Glossary · Methodology
Relative and absolute effect
In short
The same study yields two numbers: the relative change sounds big, the absolute one says what actually happened. Anyone who gives only the relative one has left out half.
If a value falls from 12 to 8, that is four points — or one third. Both statements are correct. The second sounds much better and therefore appears in almost every press release.
With risks, the difference becomes dramatic. If a risk falls from two to one percent, that is a halving — and one percentage point. Out of a hundred people treated, exactly one benefits. The relative number describes the effect correctly, but it does not reveal how many people it affects.
Our rule: if a source gives only the relative number, we look for the absolute one. If it is missing, this is stated in the list of evidence.
A conversion is not a measured value.
Where this comes up in the fact checks
- Tesamorelin – The liver fat study reports an absolute difference of 4.1 percentage points — the catchier relative figure conceals the starting value.
- Omega-3 – Large studies show relative effects whose absolute size in everyday life is small.
- Resveratrol – Many small individual findings do not add up to a big one.
Related terms: Primary endpoint · Surrogate marker
Information only, not medical advice. The BK-Score is a subjective assessment according to disclosed rules – not a scientific evaluation. Scoring rules