A confident sentence is not the same as a well-supported one. Readers should be able to tell where a claim came from, what it actually establishes and what remains unknown. That is the starting point for our editorial standards.
Know what kind of claim you are reading
A manufacturer specification tells you what a company states about a product. A museum record documents an institution's research about an object. A government safety page provides public guidance. Each can be useful, but they answer different questions.
A hands-on observation requires someone to have actually examined something. A test result also requires a defined procedure, recorded conditions and an explanation of limits. Neither can be supplied by a persuasive product description or a generated image.
Make the route to the source visible
Our educational articles link to the sources supporting their factual claims. When writing about history, an attribution such as “the museum identifies” is often more precise than pretending that every historical question has a settled answer. When discussing storage, we identify official guidance and avoid turning it into a product endorsement.
A source should support the specific sentence beside it. A long list of impressive references at the bottom of a page does not repair an unsupported claim in the text. Readers should be able to follow the chain of reasoning without having to guess.
Be explicit about the limits
We have not published hands-on product reviews. There are no test scores, performance rankings or lab certifications on this site. “Lab” is part of the publication's name; it is not a claim of accreditation or an existing test facility.
If we publish testing in the future, it will need a clearly described method, product provenance, dates, conditions, observations, limitations and relevant disclosures. Until that work exists, leaving the review shelf empty is more informative than filling it with unsupported verdicts.
Images are not evidence
Our atmospheric images are AI-generated editorial illustrations. They do not show products tested by GunReviewLab and should not be used to identify a model or assess its condition. Words and images should tell the same truth about what was—and was not—done.
How we approach evidence →
