Open imweijia opened 7 years ago
Has there been any progress or discussion on this feature? I also think that would be very useful.
To revive this discussion, what were peoples thoughts on how assignment would work in situations where the threshold N was reached and individual alternatives do not meet the required sample size to reach statistical significance?
Assuming the algorithm chosen has an element of randomness I imagine a solution would be to define threshold N per alternative and have the experiment end once all alternatives reach the specified minimum.
Just wondering if this can be configured as an option whereby the test can be paused when N sample size of participants is reached?