Posted on 11/06/2006 1:00:10 PM PST by carl in alaska
There are major, serious problems in the methodology of these national polls about control of Congress that make most of these polls useless as predictors of which party will control Congress after this election:
1) A meaningful poll cannot ask this question at a national level, because control of Congress will not be decided in a national election. Control of Congress will be decided in a large number of separate House and Senate elections. Control of Congress is decided in much the same way in which control of the White House is decided through a series of separate state by state votes for electors to the electoral college. The presidential election is decided in the competitive battleground states, and in the same way control of Congress is decided only in the competitive House and Senate races.
2) When these polls are done incorrectly at a national level, voters in districts and states where incumbents are certain to win by large majorities throw off the results by adding large numbers of responses for Republican or Democrat from voters in races that are not competitive. In reality, only the voters in districts with competitive House races will have any impact on who controls the House, and only voters in states with competitive Senate races will decide who controls the Senate. When combined with point 1 above, it's clear that national polling on this question is meaningless.
There's one and only one correct way to use polling to predict control of Congress:
A) To be meaningful, each House race and each Senate race must be polled separately and each race must be predicted separately.
B) Then the predicted numbers of House representatives and Senators from each party after the election are added up to determine which party controls the House and which party controls the Senate.
That's the only meaningful way to do this kind of prediction, and you'll find this methodology used at the top political websites. National polls that combine all voters together are useless and misleading.
So I think the best thing to do is assume that the GOP has a slight edge in retaining control of the Senate and that control of the House is about a tie right now. At this point it all depends on turnout, so please get out there and vote and prove that these national polls on control of Congress were all wrong.
I'll wrap this up with a quote from Phil Collins:
"...So wipe off that grin. I know where you been. It's all been a pack of lies."
Freepers! Get thee to a voting booth!
Thanks and ping.........
Just get out and vote.
The only poll that matters is the one tomorrow night or wednesday when the results are in.
Thanks and ping.........
Don't forget the poll taken on the 8th, when Democrats are supposed to vote. ;)
ooops.
Almost forgot!
:0)
You're welcome. This national polling about the congressional elecion has bothered me for a couple of weeks and it's time to straighten this out so that Republicans are not deceived by meaningless polls.
I heard there was a poll this Tuesday.
I think I trust that one, and I will even participate in it!
Today during a Rush Limbaugh station break for news the cBS news was talking about the polls in particular races and mentioned the race that talent's in right now and the cBS announcer declared it was "still too close to call".
Anybody else hear that?
yes, the generic polls are meaningless in any consideration of a race by race battle - but they are useful for spotting national trends, if you choose to ignore the abolsute numbers, the trendlines are still worth something.
One of the things political scientists like to do, though is to work out some algorithm that will translate the "generic congressional ballot" into actual electoral results at the district level.
I guess there's a trade-off involved. It would take a lot of work to estimate the winners in each district and you still wouldn't be certain. The generic congressional ballot is a much cruder measure, but it's what pollsters and political scientists have close at hand ready to work with.
More here.
Nope I did not hear that but that is true. MO is still close.
The only reason for the Rats polls is to depress GOP vote turnout. When they keep shoving polls in our faces showing Rat candidates ahead by two digits in "safe" red territory over and over again 24/7, it is bound to have a discouraging effect on conservatives. We just have to ignore the news pundits and the polls and watch Animal Planet. It works for me.
Its called "An Election."
It's hard to believe, after the media's deceit in the 2000 election that anyone puts any credibility in any poll from a major MSM outlet. They try to interfere with every election. Always have, always will. Republicans should be used to this by now, and the mantra from the RNC should be to ignore what the media says and VOTE!
Of course they're meaningless. The pollsters purposely survey more of those who agree with them, creating an artificial liberal majority. The only poll that matters is the one tomorrow.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.