Hi all, my first post - couldn't find his specific element covered, hope I'm not covering old ground.
Can the baseline conversion rate not being as expected (in retrospect) affect the confidence level of the conclusion of an activity, and if so, by how much, and how to calculate that?
Here's an example to put it in context:
Supposing I had run an A/B test as follows; 4 experiences including the control, 95% confidence level, 80% statistical power, 10% baseline conversion, 240,000 visitors over 20 days, with a 5% minimum detectable lift.
On conclusion of the test, I find that the baseline conversion was actually 8%. How will that affect the confidence level of the result?
The more I try to figure this out, the less it seems to make sense to me, can someone shed some light for me, please...