are incompatible with research support under these principles. References should be included because they are relevant to the research.
Institutional assessment should also recognise the limits of indicators. Publication and citation data can inform judgement, but need context about discipline, coverage and research contribution. The ALLEA research integrity framework is one cross-disciplinary reference for responsible practice; it is not a certification of a consulting service.
06
Responsible use of AI
Any proposed AI use should have a defined purpose, an appropriate assessment of the tool and a person accountable for checking the output. The relevant institutional, funder and journal rules apply to the project.
Client manuscripts, restricted datasets or other confidential material must not be entered into third-party AI systems without the necessary permission and safeguards. Availability of a tool does not by itself establish that it is suitable for the information.
AI-generated statements, code, references and figures require verification appropriate to the task. Relevant use should be documented and disclosed where required. The European Commission's living guidance provides guidance for this discussion, alongside the rules that govern the actual project.
07
Confidentiality and conflicts
The engagement should identify restrictions on access, use and disclosure of project material. Permission to review a manuscript does not imply permission to use it in marketing, training or an unrelated project.
Relevant conflicts should be disclosed before work is accepted. These may involve collaborators, competing projects, reviewers, suppliers or referral arrangements. The parties should agree how a material conflict will be managed or whether the assignment should not proceed.
The confidentiality page explains what should be agreed before detailed material is shared.
08
Questions, concerns and corrections
If a concern arises about the evidence, permissions, contribution or proposed use of material, it should be raised with the project contact. The relevant work may need to pause while the facts and responsibilities are clarified.
A correction should identify the affected output, the nature of the problem and the required action. The project record should retain a clear account of substantive changes. Decisions about a journal correction, retraction, institutional investigation or formal ethics process remain with the appropriate responsible parties.
These principles should be reflected in the project agreement and working arrangements. If your institution or journal has specific requirements, describe them before the work is scoped.