(cross posted from dailykos)
I’m a statistician. I belong to the American Statistical Association. In the latest issue of Amstat News, they have called for a do-over in the race in the 13th CD of Florida
more below the fold
I will use the analysis presented in AMSTAT News together with other information; for example, AMSTAT News refers to ‘Candidate A’ and ‘Candidate B’; I’ve substituted the actual names and parties.
Florida’s 13 CD was open; it has previously been held by the infamous Katherine Harris. The Republican was Vern Buchanan, and the Democrat was Christine Jennings. Buchanan has claimed victory. Jennings has not conceded defeat.
There were approximately 240,000 votes cast. Buchanan got 50.08%, and Jennings 49.92%- the difference was 401 votes. In Amstat’s words “arguably too close to call, even if only the usual errors were present”
But, as they point out, more than the usual errors were present. In Sarasota county (which uses touch screens) there were 18,000 votes not counted – 13% of the votes cast. In other counties in the 13th CD, undervoting ranged from 1% to 5%. Many voters in the Sarasota County reported difficulty in finding the race. The votes that DID get counted in that county favored Jennings by 6%, and, if the undercount had been similar to that on absentee ballots in Sarasota, Jennings would have picked up 830 votes, and won the election.
The short article concludes
We conclude that we don’t know who won. Our professional recommendation is simple: Do it over
This was not an editorial, this was a summary of the findings of the ASA’s Scientific and Public Affairs Advisory Committee
