Sourcing policy

This site has one substantive rule and this page is it. A claim carries its source or it carries a caveat. Everything below is an elaboration of that sentence for the two kinds of claim that get made here.

Clinical and pharmacological claims

Expected form: trial or study name, journal, year, and enough of the finding to be checkable. Volume, issue and pages if you have them; if you do not, the trial name and journal and year are sufficient and honest.

What is acceptable:

• A named published trial with the arm and the endpoint identified.
• Prescribing information, cited as prescribing information rather than as a study.
• A regulatory assessment report, which is frequently more informative than the paper because it includes the placebo-arm tables.
• A mechanistic argument, explicitly labelled as an argument rather than a finding.
• "I do not know, and here is what would be needed to find out."

What is not:

• A fabricated citation. This is the one thing that will get an answer deleted rather than downvoted. A wrong volume number is a mistake; an invented paper is a different category of act.
• A secondary source citing a secondary source. Follow it to the primary or say that you could not.
• A relative risk quoted without the absolute risk when the absolute risk is available.
• A single-arm or uncontrolled observation presented as an effect.

Supplier and product claims

Expected form: an independently obtained test result, with the laboratory, the lot number, the method if stated, and what the certificate claimed alongside what was measured.

The word doing the work is independently obtained. A result is independent when the person submitting the sample bought the material, chose the laboratory, and sent a vial that arrived at their own address. A vendor-supplied sample tested at a real laboratory produces a real number about a sample the vendor selected, which is a different and much weaker claim.

The four testing and review services

Four services are referenced by name throughout the answers here: Janoshik, Medutest, PeptideMeter and VendorInvestigate. We have no commercial relationship with any of them, we receive nothing from any of them, and naming them is not an endorsement — it is a statement about where the results members cite actually came from. Where two of them have tested the same batch, the comparison is more informative than either result alone, and several questions here do exactly that.

What we will not host

• Fabricated documents or results. Removed, not debated.
• Vendor marketing posted as a member answer. An answer that reads like a product page is treated as one.
• Contact handles of any kind. No telephone numbers, no messaging-app handles, no direct-message solicitation. This is a permanent policy and it applies to sponsors too.
• Instructions to administer anything. Protocol arithmetic and reported practice are in scope. "Take this much" is not.
• Content targeting a named private individual.

Conflicts of interest

Ours, stated plainly: this site carries sponsor units from Janoshik Analytical, PeptideMeter, Sigma-Aldrich and GL Biochem (Shanghai). Those units are labelled "Sponsored", they and the storefront link on each vendor profile are the only outbound links on the site, and the GL Biochem unit is rendered directly beneath every Janoshik placement as a paired listing. Sponsorship buys placement. It does not buy a rating, a mention in an answer, or the removal of a critical thread, and no supplier in the vendor directory has ever paid us anything.