Fact-Checking Policy
What verification means on this beat
Fact-checking a medical cannabis story is a specific craft, because the two raw materials of the beat, scientific studies and program regulations, both fail in characteristic ways: studies get inflated on the way from journal to headline, and rules get garbled on the way from statute book to social media. The method of MMJ GAZETTE is built around one habit: verify at the source, every time. This page describes what that means in practice. Errors that slip through are handled under the corrections policy, and reports of them are welcome at support@mmjgazette.com.
Research claims: the journal, not the press release
When we report a study, the reporter reads the study. Not the university’s press release, not another outlet’s summary, the peer-reviewed paper itself, and where the paper is paywalled we obtain it before writing, not after. Verification covers:
- What was actually measured. The claim in our article must match the paper’s own results section, not its most optimistic quote.
- Design and scale. Randomized trial, observational study, animal model, or cell culture; how many subjects; over what duration. Our copy states this, because a reader cannot weigh a finding without it.
- Peer-review status. Preprints and conference abstracts are identified as not yet peer reviewed whenever we cover them at all.
- Funding and conflicts. Who paid for the study and what the authors disclose. Industry funding does not disqualify a study, but readers hear about it.
- Context. Where a finding contradicts the weight of prior evidence, the article says so rather than presenting it as a lone revelation.
Preliminary work is labeled as preliminary. That is a hard rule, not a style preference. Early-phase findings, small samples, animal and laboratory results are named as such in the copy, near the finding, not in a footnote. The related editorial rule, no miracle-cure framing, lives in our editorial standards; between the two rules, a promising mouse study cannot become a human cure by the time it reaches a headline here.
Program rules and legal claims: the issuing authority, not the echo
Statements about what a medical cannabis program allows, which conditions qualify, possession limits, caregiver rules, reciprocity, or the legal status of a product are verified against the official source: the statute or regulation as currently in force, or the program authority’s official publications. Secondhand summaries, advocacy pages, and forum posts are leads, never sources. Because rules change frequently and differ by jurisdiction, every legal claim in our copy is tied to a named jurisdiction and reflects the rule as of the article’s date; unattributed, jurisdiction-free claims of legality are exactly the errors this policy exists to prevent.
Quotes, numbers, and people
Quotes are checked against recordings or transcripts. Numbers are traced to the document they came from, and derived figures are recalculated rather than trusted. Names, titles, and affiliations are confirmed with the person or an authoritative record. Market and business claims from companies are treated as claims by an interested party and attributed as such, not laundered into the site’s own voice.
The check before publication
Verification happens before publishing, not after complaints. Every article is reviewed by a second person on the team before it goes live, and the review reads the sources, not just the sentences. Claims that would earn heightened scrutiny anywhere earn it here: allegations of wrongdoing, health claims, recall details, and anything a patient might act on. On fast-moving stories, the rule is to publish less rather than verify less; a gap in an article is honest, a guess is not.
What we do when sources conflict
Sometimes the journal, the regulator, and the company all say different things. We do not pick the most dramatic version. The article reports the disagreement, attributes each version to its source, and says which source we consider authoritative and why. When we cannot verify a widely reported claim, we either omit it or report, explicitly, that it is unverified.
The limits of the method
Verification establishes what the best available sources said at publication time. It cannot see a future retraction, a rule change next month, or fraud inside a lab. When the record shifts, articles get updates or corrections under the corrections policy, with the two labels kept distinct. Readers who check our work against the sources we cite are doing exactly what we hope readers do.
