Independence is not accreditation
The two are constantly conflated, including by people acting in good faith. They answer different questions and are verified in different ways.
Competence versus interest
Accreditation — most often ISO/IEC 17025 for testing laboratories — attests to technical competence: validated methods, calibrated and traceable equipment, trained personnel, controlled documentation, participation in proficiency testing. It is audited periodically by a national accreditation body and it is genuinely meaningful.
Independence is a question about interest: who owns the laboratory, who commissioned the analysis, who selected the sample, and what happens commercially if the result is unfavourable. No accreditation standard for testing laboratories establishes any of that, because it was never designed to.
A laboratory can be fully accredited and wholly owned by the supplier whose material it certifies. Every number it publishes may be correct. It is still not a third party, and a certificate describing it as one is inaccurate on that point regardless of the analysis.
Verifying the accreditation claim
This takes a few minutes and is the most reliable check available to a reader without laboratory access.
- Take the accreditation number from the certificate. If none is printed, the claim cannot be verified and should be treated as unverified.
- Find the accreditation body that issued it — UKAS, COFRAC, DAkkS, A2LA, ANAB and their national equivalents each publish a searchable register of accredited bodies.
- Search the register for the number. It should resolve to the laboratory named on the certificate, at the address given.
- Open the scope certificate. Accreditation is never blanket: it covers named methods on named matrices. Confirm the method used on your report is inside the listed scope.
- Check the validity dates. Accreditation lapses, is suspended, and is sometimes reduced in scope after an audit.
Step four is the one most often skipped and the one that most often fails. A laboratory accredited for one technique may report another that falls entirely outside its accredited scope, with no misrepresentation on its part — the certificate simply carries less weight than the accreditation logo implies.
Three degrees of distance
Certificates rarely state which of these applies. Where the arrangement is not disclosed, the safe assumption is the second — it is the commercial norm.
In-house
The party being tested performs the analysis. The result may be technically impeccable and the equipment properly calibrated. It is still self-assessment, and should be read as such.
Commissioned
An external laboratory analyses a sample supplied and paid for by the party being tested. This is the ordinary arrangement and it is not disqualifying — but the sample was chosen by the interested party.
Independently sampled
The laboratory or a third party draws the sample from finished units, without the tested party choosing which one. Rare, materially stronger, and almost always documented when it happens.
Questions that surface the arrangement
Each is answerable in one line by anyone who knows, and each is informative whether it is answered or not.
- Which laboratory produced this report, and what is its accreditation number?
- Is there any ownership, investment or exclusivity arrangement between you?
- Who selected the sample that was analysed, and from what population?
- Was the analysed unit drawn from the batch that shipped, or produced separately?
- Are unfavourable results published, and can I see one?
The last question is the most diagnostic. Any real testing programme run over time produces results that fail. A registry in which nothing has ever failed is either very young or not publishing everything it holds.