causal
Multi-Factor Attribution Models Lack Causal Validity
Multi-factor attribution models used by digital platforms are association-based, so they cannot establish causation and are at best indicative, which is why advertisers apply haircuts to their numbers.
causalpending
Speaker
CliffordEvidence Quote
“association is not causation, and you can be as sophisticated as you want in your econometrics, but at best it's sort of indicative.”
Created: 11/16/2026, 6:56:55 PM
Your notes
Loading notes...