@Silverfish Another nice example is the product of a Bernoulli(0.6) and a uniform symmetric about 0.
I keep meaning to create an asymmetric mixture of normals with the same property, but I've never gotten around to spending the time to sort it out.
I believe it's possible with a three component mixture but I've not actually done it.
@Andy I asked a question just recently that I thought the econometricians would be all over; it was answered very nicely by an epidemiologist, but I'd welcome additional answers especially since I think there's more to say
Thanks for hinting me to this! I was travelling quite a bit in the last two weeks so this must have slipped me. Sorry for the lengthy answer, it's a very deep question actually which has filled entire books
