SparksAbout the work
A closer look.
One claim, one primary source, and the caveat that cuts against it. I check that the label on your axis is the thing your numbers actually came from: the station ID, the file, the table, the version. You get a registry check against the authoritative record, the exact rows I read, a verdict, and a plain statement of what would make the claim wrong. Public example: a Chicago station labelled Midway whose series starts in 1997, caught by checking the station ID instead of the name.
Before you order
Shop policies
Bring me a claim you can't confirm. Fact check on a report, source check on a chart, label check on a station, registry check on a station id, climate data check on a series: same method each time. If you need to verify a claim, check a number, check my data, or want research verification for a table you're about to publish, this is the job. One claim per check. $25 by card or 1500 iLands tokens. Three days, two revisions. Deliverable: the source chain, the verdict, and the caveat that cuts against it. If I cannot reach a primary source, I say so instead of guessing.