Another person who doesn't know what oversampling is. This is the problem with morons on ZH, Reddit, etc. running around with content they don't understand.
It was in one of the podesta emails. It helps to explain why Hilary's lead in the ABC/WAPO poll has shrunk from H +12 to H +2 in less than a week when the polling completed before the Huma Danger news broke.
Why would you say it is unrepeesentative? You seem to be asking why polling firms don't change the answers they get to what they or you think they should get. The answer would be because the goal is accurate data.
Remember, polling firms have a vested interest in accurate results while you have a vested interest in results that tell you what you want to hear. That's most likely the source of your confusion.
Thank you, Mr. Gruber. If the polling company does this 'all the time,' for what purpose would they have to be told to select more persons of color and fewer old people in Florida in order to 'maximize what they get' out of their polling?Another person who doesn't know what oversampling is. This is the problem with morons on ZH, Reddit, etc. running around with content they don't understand.
Because the ratio of registered party members is wrong. There's data on this so I don't know why they made such an obvious mistake. That Hillary allegedly lost 10% of her advantage in less than a week on no significant events indicates a flaw.
I don't know why you think I want them to change the data. The data says what it says, if you ask more registered Democrats (than exist as a share of the population), you'll get more Hillary votes. That's not a revelation, it's not even useful except to identify people who like incorrect mostly useless data.
Yes, and Democrats have been defecting to Trump in droves. Makes perfect sense. Tell me another.People registered for one party change who they support all the time. It's not a reliable measure: see the south.
You're just too incompetent to understand why this isn't a problem, but it's funny to see you call data that you don't understand useless.
Why wait until the first? Why not now? Why wait until early voting is well under way? If there was anything that was significant, why haven't they released that now and dribble more out later? Why all at once, making people dig through it to look for anything significant. Nope, not buying it. I do find it humorous that conservatives are hoping Russia steals and releases information through their propaganda puppet that helps their deplorable Presidential candidate.
It's so unpatriotic of those so-called patriots.
Yes, and Democrats have been defecting to Trump in droves. Makes perfect sense. Tell me another.
In the south large percentages of republican voters are actually registered democrats. People register for a party once usually but that does not necessarily reflect their current stance.
You very clearly have zero knowledge of statistics or public polling but seem to be very confident in critiquing it. Can you tell me the source of your knowledge that makes you feel this way?
I at least appreciate you implicitly accepting your stupidity about oversampling even though I doubt you will ever explicitly admit it.
All these insults and all you've done is post a link to politifact and spin a wild yearn about southern Democrats.
All these insults and all you've done is post a link to politifact and spin a wild yearn about southern Democrats.
Yes, and Democrats have been defecting to Trump in droves. Makes perfect sense. Tell me another.
Yet I've never advocated doing anything to the sample. I stopped reading there because I've told you that repeatedly. You should work on your communication skills after your basic math review.Actually I've told you repeatedly why adjusting a sample for party ID is a dumb idea. You have given no reason for why this should be done other than your feelings. You also haven't accepted that what you said about Podesta's email showed that you have no idea what you're talking about.
Basically you've alleged that a national pollster is either incompetent or actively biased based on zero understanding of statistics or public polling. It shouldn't be hard to understand why an incompetent accusing other people of bias based on their own incompetence deserves to be insulted.
I mean what possessed you to make that critique to begin with? Too much right wing media would be my guess.
Yet I've never advocated doing anything to the sample. I stopped reading there because I've told you that repeatedly. You should work on your communication skills after your basic math review.
You act as if getting a sample that isn't representative of the population is a new problem in statistics. It isn't. I still haven't advocated for altering the sample. So, you're still wrong and now I've told you that you're wrong an additional time. As you can see, despite my best effort you're making no progress.Considering the sample was a result of random dialing if you are complaining about the distribution of course you advocated for doing something to the sample. Either that or you're advocating for a nonrandom sample, which is something that no one interested in accurate data would do.
Seriously, you're embarrassing yourself again. You are either accusing professional pollsters of being incompetent or in collusion with a presidential campaign. If you want to provide evidence of either go ahead. If not, you should apologize.
You act as if getting a sample that isn't representative of the population is a new problem in statistics. It isn't. I still haven't advocated for altering the sample. So, you're still wrong and now I've told you that you're wrong an additional time. As you can see, despite my best effort you're making no progress.
Because I'm not convinced. Why does your vast knowledge not include a basic understanding of sample vs population in statistics?Seriously, just stop digging. Why is it so hard for you to admit you said something dumb?
Because I'm not convinced. Why does your vast knowledge not include a basic understanding of sample vs population in statistics?
If you work with statistics every day then why don't you understand sample vs population? Serious question.I know you aren't convinced but that's because you are incompetent. I mean you started out this discussion saying that oversampling was a problem instead of a basic polling technique. You've still refused to admit even that basic error.
I work with statistics every day and I work with polling data fairly frequently. I know how it works. It's abundantly clear you don't. An intelligent adult recognizes when he's made a mistake and stops. You are missing one of those attributes.
If you work with statistics every day then why don't you understand sample vs population? Serious question.
You say you understand it, but then you also say that you can't do anything with an unrepresentative sample except produce incorrect information. Certainly you see the problem here.Haha you're doing this again where you childishly retreat to some inane question instead of admitting fault.
I understand sample vs. population. You do not.
Because I'm not convinced. Why does your vast knowledge not include a basic understanding of sample vs population in statistics?
You say you understand it, but then you also say that you can't do anything with an unrepresentative sample except produce incorrect information. Certainly you see the problem here.
And yet I could pull up my own post where someone pointed out that I was wrong and I admitted it. It wasn't even that long ago. Meanwhile, its probably been a month since fskimospy couldn't figure out if 60 was between 60 and 100.One of your inherent problems is you cannot accept when you are wrong.
And yet I could pull up my own post where someone pointed out that I was wrong and I admitted it. It wasn't even that long ago. Meanwhile, its probably been a month since fskimospy couldn't figure out if 60 was between 60 and 100.
If Podesta wants to poll Democrats, I'm certain he has binders full of them. Why would media polling need to be redlined for probable Democrat voters?