A/B Testing

It is experimentation with two different groups
- One group is the default one (control)
  - This group is important because we want the possible agent of actions on this group and see which agent is make impact on treatment group. In that way, we see that is not other things, we did not know, is creating impact on treatment group
- The second group is the treatment group (which is the hypothesis you want to test)
A sample going to one of this A/B test
- This sample can be a person, seed, web visitor
- It is important to be randomized
Verify the different on the treatment group and analyze the impact
Always choose a metric/test statistic before the A/B testing to not create bias on the analysis
Hypothesis Tests