It's still surprising that arxiv is that high up. You'd think most people would cite the definitive publications, not the preprints at arxiv? Anyway, given tha

It's still surprising that arxiv is that high up. You'd think most people would cite the definitive publications, not the preprints at arxiv? Anyway, given that in some fields _everything_ that's written goes to arxiv, arxiv will [by definition][1] have a very high h-index. The thing is, the h-index was conceived to compare individuals, not journals. [1]: http://arxiv.org/abs/physics/0508025 http://arxiv.org/abs/physics/0508025