Methodology

Sourcing

We cite the primary release. For research, that is the paper or preprint and its dataset; for markets, the exchange or filing; for the economy, the statistical agency. Secondary coverage may point us to a story, but the figure in our copy is traced back to its origin and linked in the source list.

Statistics

• Every quantitative claim states its sample size and observation window where the source reports them.
• We distinguish correlation from causation explicitly, and we report effect sizes, not just significance.
• Point estimates carry their uncertainty — confidence intervals or standard errors — whenever available.
• We do not invent, round misleadingly, or strip a number from the caveats the authors attached to it.

Verification

A claim reaches publication only after the underlying figure has been located in the primary source and re-read in context. Where a result is contested, we say so and link the competing evidence. Data-heavy reports are checked by a second analyst before publication; those carry a peer-reviewed marker and a named reviewer.

Freshness

Each article shows a visible publish and updated date, mirrored in its structured data. When new data supersedes an old figure, we revise rather than republish, and record what changed.

Limitations

Data describes; it does not decide. Samples can be unrepresentative, methods can bias results, and a single observation of a noisy process can mislead. We flag these limits inline rather than implying more certainty than the evidence supports. Reporting is not investment, medical or legal advice.

Corrections

Found an error in a number or its source? Write to [email protected]. We correct openly and promptly. See our about & ethics page for how the newsroom operates.