Claims with numbers

Where the trick usually hides.

Questions to ask

Compared with what? Out of how many? Measured how? A percentage without a base, or a change without a starting point, is not information.

Relative risk sounds dramatic; absolute risk is what affects you.
A doubling from one in a million to two is still nothing.

Base rates

How common something is beforehand decides how to read a test or a signal. Ignoring it is why a highly accurate test for a rare condition mostly produces false positives.