The author cites several studies to back up those assertions. Agreed on the positive stereotype bit, though--I think people rely far too much on anecdotal evidence simply because the way that people talk is difficult to distill and quantify.