Scoring grid versions
The version history of the scoring methodology, what changes at each evolution and why.
The methodology evolves through explicit, numbered versions. Each new version is publicly documented and triggers a recalculation of every catalogue score, so all articles remain comparable. Nothing is changed silently.
Version v1.1 — in force since 2026-05-17
This version corrects several evaluation points to make the judgment fairer, without changing the philosophy of the score (method quality × source reliability).
- The delay between submission and publication is no longer penalized when the information is unavailable: missing data must not be presumed unfavorable.
- The evaluation of the publishing journal (bibliometric impact) has been corrected to better reflect its actual quality.
- Clinical trials overseen by an independent monitoring committee are now correctly recognized.
- Conflict-of-interest detection is more accurate, both ways: it better rewards transparent disclosures and better penalizes serious under-declared conflicts.
- The transparency assessment is adapted to the level of information actually available about the article, so the same limitation is not penalized twice.
- A few articles whose public data had been updated were recalculated.
Observed effect: the catalogue now includes a first level-A article. The rarity of level-A articles remains the reality of a medical science where full methodological excellence (open data, pre-registered protocol, full conflict-of-interest transparency) remains a minority — this is not a flaw of the tool, it is an honest measurement.
Version 1.0 — first public version (since 2026-05-01)
First public version of the methodology. Score = method quality (out of 100) multiplied by an integrity coefficient (0 to 1) reflecting source reliability, with the A / B / C / D / E tiers and the X alert signal.
You can always view a historical score as it was computed with the methodology version in force at the time: each score keeps track of its version.