According to RealClearPolitics, there have been three polls conducted in the last forty-eight hours in Iowa---the Des Moines Register, CNN, and Zogby's tracking poll. Here are the specific results from each poll:
Des Moines Register -- 32% Huckabee, 26% Romney (800 LV surveyed)
CNN -- 28% Huckabee, 31% Romney (373 LV surveyed)
Zogby -- 29% Huckabee, 25% Romney (928 LV surveyed)
Now if you were to combine all three polls into one poll, you would find that out of 2101 persons surveyed, 712 respondents supported Huckabee, while 656 supported Romney. Thus, in effect, 33.88% Huckabee, 31.22% Romney.
Given the size of this augmented, combined poll, the margin of error would be about 2% points. Here is how we come up with the margin of error. Take Mike Huckabee's number (33.88%) and multiply it by (1-33.88%). You get 0.2240. Then divide this number by 2101. You get 0.000106. Take the square root of this number, and you get the standard deviation of .0103. Multiply the standard deviation by 1.96. Voila! A roughly 2% margin of error. (Actually, 2.024% for Huckabee and 1.981% for Romney.)
Accordingly, the actual range extends from 31.86% to 35.91% for Huckabee and from 29.24% to 33.20% for Romney.
Notices that Romney's high-end number (33.20%) is less than Mike' average (33.88%).
If we wish to estimate the probability of Mike's true population average being greater than Romney's high end number, we can use the normal distribution to find this answer. Using the Excel Spreadsheets NORMDIST function, I have estimated that there is a 74.6% chance that Mike's numbers will exceed Mitt's if the election were held today.
MINOR CORRECTION: I failed to factor into the equation the fact that there is a 2.5% that Romney's high end number is greater than 33.20%. Therefore, Mike's chances of winning have been reduced slightly to 72.73% (or 97.50% times 33.20%)
Subscribe to:
Post Comments (Atom)
3 comments:
Nice job Matt.
Now I remember why I hated stat class.
/heh
Cool stuff
I heard you on the conference cal today, and looked up your blog. Reading down, I read the social-fiscal conservative post. I have written about this at http://larryperrault.blogspot.com.
I wrote that social conservatives have always been asked, "Do you want to see the Democrat elected?" And, they have usually gone along.
But I then asked, "If Huckabee is nominated and the shoe is on the other foot, wil the fiscal conservatives go along? I'm not so sure. Democrats too, are up to their necks in corporate money and commitments. And socialistic tendencies favor established financial interest: they will take a tax hike with policies that restrict entry into markets, maintaining/expanding established marketshare.
True conservative/free-market policy like Huckabee advocates is friendly to market entry and innovation for enterprises not yet established. That's why all of this so-called fiscal conservative criticism is nonsense. When he talks about being Main Street and/or small business friendly, and about things like the Fair Tax (which would fairly tax EVERYONE'S expenditures, that's what he's talking about. The established money have teams of accountants to make things tax-free and shift the tax burden downward to the middle class.
Now, which side do you think multi-millionaire investment counselor Mitt Romney, looking for sure investments in a less-dynamic and unpredictable economy would be on? You can almost see the hair of he and other big-money investors curling when the consider the Fair Tax. "Fair," means that they won't be special, anymore.
Post a Comment