Attached are the results of a three-day poll of 800 likely voters in Wisconsin on various political issues conducted by Strategic Vision, LLC on September 14-16, 2007. The
poll has a margin of error of ±3 percentage points. For the poll, 272 (34%) identified themselves as Democrats; 216 (27%) identified themselves as Republicans; and 312 (39%)
identified themselves as Independent or other party affiliation.
When Republicans were asked their choices for President in 2008, the results were New York Mayor Rudy Giuliani 28%; former Tennessee Senator Fred Thompson 24%; Arizona Senator
John McCain 8%; Massachusetts Governor Mitt Romney 7%; former House Speaker Newt Gingrich 6%; Texas Congressman Ron Paul 3%; former Arkansas Governor Mike Huckabee 3%; Colorado
Congressman Tom Tancredo 2%; Kansas Senator Sam Brownback 1%; California Congressman Duncan Hunter 1%; and 17% undecided.
I don’t think Thompson has even been in Wisconsin yet. This race is between Giuliani and Thompson, and Giuliani is losing.
Submit to Stumbled Upon!
Although I am really getting tired of politics, Fred Thompson has given me a bit of hope. I expect him to be the leader within a couple of months. Please can we get rid of the professional politicians (of all stripes). Maybe it’s time to bring back the farmers…
Now, I’m no pollster, but isn’t 216 a really, really, really small sample to draw any conclusions from?
Nope, the odds that this poll will produce an answer more than 3 percentage points different from the result of a much larger but otherwise identical poll is 20-to-1 against. I could repeat that a few more times if it would make it more convincing.
If polls are so accurate why do we need elections where as many of the great unwashed as possible are encouraged to express their utterly ignorant and often deranged opinions?
I’m familiar with statistics, but I still think JJ has a point. Subgroups in these polls don’t have their MOE adjusted, even though their sample size is much smaller. That’s why you should always be wary of when a poll says “out of black voters” or something similar. Black voters will only be about 10% of the sample size and the MOE gets a whole lot bigger.
I love it when people declare “Giuliani leads Thomson 28 to 24…Giuliani is losing!!!!!!”
That’s just weird.
Who are you, jack?
You’re not me.
I would have known that due to the MOE, Guiliani and Thompson are neck and neck in Wisconson.
Fred benefits from name recognition - Tommy Thompson was the Governor here for years. Fred, Tommy, what’s the difference? Thompson is Thompson.
Seriously, I’m from, and reside in, Wisconsin. Go Fred!
Your assertion that Rudy is “losing” would be more persuasive if Rudy has been camping out in the Badger state, but he seems to be concentrating more on Iowa, New Hampshire, South Carolina and Florida for some reason. Rudy did visit London yesterday when you wrote this — I hope you are aware that is also not in Wisconsin.
Malvolio is right on. The number to focus on the the total sample size of 800. That is a very reasonable number for estimating the make-up of the total population. Keep in mind that most of the national surveys you see have sample sizes between 1000 - 2000.
Yeah, but from what I hear about his campaign appearances, Fred may peak before he shows up.
As I noted in my post, I don’t think Thompson has ever campaigned in Wisconsin.
Giuliani is “losing” (sneer quotes) because, after months of campaiging and years of reasonably favorable name recognition, he is tied (within the margin of error) with Fred Thompson about two weeks after Thompson entered the race.
The total sample size is irrelevant. Only the subgroup size (216 Republicans) matters for the purpose of comparing Republican candidates. And for that very small sample, the Margin Of Error at the 95th percentile level is 6.67% for fairly evenly split questions. (You can calculate it yourself at various websites, such as http://americanresearchgroup.com/moe.html.) The mathematics are a little more complex in the case of multiple candidates who split up the pie more widely, but the MOE is still not substantially lower.
Bottom line: The only thing this tiny poll shows is that Rudy and Fred are in first and second place (although not necessarily in that order) and all the other candidates are in third place or lower.
If you try to overanalyze the results beyond that, you’re just fooling yourself.