Comments on: Election 2004 By The Numbers

By: Jay Reding

Jay Reding — Thu, 21 Oct 2004 15:25:45 +0000

I’ll admit Ohio was one that trended away from Bush against the previous trendlines. States can and do swing towards a candidate then swing back.

And as I’ve said before, the poll of the hour is only a signal of larger trends, which is why I’m using the RCP averaging rather than polling data.

And BTW, using the raw polling data yields similar results. Using the raw polling data indicates a closer race, but still shows Bush slightly increasing in support and Kerry decreasing slightly over time. However, the r-squared values for Bush were only .22 and the Kerry r-squared values did not show any kind of significant correlation (I believe is was something like .06). This is why I didn’t include them here and why using raw polling data is too statistically incoherent for analysis.

By: The American Mind

The American Mind — Thu, 21 Oct 2004 15:15:52 +0000

Trends
The trend in national polls favors Bush. However, that doesn’t really matter since the Electoral College is picks the President….

By: Mark

Mark — Thu, 21 Oct 2004 14:38:12 +0000

The polls that appear to have a genuine methodology, be it NBC, Rasmussen, Zogby, TIPP or the new Marist poll show the race dead even or an insignificant +1 for Bush. Most of the polls showing Bush leading tend to oversample Republicans, particularly Gallup. The only poll that concerned me this week is the Fox News poll. As messed up as the network is, their polls have traditionally been reasonable. However, their internals seemed questionable, so I’m cautiously dismissing it is a bad sample. Mason-Dixon seems to skew right, particularly this election sample. They very well might be right, but I’m yet to be convinced.

The bad news for Bush is that most polls indicate Kerry with a comfortable lead in the battleground states. The new Marist poll shows the race tied nationally, but with Kerry leading by eight points in the battleground states. Meanwhile, domestic issues are still cited as the main concern among battleground state voters, which benefits Kerry.

Your predictions of the race thus far have been anemic. Just a month ago, you ridiculed me for insisting Ohio was still a battleground state, despite Bush’s uptick in the polls. You were and continue to be convinced that the poll of the hour is indicative of how voters will feel on November 2, as evidenced by your repeated embarrassing aboutfaces on Ohio. I have always been less certain of this election’s outcome and am not ruling any scenario out. As it stands now though, I’d give Kerry 55-45 odds.