Relative Without Absolute
Twice as likely as what?
The idea
A percentage change is uninterpretable without the underlying rate it changed from.
In the real world
Doubling a risk means very different things starting from 1 in 10 and 1 in 100,000.
Going deeper
A relative change is accurate and carries no information about magnitude. Doubling a risk describes both a serious change and a negligible one, depending entirely on where it started.
This is the standard mechanism by which a small effect is reported as a large one without anything false being said. The single number that resolves it is the baseline rate: doubling 1 in 10 matters, doubling 1 in 100,000 does not in any practical sense. Where a report gives only the relative figure, that omission is usually doing work.
Where it stops applying
Relative measures are the appropriate form for comparing effects across populations with different baselines. The problem is presenting one without the other, not using them at all.
Why it matters
It is the standard way a small effect is reported as a large one, entirely truthfully.
Try this today
When you see a percentage increase in risk, look for the baseline number.
Test yourself
A headline reports that something doubles a risk. What single number do you need before deciding whether to care?
Show the answer
The baseline rate. Doubling 1 in 10 is a serious change; doubling 1 in 100,000 is almost nothing in absolute terms. The relative figure is accurate in both cases and carries no information about magnitude on its own.
Learn this in the feed Answering from memory, then again days later, is what makes it stick.