pageok
pageok
pageok
Extreme Measures:
From the LA Times:
But behind the elaborate preparations and gung-ho attitude, the television networks are heading into the night with a sense of cautious restraint -- especially when it comes to exit polls -- all too aware of the implications of a botched call.

"We don't want ever a repeat of what happened in 2000," said Phil Alongi, executive producer of NBC's special events, referring to the networks' haste in awarding Florida to Al Gore, then giving the state to George W. Bush before realizing it was too close to call.

"We learned so many lessons across the board," he said. "One of the first: Get it right."

To do so, the networks now follow strict rules that govern projections, examining not only exit poll data but actual vote tabulation and turnout information. NBC -- which keeps its decision desk isolated from the calls made by competing networks -- will only call a winner once its statisticians conclude that the chance of an error is less than 1 in 200. And no calls will be made until all the polls have closed in a state.

Extreme measures are taken to ensure that early data from the exit poll does not leak out, as it did in 2004, when the first wave of surveys showing John Kerry in the lead rocketed through cyberspace.

For much of the day, only a small group will have access to the exit poll, which is being conducted by Edison Media Research and Mitofsky International for the National Election Pool, a consortium of the networks and the Associated Press. Three members from each outlet will be sequestered in an undisclosed location in New York, where they will analyze the results of questionnaires filled out by 100,000 voters nationwide. Their cellphones and BlackBerries will be taken away until 5 p.m. ET, when they will be allowed to share the data with their newsrooms.

It's the same procedure that was used in the 2006 midterm elections and effectively prevented the release of incomplete data, much to the relief of network executives.

"Exit poll information in the hands of trained professionals is perfectly fine," said Sam Feist, CNN's political director. "Exit poll information in the hand of the general public, who may not understand what it means or stands for, can be dangerous."

That's because exit polls are designed to provide a demographic portrait of voters, not to predict the winner of a close race. The early waves of data can be especially misleading because they do not necessarily reflect an accurate sample of the electorate.
Mhoram:
And yet, somehow, I'm sure that all of this proves a liberal bias.
11.4.2008 3:55pm
cboldt (mail):
"Exit poll information in the hand of the general public, who may not understand what it means or stands for, can be dangerous." That's because exit polls are designed to provide a demographic portrait of voters, not to predict the winner of a close race.


Exit poll information in the hands of the public is a bad thing, because it isn't an accurate predictor, yet disparate polls (some of which are necessarily NOT accurate predictors) are okay.
.
Polls are a means to transform the public, not to inform it.
11.4.2008 3:57pm
TruePath (mail) (www):

"Exit poll information in the hand of the general public, who may not understand what it means or stands for, can be dangerous." That's because exit polls are designed to provide a demographic portrait of voters, not to predict the winner of a close race.


Voting in the hands of the general public, who may not understand the the issues or what policies a canidate stands for can be dangerous. That's why only social scientests should be allowed to vote.

WTF? dangerous?!?! That doesn't even make sense. What are the voters going to do with it, give each other paper cuts? Could early info give voters incorrect perceptions and even influence the outcome of the race as a result? Sure, but that is the essential nature of any communication to the voters. What makes polling data any differently.

I mean suppose CNN decided not to release it's economic data because we might make bad choices or did anything like this in some other area than election law. Rather than backing them up most people would be outraged.

-----

Of course whether I'm outraged or not depends on whether early release of the data benefits the candidates I think should win or not.
11.4.2008 4:40pm
A. Zarkov (mail):
This site gives graphs, tables and analysis so you can make predictions on who will win from early returns from key states. I guess this information could be "dangerous."
11.4.2008 4:46pm
EH (mail):
And yet, somehow, I'm sure that all of this proves a liberal bias.

Absolutely, just like the removal of two Republican poll workers in Indiana for using party affiliation to challenge voter eligibility.
Indianapolis Star:
The removal of two Republican election workers from a Warren Township polling site - for using improper methods to challenge voters' rights to cast a ballot - has prompted local Republican Party leaders to issue a statement of regret.
The story goes on to tell us that this was just a couple of bad apples.
11.4.2008 4:59pm
Opher Banarie (mail) (www):
"...sequestered in an undisclosed location in New York..."

Do we know if they mean New York City, or someplace like Rochester?

Is the room lead-lined to prevent snooping? Was it checked for wiretaps? Wow, this could be more secure than that place Dick Cheney vanishes to....
11.4.2008 5:07pm
Sarcastro (www):
A. Zarkov

Dangerous?
11.4.2008 5:08pm
Barbara Skolaut (mail):
Here's a free clue for the broadcast media: There's no need to predict who will win any election based on a survey of a few voters. The only reason the networks do this is to impress themselves by pretending to "scoop" their competitors. (There's also the possibility of influencing the election in the western half of the U.S. by predicting the winner will be the party the "news" people want to win, thereby hoping to discourage the members of the other party from bothering to vote, but let's not go there right now.)

The only poll that really counts is the one coming out of the voting booths. I sincerely wish the networks would keep their freakin' mouths shut about "how it's going" UNTIL VOTING IS FINISHED.

And no, I'm not holding my breath. >:-(
11.4.2008 5:12pm
EH (mail):
I sincerely wish the networks would keep their freakin' mouths shut about "how it's going" UNTIL VOTING IS FINISHED.

I actually wouldn't mind seeing a law about this. Unconstitutional, probably, but such is dreamland.
11.4.2008 5:23pm
cboldt (mail):
-- I actually wouldn't mind seeing a law about this. --
.
Not commissioning then talking about polls would sure leave a substantial void to fill with something else. Probably car chases and missing blonds.
11.4.2008 5:28pm
Splunge:
To do so, the networks now follow strict rules...

Ah. Those many layers of editing and fact-checking that, as we all know, make the mainstream media so much more reliable and objective than the wild-eyed amateur upstart operations in the blog-o-sphere.

I am vaguely reminded of stories of Hitler in the Kanzlerei bunker in 1945, carefully planning where to deploy military units that had long since ceased to exist except on paper.
11.4.2008 5:45pm
KG2V:
Heh - Want some news from the inside - there isn't any data released to the various "decision desks" until 5:00pm eastern. I actually run the computers that read the data from AP and EMR for one of the networks - but I don't look at the actual data (I could, but what I don't know, I can't tell) I do know that folks were seriously upset in 2004 - and even the leaks in 2000 and 2002 annoyed them. Back circa 2000, the guys who call the races had TVs to watch the other networks - no more. I know in 2000 the newsroom folks put presure on the desk to call the race, now, they are isolated from annoying them. Even us IT guys are isolated - I'm actually 4 floors away from the folks who call the races, and the only time I'll actually see them is when I go to get dinner in about 30 minutes
11.4.2008 5:51pm
John Moore (www):
Probability of 1 in 200? What are they smoking.

This is the same sort of thinking that led to the Credit Default Swap part of the financial meltdown.

The probability is only meaningful within the assumptions made by the pollsters and their statisticians. It is not a true estimate of the likelihood of success.
11.4.2008 6:24pm
tsotha:
That's because exit polls are designed to provide a demographic portrait of voters, not to predict the winner of a close race. The early waves of data can be especially misleading because they do not necessarily reflect an accurate sample of the electorate.

What a load of crap. Exit polls are designed to demoralize one's political opponents, like most polls in politics, and in general they work pretty well for that purpose. We know how to do relatively accurate polling - I don't see how anyone could look at the 2004 exit polls and think the networks were trying to get it right.
11.4.2008 6:31pm
Toby:
In my fantasy on how to improve elections, there is the following:

1) If you can't vote for the candidate, you can't donate. No out-of-state PACs for in-state races.

2) All funds listed on the web before deposit. Exceptions require pay back triple - once to the person who finds it, twice to the local designated agency. Pick the home for stray cats or local schools - I don't care.

3) No polls in the news. With nothing to talk about, car chases and missing blondes may predominate, but occassionally the press might report on the issues instead of the horse race.

Then set free the dogs of war...

Sociological research? Please.
11.4.2008 6:46pm
MikeDT (mail):
Exit polls are designed to demoralize one's political opponents, like most polls in politics, and in general they work pretty well for that purpose.

Do you really believe this? Do you think they're just designed by Democrats, or do both sides do this?
11.4.2008 11:07pm
David Schwartz (mail):
Exit polls have several huge problems. The biggest one is selection bias. In the context of this election, Obama supporters are much more likely to be vocal about their support, so they were likely over-represented in the exit poll data. This mean inferences like "Obama was ahead in the exit polls, so he'll be ahead in the official results" are dangerous.

In the hands of a competent statistician, however, these kinds of things are very easy to correct for. For example, suppose you have official data from only 5% of the districts. You see a slight lead for someone, but you can't call the election with only 5% of the districts.

Now, imagine you also know that the exit polling data (regardless of what it is) is substantially identical between the districts that have reported and those that have not. That can easily be the difference between being able to call a contest immediately and not being able to call it for some time.

This is precisely how exit polling data was used in calling this election. The raw results were never used (by at least the major networks) to suggest that official outcomes would be anything like those results. (Though, surprisingly at least to me, they pretty much were.)
11.5.2008 6:23am