John,
If you’re an experienced A/B tester, you know you can’t simply put up a control and treatment and pick a winner based on the results. Because the better numbers for the winner may simply be random chance.
So you need a big enough sample size to ensure statistical validity.
However, sometimes too many samples can also deceive. This is the validity threat from oversampling.
Read on to learn more about the occasional downside of big numbers.
Daniel Burstein
Senior Director, Content & Marketing
MarketingSherpa
P.S. Our job is to help you do your job better. Reply to this email to let us know how we can help. |
|
