What proportion of users had a primary telephone contact? I think you'd need to be skipping over a lot of users (those without a primary telephone contact) to h

What proportion of users had a primary telephone contact? I think you'd need to be skipping over a lot of users (those without a primary telephone contact) to hit the pathological case that's implied.