What is the best way to test the difference in means for two samples of data that follow a Tweedie distribution? (zero inflated, tweedie power parameter ~1.7)
The data represents the number of purchases made by users on an online website, where most visitors don’t make a purchase, hence the zeroes.
I would like to calculate the p-value and confidence intervals.
Is a t-test appropriate? How about a permutation test or bootstrapping?
My data is very large (several million users), so a method that would be computationally efficient would be desirable, although I don’t mind computation intensive processing if it would result in more sensitivity/accuracy.