Originality
Similarity at or below the limit (20% overall, 5% from any single source, references excluded); no duplicate or redundant publication.
Evidence: similarity report (e.g. iThenticate)
The criteria a scientific, high-quality paper must meet — from research integrity to presentation.
Science in Quality is an accreditation system that certifies whether a scientific paper meets defined standards of scientific quality and research integrity. It is not an institution: it is a set of criteria, a scoring method and a review procedure.
Seven gatekeeping requirements are pass/fail — failing any one means no accreditation, whatever the score. The scientific quality and presentation criteria are then scored and weighted, and the total decides the accreditation level.
A paper must pass all seven before it is scored. These protect the integrity of the accreditation itself.
Similarity at or below the limit (20% overall, 5% from any single source, references excluded); no duplicate or redundant publication.
Evidence: similarity report (e.g. iThenticate)
Ethics committee approval for studies on humans or animals; informed consent where people take part.
Evidence: approval number and committee name
No fabrication, falsification or inappropriate image manipulation; raw data available on request.
Evidence: signed integrity declaration
Every author meets the ICMJE criteria; no guest or ghost authors.
Evidence: CRediT contribution statement; ORCID iDs
All financial and personal interests and all funding sources declared.
Evidence: declaration section in the paper
Any AI use in writing, analysis or images is disclosed; AI is never listed as an author; authors take full responsibility.
Evidence: AI-use statement
No citation manipulation or coerced citations; self-citation limited to what the work needs.
Evidence: reference list, checked by the reviewer
These ten criteria decide whether the paper adds reliable new knowledge.
New knowledge, a new method or new data; the research gap and the paper’s contribution are stated explicitly.
A clear, specific and answerable question; hypotheses where the design calls for them.
Current, relevant and critical coverage of the field; the work is positioned against the state of the art.
A design suited to the question, with enough detail to repeat it: materials, equipment, parameters, software and versions, controls or baselines.
Correct methods with their assumptions checked; uncertainty reported (error bars, confidence intervals, error analysis); comparison with benchmarks.
A data and code availability statement; data shared in a repository where possible, following the FAIR principles.
Complete, clearly presented and consistent with the methods, including negative or unexpected results.
Results interpreted against prior work, with mechanisms explained; no claims beyond the evidence.
Limitations stated honestly, with their effect on the conclusions.
Conclusions follow from the results; theoretical or practical implications and future work are named.
These ten criteria decide whether readers and reviewers can understand, check and cite the work.
Accurate and specific, without exaggeration.
Background, aim, method, key results with numbers and the conclusion, within the word limit.
3–6 terms that a researcher in the field would search for.
IMRaD (Introduction, Methods, Results and Discussion) or the field’s accepted equivalent.
Clear, concise and grammatical English; terms and abbreviations defined at first use.
Legible, axes labelled with units, self-explanatory captions, at least 300 dpi; each one cited in the text.
SI units, consistent symbols, a nomenclature list where many symbols are used.
Complete and accurate, one consistent style, primary and recent sources, DOIs where they exist.
The checklist that fits the study type is followed (e.g. CONSORT for trials, PRISMA for systematic reviews, ARRIVE for animal studies, STROBE for observational studies).
Author names, affiliations and ORCID iDs; corresponding author; funding, acknowledgements and data statements in place.
A paper that passes all gates gets a weighted score out of 100, and the score decides its level.
The author submits the paper with the evidence for G1–G7. Any failed gate ends the review with “Not accredited” and a written reason.
At least two reviewers score S1–S10 and P1–P10. Reviewers declare any conflict of interest before they start.
Where two reviewers differ by more than 1 point on a criterion, a third reviewer scores it.
Scientific quality counts 70% and presentation 30% of the final score out of 100.
S̄ and P̄ are the mean scores of S1–S10 and P1–P10.
Total ≥ 85 and no criterion below 3
Total ≥ 70 and no criterion below 2
Total 60–69, or one criterion at 1; the author revises within a set period and is re-checked
Total < 60, any S criterion at 0, or any gate failed
The complete Science in Quality criteria as a printable PDF.
ACCEE assistant
Automated assistant · answers from this website · a person on WhatsApp
Some answers are written by an AI model (Claude, by Anthropic) from ACCEE’s official information, so please do not type personal details. Those questions are saved without your name or IP address. Privacy