In These Times

fyi Klein is right: 4 is larger than 1. But that has nothing to do with the validity of findings in Sergio BenDixon's polls in the 1 per cent range. Margins of error near the boundaries of 0 and 100 are treated quite differently than the margin of error near the midpoint of 50. (because -3 could never be a possible answer). Klein might want to look that up since it shows such a lack of basic understanding of statistics that it suggests the rest of his observations are about as accurate.

Posted to Trending Toward Inanity