11% of Americans can't find the USA on a map

[Provost Harrison]

Originally posted by Sagacious Dolphin


Much of a muchness that one. Reno

What is further west, Barcelona, Spain, or London, England?

I could tell you it is LA not Reno quite easily, whereas between London or Barcelona...well, they both lie on the Greenwich Meridian so that isn't a fair question...

[Anodyne]

OR; What is the westernmost capital in Europe?

[rah]

Originally posted by Sava
silly rah, that was my whole point... many of these so-called "national" polls aren't true representations of the demographics of America. I assume that you are in the polling industry. And you are correct, the people that do not effectively balance polls do give your industry a bad name. Do I think every single poll is crap? No. You should stop generalizing my words.

My overall attitude on polls is that they should be taken with a grain of salt, that's all.

It's nice that you assume I have no knowledge of the polling industry. But I will stop there as I'm not about to stoop to your level of name-calling.

Ah, your story changes when you get challanged. No surprise. Read what you wrote an I quoted from you in my earlier post.
Based on that, no, you don't understand. If you want to weasel out of the statement and retract, fine. If you still stand by what you posted...........I'll be surprised since your later post leads me to believe that you actually do understand it a little.

RAH
And yes, I take it personally when people attempt to trash the industry that I work for when they don't know what they're talking about.

[Provost Harrison]

Originally posted by Ade
OR; What is the westernmost capital in Europe?

Good one. And I am not entirely sure...is it Lisbon?

[Dis]

what about reyjavek? I have no idea how to spell it. It's a city in Iceland and I also believe its the capitol of iceland.

[Felch]

Originally posted by Ming
What's more important is that the sample is "representative" of the US and NOT just random

Granted, but randomness is the easiest way to make something (hopefully) representative. Point is, a small sample size is still valid.

[rah]

On a random sample for the US, it would have to be considerably larger than 1000. I've got a guy working for me that could tell me exactly how big it would have to be, but he went home . Assuming true random, and not biased random, (I.E. any phone survey is biased random (without demo weighting on the back end) due to the lower incidence of phones as income decreases.)

But I'll bet it's around 5000. Which I'm sure we would all agree is still considered SMALL. (I really liked your use of (hopefully)).

RAH

[Ming]

Originally posted by Felch X
Granted, but randomness is the easiest way to make something (hopefully) representative.

Actually, it's the exact opposite. A random poll has no hope of being representative and accurate. As Rah points out, it's a science to get a small sample that truely reflects reality.

[rah]

It all comes down to what you consider small. But you are correct, for a random sample to have any hope of being a rep sample, it has to be larger. And since every survey we send out costs, it's cheaper to take the time to make it a rep sample and go with the smaller sample.


And most random samples are likely to be biased in some manner based on the selected methodology, so it's always preferred to weight the answers on the back end if you can collect demographics on this type of sample.

[OneFootInTheGrave]

Interesting only 5000 to get the good prediction for 300 million....
Can this be right?

At least you have to choose different racial/age/geographic parts.

Can't make it in one city can you?

[Dauphin]

Originally posted by Provost Harrison


I could tell you it is LA not Reno quite easily, whereas between London or Barcelona...well, they both lie on the Greenwich Meridian so that isn't a fair question...

I'm almost certain Barcelona doesn't lie on the meridien (close, but definitely not) and I'm sure Reno is west of Los Angeles (just).

[rah]

As I said, that was a guess (even though from experience i doubt it's much higher than that) The trick is your random methodology. If I based it on percentages of pop by AREA CODE and did a raised digit of selected phone numbers from the phone data bases (raised last digit is used to not eliminate unlisted numbers, and maximize working exchanges) Then weighted for the lower income homes, I think 5000 would be more than sufficient.

Random, but not completely random.

[tinyp3nis]

Ok, I didn't get that biggest religion right either... I wasn't sure what it was, only thing I was sure it's not christianity...well I was wrong. I recall being a kid and believing that christianity was the biggest, and then getting information about (from source that I can't remember but remember being considered reliable source by myself) christianity not being the biggest.... It's a conspiracy!!!!

I allways found polls (especially the ones about popularity before elections) so accurate that they make the counting of the votes boring to watch because you already know the results...

[orange]

Reno is west of Los Angeles and Barcelona is well east of the PM.

[orange]

Here's a good one...

Don't look at a map

Which is closer to the Prime Meridian (that is, further west)

Moscow, Russia or Jerusalem, Israel