Does the data distribution itself need to be normal? I’m not sure but I think that the sampling distribution of the data needs to be normal. Correct me if I’m wrong. :)
Hey Lily, you are correct, the sampling distribution is the one that needs to be approximately normal. In my case, the samples were the conversion rates of the two groups of users (pre CTA and post CTA). Hope that clarifies it :)
Does the data distribution itself need to be normal? I’m not sure but I think that the sampling distribution of the data needs to be normal. Correct me if I’m wrong. :)
Hey Lily, you are correct, the sampling distribution is the one that needs to be approximately normal. In my case, the samples were the conversion rates of the two groups of users (pre CTA and post CTA). Hope that clarifies it :)
In the same vein, how would you test the equal variance assumption for a proportion, like conversion rate?