QUOTE (jammonius @ Sep 14 2008, 04:59 PM)

Wait a minute, bushwa. This is not a subject that should require documentation ...
Well, then I guess I'm just dumber than the average bear, because I
am continuing to ask for substantiation of the claim. Either you can document it, or you cannot. Are RRMB standards too stringent for you?
QUOTE (jammonius @ Sep 14 2008, 04:59 PM)

There's nothing controversial about the claim that polling and pollsters have an agenda.
If it's something other than the two points listed in my initial question, something OTHER than obtaining a reliable result in exchange for compensation, there sure as hell is something "controversial" about it. If the "agenda" you allege were NOT "controversial," then WhyTF would you waste time alleging it exists?
QUOTE (jammonius @ Sep 14 2008, 04:59 PM)

You do acknowledge, without the necessity of "documentation' that there are more registered Ds nationwide than there are Rs, don't you?
So, on that basis, all polls should poll more Ds than Rs. However, pollsters ALMOST NEVER poll more Ds than Rs in any given sample.
WOW! CHECK OUT THE LINE ABOVE THAT I UNDERLINED AND BOLDED. You may as well have handwritten the statement, "
I do not know the first good goddamn thing about political polling," then sworn to it under oath, had it notarized and signed by 3 witnesses.
Thanks for citing an easy example that provides an opportunity to demonstrate for you as to why some folks having a little bit of information can be so fucking obnoxious when they're not intellectually developed enough to recognize there's also a great deal they DON'T know.
Dem registrations DO generally outnumber Republican registrations. And sampling DOES generally include more Republicans! What a find! OMG, what a scandal! WhyTF has no one ever noticed this before?! You have single-handedly broken the code and revealed the gigantic fraud! YAY!
Boyo, if your assertion of a nefarious agenda relies on
that, you have just flushed your theory down the toilet. See, um, actually, my friend, this observation of yours is because Republicans typically have a significantly higher turnout rate than Democrats - often high enough not only to balance the registration numbers, but to even overcome them.
If one is conducting polling on a partisan political race and establishes a sampling that directly represents registration figures, they're going to get useless numbers that will humiliate them when the election is over. For folks interested in and knowledgeable with regard to polling, that reality is sort of a kindergarten basic.
Gotta tell you, Jam, I'm trying desperately to think of an analogy for this statement of yours that will reasonably communicate the amazing extent of your inadvertent admission. I don't claim to be good at analogies, but this is perhaps a bit like me maintaining I am the world's foremost authority on mathematics, and starting out by declaring that 2 + 2 = 832,165.07
Sampling weights for political parties have to be adjusted from case to case, and the offices, ballot measures and even the candidates in play in a given election have an impact on the weighting, from national and statewide samples, right down to the individual precinct level. (I suspect naming Palin forced virtually ALL of the polling firms to reassess and adjust their sampling.)
There is no across the board rule of thumb such as 135 Democratic registrations = 98 Republican registrations to apply across the board. But that's why people who know WTF they're doing - like statisticians, math scholars, marketers, sociologists, historical researchers and such - are doing the polling, instead of dorks who are simply able to count registrations and scribble hash marks on a piece of paper, then tally them up five at a time.
QUOTE (jammonius @ Sep 14 2008, 04:59 PM)

I'm prepared to discuss this, but I see no necessity to restate the ABCs of polling methodology here as I assume you know this already.
For sake of a baseline, however, what is your understanding of the typical polling sample and of the percentage of party affiliation typically used?
Actually, you have unintentionally and unknowingly demonstrated that you're not past the "A" of polling - I'm guessing you
can count. But you are nowhere near the Bs and Cs yet.
Now, unsolicited I've already provided documentation demonstrating that those polling folks were (gasp) well aware of and involved in study years ago of the "hot topic" that Ron Paul's supporters, and apparently you and a couple of RRMBers recently "discovered." I think that before answering any questions from you, I'll just hold off until you provide the documentation and substantiation that I've asked for a couple of times now.