The Republicans cheated again--and still lost!

[Skeptic]

But your link does not support claims of "observation bias", but rather my explanation: the "self-biasing sample":

"The most likely source of this error is differential non-response rates for Democrats and Republicans..."

It gives no evidence --- nor even mere opinion --- that the discrepancy is caused by the pollsters.

"Observation bias" doesn't mean the observers are biased. It simply means that sometimes, the very act of making an observation biases the results. The pollsters caused the discrepancy simply because they asked people who they voted for, which is a question that democrats are more likely to answer.

 

[Darth Rotor]

Er ... yes it is.

A smaller proportion of Republicans reveal their voting preference than Democrats. This is what the figures show, and what the article you cited actually said.

There are a lot of people in America who vote, but who are not registered as Republicans or Democrats. (People like me) So categorizing people as voters by "Republican" and "Democrat" is not as accurate as categorizing the folks being voted for in those terms.

Kinky For Governor in 2006
Ended with Friedman on a ranch in the styx
Kinky for President in 2008
For VP Willie Nelson -- now that would be great.

DR

 

[senorpogo]

So the Republicans rigged the election, but were only smart enough to give themselves enough votes to both lose the election and get caught cheating?

 

[Dr Adequate]

There are a lot of people in America who vote, but who are not registered as Republicans or Democrats. (People like me) So categorizing people as voters by "Republican" and "Democrat" is not as accurate as categorizing the folks being voted for in those terms.

For the purposes of this discussion, which is about the 2006 elections, could we agree that by "Republican voter" I mean someone who voted Republican in the 2006 elections?

Or do I have to come over there?

 

[Dr Adequate]

..or that they were as likely as Democratic voters to speak to someone in their own demographic, but the pool of poll workers was skewed to a demographic that votes Democrat.

A good point.

I'm not saying that's the only explanation (I suspect a variety of factors in play) but rather that it is incorrect to ascribe an untested motive.

Thus far, you have speculated about their motives, and I have not.

 

[Dr Adequate]

"Observation bias" doesn't mean the observers are biased.

No? OK.

Personally, I think it is useful to use two different phrases for two completely different things, but have it your way.

However, even if we agree to make a gray soggy amorphous mess of the English language, I was talking about the pollees, while Cylinder was talking about the pollsters.

He may well have a point, but we are in fact referring to different phenomena.

 

[Skeptic]

No? OK.

No, it doesn't. Obervation bias simply means that the observed mean (or variance, etc.) might sometimes be different than the actual mean (or variance, etc.) because the sampling from the population isn't random (in a statistically well-defined sense of "random").

There's nothing in this definition, either in theory or in practice, that implies that the reason the sampling is not random is because the observers are psychologically biased toward some result. Observation bias can occur (and often does) even when the "observer" is an inanimate object, such as a camera.

However, even if we agree to make a gray soggy amorphous mess of the English language,

We're not. "Observation bias" is a technical term which has a well-defined meaning, which I outlined above.

 

[Dr Adequate]

Encore.

Personally, I think it is useful to use two different phrases for two completely different things, but have it your way.

However, even if we agree to make a gray soggy amorphous mess of the English language, I was talking about the pollees, while Cylinder was talking about the pollsters.

He may well have a point, but we are in fact referring to different phenomena.

You are right in saying that the term "observation bias", as a technical term, covers both what he was talking about, and what I was talking about, but they are still two different things.

 

[Ziggurat]

Can't we just agree that the original poster is an idiot who doesn't even have the guts to come back and defend his post or concede his mistake?

I'm trying to be a uniter here

 

[Dr Adequate]

It would be a bit of a broad brush to call PL an idiot, but yes, the OP is idiotic, I think we all agree.

Let's all ****!

---

What, none of you is female?

Okay, you all make love and I'll just sit here and, er, actually, now I think about it, not watch.

I have a book and a sandwich. I'm happy.

 

[Darth Rotor]

For the purposes of this discussion, which is about the 2006 elections, could we agree that by "Republican voter" I mean someone who voted Republican in the 2006 elections?

Or do I have to come over there?

You could, if you could show me that voters confine their vote by party affiliation. Come on over in any case, I'll cook you up some chili, get out beer (Shiner Bock) and regale you with stories of LBJ and dead men's ballot boxes.

I see your point, but a lot of us vote the person, not the party.

I voted for a Democratic judge in one district, two Democrats for county jobs (I know one personally) Kinky for governor (Independent), against Hutchinson for Senator(R) (so for a Libertarian), a Republican for state rep, and Republican for my district rep to US Congress. (He lost, his last name isn't Hispanic enough to win.)

So, what am I, Doc: Democrat, Republican, Libertarian, or Texan?

DR

 

[brodski]

You could, if you could show me taht voters confince their vote by party. I see your point, but a lot of us vote the person, not the party.

which is a major difference between the US and the UK, here people almost exclusively vote party and not personality, which may be why Dr A has chosen to couch things in the terms he has.

 

[Cleon]

Perhaps the demographics of republicans suggests not that they are reluctant to admit, but rather are reluctant to be engaged.

And you'd think with all the talking points about how important the family is, "defending" marriage against teh gay, etc., you'd think Republicans would be more interested in getting engaged.

 

[Bill Thompson]

As for the first time they "cheated": a supreme court decision is not cheating.

 

[Dr Adequate]

You could, if you could show me that voters confine their vote by party affiliation. Come on over in any case, I'll cook you up some chili, get out beer (Shiner Bock) and regale you with stories of LBJ and dead men's ballot boxes.

I see your point, but a lot of us vote the person, not the party.

I voted for a Democratic judge in one district, two Democrats for county jobs (I know one personally) Kinky for governor (Independent), against Hutchinson for Senator(R) (so for a Libertarian), a Republican for state rep, and Republican for my district rep to US Congress. (He lost, his last name isn't Hispanic enough to win.)

So, what am I, Doc: Democrat, Republican, Libertarian, or Texan?

DR

Okay. For the purposes of this discussion, could we agree that by "Republican voter", we mean someone who voted Republican in the 2006 Congressional elections? Since that is what we're talking about?

Or do I have to come over there with a huge enormous gun and a squadron of killer paragliding weasels?

You know what I mean, dammit. I mean what the exit pollsters meant when they asked the question discussed in the OP. This is what we're talking about.

Oh, WTF. Unleash the paragliding weasels!

 

[ref]

So the Republicans rigged the election, but were only smart enough to give themselves enough votes to both lose the election and get caught cheating?

Here is a hypothesis what could explain that claim.

It would be even more obvious rigging to give a landslide win for the Republicans, given all the polls favoring the Democrats before the election.

The more convinient way to do it without getting caught would be exactly that. Rig the election by the minimum expected amount of votes needed to win, not to cause some very strange surprise result that contradicts all the polls.

What I'm saying is, 'Hacking Democracy' clearly showed that the voting machines and their results can easily be hacked by one single person. Which of course gives any malicious vote counter in the right place of hierarchy all the power in the world. So in theory it can be done. There are machines that are made by different manufacturers, so if the hacking is done with only one manufacturers' machines, you would have to predict the results from the other manufacturers' vote machines to give you the amount of rigged votes needed to win. Bad prediction, bad result. You lose the election.

I guess that is what is meant, when talking about the rigging of 2006

 

[Skeptic]

Here is a hypothesis what could explain that claim.

It would be even more obvious rigging to give a landslide win for the Republicans, given all the polls favoring the Democrats before the election.

The more convinient way to do it without getting caught would be exactly that. Rig the election by the minimum expected amount of votes needed to win, not to cause some very strange surprise result that contradicts all the polls.

So if the Republicans win, they fact that the elections' results contradict the polls prove they cheated. But if the Republicans lose, the fact that the election results did not contradict the polls merely prove the Republican cheated in a secret, easy-to-hide manner.

In other words, regardless of who wins, viola! : we have "proof" Republicans cheated.

 

[ref]

So if the Republicans win, they fact that the elections' results contradict the polls prove they cheated. But if the Republicans lose, the fact that the election results did not contradict the polls merely prove the Republican cheated in a secret, easy-to-hide manner.

In other words, regardless of who wins, viola! : we have "proof" Republicans cheated.

It was a hypothesis. If you would do something like that you would probably be doing it that way.

The given results did not contradict the polls before elections. But some argue the exit polls still contradicted the given results.

There is no proof wrongdoing. Only those exit polls some are talking about.