Trust & Legal
Research Methodology
Every guide on Learn Dark Web is researched the same way: primary sources first, the questions real users ask, and dated verification for anything that can change. This page explains how that works.
How a topic is researched
Research for a topic happens in four steps, in this order.
- Start with the questions real users ask. Before writing, we look at what people actually search for and ask about a topic — the wording of the question usually reveals the misunderstanding the guide needs to clear up.
- Go to primary sources first. The authoritative source for a fact — the Tor Project for Tor, the National Institute of Standards and Technology (NIST) for standards, the Cybersecurity and Infrastructure Security Agency (CISA) for guidance, CVE.org and the National Vulnerability Database for vulnerabilities, certifying bodies for certification details, and official law and policy pages for legal claims.
- Check the search results and competitors. We read the pages already ranking for the topic, to see what they cover, what they miss, and where they disagree with each other or with the primary sources. A disagreement between sources is a warning, not something to smooth over.
- Verify, then write. Every factual claim is checked against a reliable source before it is published. A claim that cannot be verified is left out or labelled as unconfirmed — it is never presented as fact.
Citation standard
Sources are named, not implied. When a guide states a fact that comes from a specific source, the source is named in the text so a reader can find it. Statistics, fees, counts and dates each need a named primary source. Anything that can change — a law, a country restriction, a fee, a statistic — carries a "last checked" date on the page, so a reader can see how fresh the claim is.
Contested claims — where reputable sources disagree — need two or more independent sources. "Independent" means genuinely separate: two pages quoting the same original survey are one source, not two.
What we do when sources conflict
Sources conflict often in this field, and the conflicts are usually honest: different vintages of the same figure, different methods, different definitions. When we find a conflict, we do one of two things.
- Show both sides. If the disagreement matters to the reader — for example, two widely quoted figures for the same thing, measured in different years — the page says so and gives both, with the dates and the sources attached.
- Omit the claim. If we cannot resolve the conflict and it does not matter to the reader, the claim is left out entirely rather than dressed up as settled fact.
We never resolve a conflict by picking the number that sounds best, or by averaging the figures, or by citing only the source that agrees with the point we wanted to make.
The standards we work to
Our research follows named, public standards rather than our own instincts. Three are worth naming because they shape whole areas of the site.
- NIST SP 800-63B Rev. 4 (final, July 2025) — the National Institute of Standards and Technology's digital identity guidelines. Our password guidance follows this standard: long passphrases over forced complexity, and no routine password expiry.
- FTC Endorsement Guides, 16 CFR Part 255 (revised July 2023) — the Federal Trade Commission's rules on endorsements and material connections. They are the basis of our disclosure rules: any material connection must be disclosed clearly and conspicuously, where readers cannot miss it. See our affiliate disclosure.
- Google Search Central helpful-content guidance (updated October 2026) — our editorial standard for what makes a page genuinely helpful: original, accurate, written for readers rather than for search engines, with a real author standing behind it.
These are standards we cite and follow. We do not claim any affiliation with the bodies that publish them.
What we do not do
- No invented volumes. We do not publish search volumes, traffic figures, market sizes or "how many people" figures unless a named primary source states them.
- No fake metrics. No invented ratings, no invented review scores, no invented rankings of tools or services.
- No unverified claims presented as fact. If it was not checked, it is not stated as fact — it is omitted, or clearly labelled as unconfirmed.
- No fees, earnings or regulator fine figures we could not verify. Certification fees change, bug-bounty earnings are mostly unattributed, and regulator fine figures conflict across sources. Where we could not verify a figure against a primary source, we do not publish it — we link to the official source and tell the reader to check the current figure there.
- No testing on live systems we do not own. We do not scan, probe or demonstrate on any system without ownership or written permission.