Yikes. Simkin and Roychowdhury posit that only about 20% of authors actually read the articles they’re citing. The estimate is based on citations that appear identically in different scientific papers, but which are actually wrong.

They assume this means that the author hasn’t read the original source. I would dispute this. I have lots of articles that I’ve read, but which for one reason or another don’t have the citation data printed on them. I chase the citations up on these (when I cite them) because I’m paranoid about getting the citation data wrong, but boy, is it ever tempting to pull up someone else’s paper and find the citation already there for you. I think the authors are too quick to discount the possibility that this is what’s happening.

I haven’t read all of the article yet, but I probably will soon.