From what I gathered, the difference is that, urban kids (why is it urban kids?) relate to animals by relating them to humans; whereas elsewhere, animals are related to other animals.
To quote the original article, "inferences are asymmetric, with inferences from human properties to mammals emerging as stronger than inference from mammals to humans, and (3) children’s inferences violate their own similarity judgments by, for example, providing stronger inference from humans to bugs than from bugs to bees (Carey 1985;1995)." emphasis mine.
If this is what the submitted link is talking about, I don't believe that's anthropomorphism.
