Editorial Methodology
Not every claim needs the same kind of evidence. The best source for today's price is the official order page. The best source for whether an ingredient reduces gum inflammation is a peer-reviewed clinical trial. The best source for a regulatory claim — like whether something is "FDA approved" — is the regulation itself, not a marketing page. This page explains, concretely, how we match claim to evidence across every page on this site, so a reader can judge our work rather than just trust it on faith.
Our Research Workflow
Before a claim is published or updated on this site, it generally goes through the same sequence:
- Define the exact claim. "Supports gum health" and "treats gum disease" are different claims requiring different evidence, even though they sound similar.
- Classify the claim type. Is it commercial (price, packaging), regulatory (FDA status), scientific (ingredient mechanism), or experiential (what a specific product is like to use)?
- Find the most direct available source. A systematic review beats a single study; a single human trial beats a lab study; a lab study beats no evidence at all — but each is labeled for what it actually is.
- Check date and independence. Is the source current? Is it independent of the seller, or produced/funded by the brand itself?
- Check for a match on route, dose, and outcome. A study on xylitol chewing gum doesn't automatically transfer to a xylitol mouthwash or a xylitol chewable tablet — format matters, and we try to say so explicitly rather than blur the distinction.
- Record the limitations. Small sample size, short duration, animal or in vitro model, single study rather than a replicated body of evidence — these are noted, not hidden.
- Write a conclusion no stronger than the source. If a study shows a laboratory effect on isolated bacteria, we don't present that as proof the finished consumer product works in a real mouth.
- Recheck related pages for consistency. When new evidence changes our understanding of an ingredient, we look for every other page on the site that references it.
Source Hierarchy for Scientific Claims
When multiple types of evidence exist for the same question, we generally weight them in this order, from strongest to weakest:
- Systematic reviews and meta-analyses of multiple randomized controlled human trials
- Individual randomized controlled trials in humans
- Non-randomized human studies or small human trials
- Laboratory (in vitro) or animal studies
- Mechanistic reasoning without direct testing
Testimonials and personal anecdotes are not treated as clinical evidence anywhere on this site, even when we include them to describe a genuine customer experience with using the product day to day. The full version of this ranking system, with worked examples, lives on our Research Library page.
How We Tag Claims
Internally, and increasingly in how we word our published content, we think about claims in five categories:
| Tag | What it means |
|---|---|
| Verified Fact | Directly supported by a reliable, checkable source (e.g. "no dietary supplement is FDA approved" — a regulatory fact) |
| Seller Claim | Attributable specifically to the brand or manufacturer, presented as such rather than as independently confirmed |
| Independent Evidence | Backed by research from a source unrelated to the seller — a university, government agency, or independent journal |
| Limitation | A noted gap between the evidence available and the claim being made (e.g. ingredient-level evidence used to describe a finished multi-ingredient product) |
| Unverified | Insufficiently supported by current evidence — we say so rather than omit the caveat |
The Evidence Escalation Audit
Marketing language has a natural tendency to drift upward in strength over time — "may support" quietly becomes "supports," which quietly becomes "proven to." Before anything is published, we check whether the wording has become stronger than what the underlying source actually says. A study showing a reduced bacterial count in a petri dish should never be summarized on this site as "kills the bacteria that cause gum disease in your mouth" — that's a real, common type of overreach we specifically check for and correct.
Zero-Trust Review of Our Own Past Content
We don't assume a page we published previously is still accurate simply because it was accurate when we wrote it. When we update a page — or when new research becomes available — we re-audit it as if we were reviewing someone else's work for the first time, rather than defending our own prior conclusions. This is why pages across this site carry a "Last updated" date, and why our Research Library is maintained as a living document rather than a one-time reference list.
Related Editorial Pages
This methodology works together with two other pages that explain how we handle evidence and commercial interest: our Sources Policy, which covers how we select and reject specific sources, and our Affiliate Disclosure, which explains the firewall between our editorial content and our commercial relationships.