Richard Spencer @RichardBSpencer
You're assuming that, since 28% of this not-at-all-random sample is infected, 28% of the population at large is infected. That is not sound reasoning. A much small percentage of the population is likely infected. — PolitiTweet.org