I am running A/B testing . Created 4 groups with 3 group as variant and one as Control.
I found the confidence values for this hypotheses
More Number of events(Group) per day Drive Higher Uninstalls
| Group | Uninstall | Base | UninstallRate | Delta | Confidence |
|---|---|---|---|---|---|
| Group A | 508 | 33704 | 1.51% | 0.04% | 0.310 |
| Group B | 514 | 34052 | 1.51% | 0.04% | 0.293 |
| Group C | 545 | 33814 | 1.61% | -0.07% | 0.515 |
| Group D | 523 | 33854 | 1.54% | 0.00% | 0.000 |
Now I want to know how can we statistically conclude with confidence that number of events(Group) DOES NOT impact uninstalls?