Projected New Senate: 48 Democrats 52 Republicans
News from the Votemaster
Three Senate polls in hotly contested states today, all good news for the Democrats. In Tennessee, Rep. Harold Ford, Jr. (D) is in a statistical tie with former Chattanooga mayor, Bob Corker, with Ford leading 43% to 42%. Considering that they are fighting to replace Sen. Bill Frist (R-TN) who is retiring to run for president and that Tennessee is largely Republican, Ford is doing very well to be in a tie.
In Montana, state Senate president Jon Tester (D) continues to outpoll incumbent Sen. Conrad Burns (R-MT), largely due to Burns close ties to disgraced lobbyist Jack Abramoff. Tester has been leading since the middle of August.
In Ohio, Rep. Sherrod Brown (D), who has been leading incumbent Sen. Mike DeWine (R-OH) all year, is still slightly ahead, but his lead has been cut to an insignificant 45% to 43%, a statistical tie.
Over in the House, New Mexico attorney general, Patricia Madrid is now in an exact tie in NM-01 against incumbent Rep. Heather Wilson (R-NM), at 44% each. A lot of mud has been slung in this campaign. Wilson is also having to spend a lot more effort defending her vote for the war in Iraq, which Madrid has repeatedly raised as an issue.
Since there is no other major news today (although I will keep an eye on FL-16, where more news is leaking out every day), this is probably a good time to discuss polling methodology. Probably the least understood but nevertheless most overrated term in polling is "margin of error." If Smith polls 53% and Jones polls 46% with a margin of error of 3%, what this means is that there is a 95% chance that Smith is favored by between 50% and 56% of the voters and Jones is favored by between 43% and 49% of the voters. Thus Smith is (barely) ahead by more than the margin of error.
But does that really mean much? Unfortunately, no. Margin of error has to do with how much error is introduced in the polling results by the fact that only 600-1000 people are polled. With bad luck the pollster might happen to get too many Democrats or too many Republicans in the small sample (although they try to correct for that). If the pollster asked every registered (or possibly, likely) voter in the state, the margin of error would be 0, but the results might still be highly inaccurate.
In general, statisticians distinguish between precision and accuracy. Precision has to do with having a small sampling error; accuracy has to do with actually determing the true value you are trying to measure. The classic example is to ask a million people the height of the Emperor of Japan. You will get a very precise (i.e., reproducible) answer down to a fraction of a millimeter, but it won't be very accurate because most people don't know the height of the Emperor of Japan.
Another classic example is the 1936 Literary Digest poll that predicted Republican Alf Landon would crush Democrat Franklin D. Roosevelt. They sent 10 million mock ballots to addresses taken from telephone books and got back 2.4 million replies, an immense sample, with a margin of error less than 1/10 of a percent. But it turned out that anybody rich enough to have a telephone at the bottom of the Depression was probably a Republican. The sample was anything but random.
The PublicMind poll at Fairleigh Dickinson University in New Jersey ran a wonderful controlled experiment in July that illustrates the problem. They asked 1/3 of the respondents the following (simplifed) questions in this order:1. Do you believe the country is going in the right direction or the wrong direction?
2. Would you rate the job the President is doing as good, fair, or poor?
3. Do you approve of the way President Bush is handling his job?
4. Is the military effort in Iraq going very well, fairly well, not too well, or not well at all?
5. Did the U.S. do the right thing going to war in Iraq or was it a mistake?
Then they asked them whether they were going to vote for the Democrat, Bob Menendez, or the Republican, Tom Kean, Jr., for senator.
A second 1/3 was asked questions 1-3 before the question on the Senate race and the final 1/3 was asked about the Senate first, then the five questions above.
The results are significant. In groups 2 and 3, Menendez and Kean were virtually tied. But in group 1, Menendez held an 8% lead, In other words, just bringing up the subject of Iraq, causes a NJ state senator (Kean) who had nothing to do with the Iraq policy, to tank. While this result suggests a strategy for the Democrats (talk about Iraq all the time) and one for the Republicans (hide the fact that you are a Republican), it clearly illustrates our point that methodological issues often dominate in polling. To really understand what a poll means, you have to look at the exact questions asked, the order they were asked in, and fair number of other issues.
The media rarely talk about these things, but they are crucial in understanding why different contemporaneous polls often give very different results: it may depend on what other questions were asked before the one reported. Other key factors are whether the poll was conducted by a human interviewer (e.g., Gallup, Research 2000), a recorded voice (e.g., Rasmussen, SurveyUSA), or over the Internet (e.g., Zogby), what day of the week and time of day the call was made, and other factors. Conclusion: margin of error deals with errors introduced by the small sample used, but says nothing about the (far more important) methodological issues. For more on polling, see the Polling-FAQ above.
Projected New House*: 216 Democrats 218 Republicans 1 Tie* Where no independent polls exist, the 2004 election results have been used. See complete House polls.
See the details of the Senate and House races with photos, maps, links, polls, etc.
If you like this site, please announce it to news groups and blogs and tell your friends about it. If you have your own blog, please click on "For bloggers" above.
-- The Votemaster