We should be careful discussing the Iowa results

I have seen several times discussions that Obama beat Hillary amoung Democrats. We actually don't know if he did or he didn't. The 'results' are from entrance polls which have a margin of error, just like any other poll. So Obama 'winning' 32 to 31 is really a statistical tie. He clearly won the caucuses of that there is no doubt. But we will never know the true number of voters in each camp to a moral certainty. This isn't an anti Obama thread, though I publicly have expressed my dislike of him. It is simply a math primer so to speak. We need to be careful with these breakdown numbers from polls.

And the truth gets buried somewhere along the way. Whether Obama won among Democrats or not is irrelevant at this point. Well, it is, but whether he did or didn't isn't the question anymore.

The perception is that he did, so he did.

And no, people make up own mind, and in this day and age of information access, if you blindly believe anything anyone tells you (well, except my Dad--he's the most honest person in the world) then you are the problem, not the media.

winning Iowa was enough for Obama to get the nomination barring Hillary winning NH or running the table on Super Tuesday. Who won Iowa's Democrats likely will have nothing to do with it. But as a math teacher, and a statistics teacher in particular, I do care that we have numerate people. Numerate people don't confuse polls of a sample with the actual numbers of a population.

Deal with it.

incidently I don't have a candidate. Had Edwards, Richardson, Biden, Dodd, or Kucinich won Iowa I would have been happier than I am today and in all but Kucinich's case happier than if Hillary won. I do care, when supposedly educated adults, make fundamental mistakes about basic statistics. Innumeracy is a prime reason Repubicans are able to win elections with idiotic platforms. Numerate people would know better than to believe tax cuts increase revenue.

Spin on.

But, one must understand how the Iowa caucuses work. To me, they sound like a cluster f**k, but what do I know.

I think most people know the caucus gives supporters of candidate who got less than 15% to pick a second choice. What they may not know is that the precincts do not report (or pick delegates) based on the actual number of supporters, but on percentages. I don't totally understand it, but it almost sounds like a popular vote vs. electoral vote kind of thing.

When the media reports percentages, they may be looking at results of polls (popular vote) with no consideration to how the Iowa caucuses are actually run.

Am I with you?

Obama won Iowa by one delegate, in a race that requires more than 2,000 to win.

but a showing of the preference of the electorate as a whole. So Obama had the highest level of support, Edwards second highest and Clinton third highest. That is all.


Everyone trying to make the Iowa Caucuses fit into a primary model is going to continue to have headaches.

(I am no expert in the arcane ways of Iowa Caucus nor am I good at math)

The networks take entrance polls.

The caucuses are held and delegates assigned based on the results of alignment and realignment.

The state party (or the state dept?) calls in the delegate counts to the networks.

The networks then walk it back, accounting for the alignment/realignment and delegates awarded and arrive at their final figures.

This is what I've gathered from reading Pollster.com

But I take your point, dsc, that the vote ID'd as Dem going in may not be trackable (from what we know - perhaps the voter registration changes are somehow followed through the process) in terms of the vote coming out. As one who has been saying Obama beat Hillary among Dems, I appreciate your pointing out it may not be possible to really know this with any certainty.

Obama got the most delegates (snicker) so he won in Iowa.

Those who make it a tremendous victory for Obama are delusional.

I don't deny that, but I do want us to refrain from discussing numbers as being written in stone when they aren't.