How We Evaluate Evidence
Methodology
Every wellness claim on WellnessInsider is graded against a published rubric before it reaches a reader. This page describes the rubric and the sourcing principles that govern it.
The A–D Evidence Grade Rubric
Each claim cluster in an article receives one of four grades. The grade is derived deterministically from the supporting sources attached to that claim — it is never invented or averaged across the article as a whole.
Grade A
Replicated primary evidence (≥2 primary sources)
Grade B
Single primary source
Grade C
Secondary source(s) only
Grade D
Unattributed — no supporting source
Grade thresholds: A requires two or more primary sources supporting the same claim. B is a single primary source. C is one or more secondary sources only (no primary). D is unattributed — a claim with no supporting source attached at all.
Sourcing Principles
Primary sources first
Every health claim is traced to a primary source: a peer-reviewed study, a clinical guideline, or a regulatory statement. Secondary sources (news articles, other wellness media) are used only for context, not as evidence.
Named sources, not link-stuffing
We link to the actual source — the abstract, the guideline document, the FDA communication — not to a secondary article that itself cites it. Reader-checkable at every step.
Contested evidence is named as such
When the evidence is mixed, preliminary, or in active scientific debate, we say so explicitly rather than presenting the strongest supporting study as settled consensus.
No absolute health claims
Phrases like "cures", "guaranteed", "miracle", or "clinically proven to cure" are categorically banned from our content. The validator flags them as hard failures before any article is published.
Editorial Standards
How we select sources, handle AI-assisted drafting, and maintain the review process — the operational complement to this grading rubric.
Read Editorial Standards