Weighing the pig: the mishandling of reference statistics