Margin of Error
Charles Stewart is a political science professor at MIT. He’s studied elections for a long time and we talked to him about whether you can determine a margin of error for the general election. We asked him if there comes a point where the margin of error is close enough that the result is the same as flipping a coin.
If you do not see flash audio player please install the latest flash player.
Dec 21 program on WCPN
Interesting program but the comments on margin of error were off target. Polls which you correctly cited as having a 3 to 5 % error are samples of a larger population based on a small subset of the target population. An election is an enumeration,a counting of all the votes cast. While a 1 % error rate might sound good compared to a poll it is not the appropriate comparisonn Check with your NPR statistician to clarify the difference.
Comment by andrew konick — December 21, 2007 @ 3:34 pm |
+0