Dr. Peto’s Response to an Editor’s Request for Statistics

“The patients were subdivided into 12 … groups according to their medieval astrological birth signs.” When the tongue-in-cheek zodiac subgroups were analyzed, Geminis and Libras were found to have no benefit from aspirin, but the drug “produced halving of risk if you were born under Capricorn.” Continue Reading →

Mitt Romney and Paul Ryan Will Win in a Landslide

My methodology is simple: read all the polls, rip them up and throw them away. Read all the editorial endorsements, tear them up and throw them away. Listen to my gut instead. Yeah, I know it’s not scientific, but most polls have a “refuse to answer” rate around 90%. That means their numbers are based on a sample that is far from random. Continue Reading →

Twitter a Polling Alternative?

This is typical media stupidity. Tweets are public unless you work very hard to keep them private. Most people on Twitter do not reveal much about their personal lives. There are many fake personas, too. Our cat has a Twitter feed (made up mostly of other household pets and other animals). People routinely lie. People will also lie to a telephone pollster, but at least they don’t have the incentive of their response being visible to the entire world. Continue Reading →