posted
I have never been clear on how the networks arrive at their conclusions when they call a state as having been "won" by a presidential candidate when fewer than half of the precincts have been counted. Nor on how meaningful it is when they do so. What's the deal?
Posts: 4534 | Registered: Jan 2003
| IP: Logged |
quote: Using exit polls, the TV networks and the Associated Press employ a range of statistical methods to project a final vote count based on data received from the same two sources:
The Associated Press: The Associated Press is providing vote totals for each race. The AP will gather data from reporters and freelance interviewers based in each county where votes are tabulated.
Edison/Mitofsky Research: To assist the TV networks in collecting and evaluating AP voting data and to conduct exit poll interviews, Edison Media Research (EMR) and Mitofsky International of Somerville, NJ are conducting exit polls. In these exit polls, interviewers ask voters their opinion on a variety of relevant issues, determine how they voted, and ask a number of demographic questions to allow analysis of voting patterns by group.
Using exit poll results gathered at scientifically selected representative precincts, a small percentage of election voting results from the AP, and a number of sophisticated analysis techniques, Edison/Mitofsky recommends projections of a winner for the Presidential race.
posted
they don't use the exit polls at this point in the night. At least most networks I have seen are no longer paying much attention the exit polls (they were off this year) They use statistics looking at the numbers that have come in to decide who is likely to win. They also look at what counties are left, and which party that county is thought to lean towards.
The more votes come in, the easier it is to make the call.
posted
I'm confused because the networks call some states before many votes are counted. For example, right now only 31% of precincts have reported in California, and it's already been called for Kerry. If they're not using exit polls to at least some degree, how can they make that claim?
Posts: 1947 | Registered: Aug 2002
| IP: Logged |
posted
Because it is legitimate to assume (as a first approximation) that the rest of the votes are going to come out with the same distribution. So if 60% of the first half of the votes are Kerry, say, then you can assume with some justification that the next half are also going to be 60% Kerry. Of course, this grows less accurate as the sample gets smaller. If 100% of the first vote cast is Kerry, it is not legitimate to assume that the rest are going to be Kerry.
This is a first approximation, mind. If, say, the rural districts come in first, and you happen to know that urban districts tend to support Kerry, then you are justified in not calling the state for Bush even if 60% of the vote so far is for him. The first approximation mentioned above is in the absence of other information.
Posts: 10645 | Registered: Jul 2004
| IP: Logged |
Printer-friendly view of this topic
![[Smile]](smile.gif)
