Leading indicators · The causal assumption · Element
The check
Whether the link has held when tested against what actually happened.
The part
What it is
If the indicator moved and the outcome did not, the assumption was wrong and the indicator should go. This element is where that gets caught.
Checking it annually is usually enough; checking it never is what allows a broken indicator to be tracked for years.
The smallest level
The attributes it carries
Each attribute is one of the framework’s shared types. What a date is, and how it is written, is defined once for the whole model — the note here says what it means in this particular place.
The other elements in the causal assumption
The claim
The stated causal link between the two measures.
LearnThe evidence
What supports the claim, and how strong that support is.
Learn