establish causality between the change and the effect
Start here: Statistics for A/B testing
Guilherme Coelho
616
This is incorrect. A significant P-value establishes that the OECs are statistically different but does not establish causality. This is a very common mistake.