Do you have any data showing that religious people deviate only slightly from non-religious people?