Worth a read:
http://seosherpa.com/seo-experiments/
Thanks.
Any that strike you as... improbable?
Maybe that the image alt text seemed to count more than text links in #12.
#10, We've been saying here for a long time that nofollow links are worth getting, so that's not really a surprise, but its good to see it verified out there.
#14, First link bias just seems like sloppiness/laziness in the algorithm.