Relative vs absolute risk
A big-sounding relative change applied to a tiny starting risk is still a tiny risk.
Try it
Choose a baseline risk and a headline percentage and see the people.
Hypothetical unless you press the pill button, which sets 1 in 7,000 (1.43 per 10,000) rising 100%.
■ baseline ■ extra cases. 100 × 100 = 10,000 people (rounded to whole dots).
What is going on
Relative risk tells you the proportional change. Absolute risk tells you how many extra people are affected. Headlines prefer relative figures because they are bigger, but only the absolute figure tells you how worried to be.
The helpful question is always “out of how many?”. Ten thousand dots make the answer visible.
How to spot it
- “Increases risk by X%” with no baseline figure.
- Only relative numbers in a press release.
- Ask: what is the risk without it, and what is it with it, for 1,000 or 10,000 people like me?
The 1995 pill scare
In October 1995 the UK Committee on Safety of Medicines warned that third-generation contraceptive pills carried around twice the risk of venous thrombosis of earlier pills. In absolute terms the risk rose from about 1 in 7,000 women a year to about 2 in 7,000. Gigerenzer and colleagues describe how many women stopped taking the pill, and how the following year saw an estimated rise in unwanted pregnancies and abortions.
The dots in the demo show whatever numbers you set; they are hypothetical, apart from the preset that reproduces the pill figures as described by Gigerenzer et al.