Thursday, December 08, 2005

Fun With Statistics

Among other logical "issues," distractions, extraneous issues, and name-calling, one poster to the comments here thinks Bush has reversed his trend in poor ratings. Now reverse would suggest more than a point (or two or three) fluctuation. Given the impressive nose-dive Bush has taken, reversal (of fortune, so to speak) suggests he's regained his former position. The commenter's spin either 1) demonstrates how to lie with statistics, 2) illustrates his optimism, or 3) depends on parsing what the meaning of the word "reverse" is. A point or two or three? Maybe he's thinking that the act of shifting gears is enough.

So, for fun the reader can look at a summary of polls here, the long term change of some of these polls (not selected or parsed by me) here, and the one the reader cites here. For those who'd rather cut and paste, the links are below.

For example, in the Rasmussen poll, on December 7th the favorability rating was indeed 45%. But look at Dec. 5th (48%). Take a look at Nov. 21 (45%). Yes, things appear up a bit from Nov. 17th (43%). But a majority still disapprove of the performance of George W. Bush in this poll. Reversal of fortune? On the right are daily trackings which, not surprisingly ('cause that's what tracking polls do) actually show fluctuation with Bush up or down the past few weeks. The Bushies are all so desperate that they think they can convince the majority of Americans who are not buying what Bush is selling that we are wrong (not) by using polls, intimidation, . But you've gotta admire their optimism.

One other note: The blog bullies try to make anyone who comments against government policy as not in the mainstream. Guess what, they'd be wrong. Using one poll on any particular day or even comparing one poll to two points in time doesn't give the complete picture. I haven't even discussed here what voters think of Bush's specific policies. And that's a sight to be hold (and fodder for another column). For example, even in New Hampshire, a poll showed that "Almost 60 percent of voters polled last week indicated they believed the war in Iraq was not worth fighting and 53 percent said they felt misled about whether Iraq had weapons of mass destruction." But I get ahead of myself. But I wouldn't claim that one data point means all that much. Trends suggest what they suggest. And we infer that the Man has a few problems with the electorate.

The reason the poster and others persist in their spin is they think they can convince folks there's a train taking off to happy Bushland and that we all should jump aboard. I think there's a better chance of me riding a motorcycle than that (and that's not saying much.) And I doubt the rest of the public will be buying the Bush-BS this time around. No matter how many (or few) approve of Bush's performance, that data doesn't erase the mismanagement and deception.
Summary of Polls

Over Time

Rasmussen Poll cited by the reader:

New Hampshire Poll