I've got a significant core population (1.3MM), along with the probability of an event occurring within that population (p of a person making a purchase at a specific store across NA vs. purchases made at any store across NA). Because there are so many stores, the p for any specific merchant is extremely low (<0.1%).

I've also got a defined sample population that is very small and cannot be increased (~50 people). Performing binomial analysis is proving to be too sensitive, because every store in the sample of 50 is well outside the normal population with a single event.

Is there another method that could be used that might be more suitable?

Thanks,

What?? In simple english please.

Quote:

---

Originally Posted by tickle

What ?????????? in simple english please.

---

Binomial does not appear to be the best choice for discrte event analysis when the probability of the event is so low. Is there a more appropriate model?