Method
How our content is made
Rigor is not a promise, it is a protocol. This page describes the path every piece of content published by ODERSA follows, from the first fact-check to a public correction.
It applies to the pages of this site just as much as to the training cases in Before You Believe It (Avant d’y croire), and it exists for a simple reason: a reader, a teacher, or a funder should be able to know where what they are reading comes from, and check it without taking our word for it.
From idea to publication
From idea to publication
Check before writing
Every fact, every figure, every definition is checked against public sources before it goes into a piece of content. Anything that cannot be verified is rewritten as a generic example, or dropped. Content that cannot be checked does not get published, no matter how compelling it looks.
Write for real readers
The language is calibrated to the intended audience: short sentences, vocabulary explained, concrete examples. Training cases are realistic and entirely fictional: never a real brand, a real person, or a real organization inside a made-up example.
Cross-review
No content is approved by the one person who wrote it. An independent review checks three things: the facts (do they hold up?), the clarity (would a reader in a hurry understand it?), and the language (each language version is reviewed on its own terms, not as a translation).
Run automated checks
Before every publication, automated checks sweep the entire site: no broken links, display contrasts calculated to the WCAG AA standard, no leftover work-in-progress markers. These checks are blocking: if even one fails, nothing gets published.
Publish, then account for it
Every published figure comes with the public address where anyone can recount it. The data downloads freely, in an open format. And when an error is reported, it gets corrected, and the correction is public.
The rules that do not move
The rules that do not move
Open license
Everything is published under CC BY 4.0: reusable, adaptable, translatable, with a simple source credit.
Completely free
No paid content, no account, no ads, on any of our sites.
Zero collection
No cookies, no trackers, no third-party requests: privacy is built in, not just declared.
Multilingual from birth
Languages are planned at the design stage, never bolted on afterward.
What we do not do
- No estimated “beneficiaries” and no figures inflated by calculation: our sites carry no measurement tool, and the audience figures we publish are counted by our host, on its own side, with their date and their method.
- No announcements: a resource appears on this site once it is online and verifiable, not before.
- No real misinformation reproduced: training cases are fictional and realistic, so people learn the mechanisms without spreading actual misleading content.
- No real people or real brands in the examples: teaching the mechanism does not require putting anyone on trial.
- No jargon for show: a sentence that means nothing to a reader in a hurry gets rewritten.
If you find an error
Every page on our sites carries a “Report an error” link. Tell us what is wrong and where: we check it, we fix it, and the correction is public. An outside report is one more check on our work: it is how content gets better.
Report an errorTo go further
To go further
Our commitments
Commitments kept by design, each paired with a way to check it yourself.
Read our commitments →Impact & figures
Public counters, dated, with the address where you can recount each one.
Recount our figures →License & reuse
What the CC BY 4.0 license allows, and the attribution line to copy.
Reuse our content →