Tuesday, September 25, 2007

A-B TAB Learns Exit Polling From Enron's Florida Unit

The A-B TAB decided to send their reporter out for a little bit of exit polling today, which they posted at 6:47 pm:
TAB reporter Rich Cherecwich recently headed to the polls for the evening and has done a survey of (so far) 10 people. The votes break down like this: 3 votes for Alex Selvig; 3 each for Ciommo, Glennon and Hanlon, and 1 for Schofield.
Hmmmm. They appear to have counted as well as Enron's accountants. And they posted exit poll results more than an hour before the polls closed -- bad practice.

Let me warn anybody against believing these results: the margin of error for a sample of ten -- assuming they were, in fact, counted correctly -- is 32%. All of the candidates have statistically consistent results for their survey, i.e., within the error bars of their tiny survey, it is a dead heat.


Oh, and the TAB keeps coming at it with their exit poll results (at 8:01 pm):
Rich just talked to five more voters at Jackson Mann — four voted for Schofield and one for Selvig.

According to our extremely unscientific poll of 19 people, this puts Schofield and Selvig in the lead with 6 votes total for Schofield and 5 for Selvig. Ciommo and Hanlon each got three of those 19 and Glennon got 2.
Amen to the "unscientific poll" statement!

The statistical error bar on a sample of 19 is 4.3, therefore Glennon's two votes are statistically indistinguishable from Schofield and Selvig's six each. Still a dead heat.

No comments: