Talk About Liberal Media Poll Bias! Fox News must have been infiltrated!

[Rainsford]

Thanks for the info.

TBH, I don't understand how you cannot adjust for party affiliation and be accurate.

Fern

My understanding is that unlike pretty much every other factor, there is basically no way to get more accurate data for party affiliation than just asking people and extrapolating. Your overall margin of error for ALL factors should actually be pretty small if you ask enough people, but on factors where we have known good data from other sources (e.g., the census), it's possible to make the sample even MORE representative. For the polls, what would this data be? Many (most?) ask what you consider yourself RIGHT NOW. Voter registration data would get close, but I can't imagine it would be MORE accurate than just asking enough people.

 

[Rainsford]

If your polling is left to whims of chance in terms of the breakdown for Dems/Repubs it cannot possibly be accurate unless by mere coincidence.
...
Fern

That's not really how polling and statistics work. In fact, their underlying math is based on exactly the opposite. A randomly selected sample of sufficient size has an extremely good probability of accurately representing the whole group. "The whims of chance" work surprisingly well if your sample is large enough, because accuracy by "mere coincidence" is actually the most likely outcome.

Look at it from the other side. If the population is 50/50 Democrat/Republican and you randomly pick 1000 people, you're saying that you'll probably end up with some number of Democrats NOT all that close to 500. For that to be true, your random selection process would have to somehow select Democrats more than 50% of the time. For small samples, that can certainly happen, because the 50% number is after all just an average. But pretty soon, it becomes less and less likely that your sample will deviate significantly from the whole population. And the reason is that significant deviation in large samples requires more and more individuals in the sample to be different from the average...which is less and less likely the more individuals we're talking about. 5 Democrats who you'd expect to be Republicans might sound possible, 50 is a lot less likely.

 

[dali71]

You're the reason this forum is a haven for Republican/Libertarian douchebag zealots, aren't you?

Moderator call outs of any kind are strictly forbidden here. Simply put, you may not make reference to a poster's other mod status when they are posting on their personal account.

Perknose
Forum Director

Please just permaban him and be done with it.

 

[Schmide]

You have to admit though that the entire media has a liberal bias when they show Ohio being neck and neck, if not being in Obama's favor.

Maybe reality has a liberal bias from your perspective?

Fox News did it's own poll of who their viewers thought would win the state and it was Romney 90%, Obama 10%.

It was a call in poll dude. Kind of like, if you sample 95% conservatives you will probably get 90% for Romney.

I think that makes it pretty obvious just how liberal every other form of news media in the world really is.

Step out of the bubble dude. Reality is awesome.

 

[nyker96]

whoever did the fox news poll needs to be fired! they need to keep hiring new pollsters until someone can poll a 15 point lead for Romney!

 

[Rainsford]

whoever did the fox news poll needs to be fired! they need to keep hiring new pollsters until someone can poll a 15 point lead for Romney!

The problem with intentionally biasing polls, and another reason the argument that polls are biased against Romney doesn't make a lot of sense, is that we'll all know relatively soon how full of crap a polling organization is. Unless polling results can directly influence results in a predictable direction, biased polls will look pretty dumb come election night.

For example, the guy behind unskewedpolls.com is taking a big chance. It's one thing to post opinionated "analysis" of something when it's hard to prove one way or another. But if the real polls turn out to be accurate on election nigh, the unskewedpolls.com guy will end up looking like a colossal moron. It's the difference between producing propaganda and using your own product. Now some random dude with little to lose can take that risk, but it's hard to buy the idea that polling organizations with reputations to maintain would do the same thing.

 

[Schmide]

the unskewedpolls.com guy will end up looking like a colossal moron.

Nevahar I say nevahar

 

[Throckmorton]

The Fox morning show was going on and on about polls being biased and the "liberal media" trying to create a bandwagon effect. They didn't say a word about their own poll.

 

[Tom]

You have to admit though that the entire media has a liberal bias when they show Ohio being neck and neck, if not being in Obama's favor. Fox News did it's own poll of who their viewers thought would win the state and it was Romney 90%, Obama 10%.

I think that makes it pretty obvious just how liberal every other form of news media in the world really is.

that poll has to be skewed, no way 10% of Fox viewers would vote for Obama.

(ripped from Colbert)

 

[CallMeJoe]

Maybe reality has a liberal bias from your perspective?

It was a call in poll dude. Kind of like, if you sample 95% conservatives you will probably get 90% for Romney.

Step out of the bubble dude. Reality is awesome.

You appear to have mistaken the most blatant parody account in P&N.

 

[Jhhnn]

The Fox morning show was going on and on about polls being biased and the "liberal media" trying to create a bandwagon effect. They didn't say a word about their own poll.

Bandwagon? After the debates, it'll look like this-