Re: Do the phone polls capture the young vote?
 

"thing2" <thing2@nowhere.commy> wrote in message
news:g4a9u5-hpv.ln1@news.vuw.ac.nz...
> Just looking at the USA polling and there seems to be a clear % of <30
> who are democrat. Translating that here and that our polling does not do
> mobile phones and I just wonder what or if correction is needed...
>
> regards
>
> Thing

All polls / surveys / "studies" are ridiculous and useless, including th
idiotic Neilsen ratings and the box office numbers for movies. Most polls /
surveys ask about 50 people out of a population of 4 million and yet people
blindly expect realistic results?!

Re: Do the phone polls capture the young vote?
 

"Your Name" <your.name@isp.com> wrote in message
news:gesuef$kdr$1@lust.ihug.co.nz...
>
> "thing2" <thing2@nowhere.commy> wrote in message
> news:g4a9u5-hpv.ln1@news.vuw.ac.nz...
>> Just looking at the USA polling and there seems to be a clear % of <30
>> who are democrat. Translating that here and that our polling does not do
>> mobile phones and I just wonder what or if correction is needed...
>>
>> regards
>>
>> Thing
>
> All polls / surveys / "studies" are ridiculous and useless, including th
> idiotic Neilsen ratings and the box office numbers for movies. Most polls
> /
> surveys ask about 50 people out of a population of 4 million and yet
> people
> blindly expect realistic results?!

Not true for election polling, maybe for the Nielsen ratings and box office.
It's a popular but misguided belief that election polling and similar
studies are "useless". I used to do computer admin at a market research
company, and they used a tracking based system with the size of each sample
being 750 (tracking as in, doing a series of surveys and tracking the
results) which ran all the time, not just during election year, with the
time intervals between each being weekly, changing to nightly leading up to
an election. They usually predicted the results to within a couple of %,
best year I think was a little under 1% off. The demographic questions that
they ask at the end of surveys allow them to see how close to being
representative the sample of people they talked to was by comparison to the
results of the last census and there are ways to adjust the weighting of the
poll results accordingly. That's somewhat of an oversimplification, but
that's pretty much how it works.

Re: Do the phone polls capture the young vote?
 

In article <mrd9u5-hd7.ln1@news.vuw.ac.nz>, thing2 <thing2@nowhere.commy> wrote:

>I think the numbers are 800~1000 people....they must have some logical
>basis on looking for a sample that size, its going to be a balance of
>numbers, time taken and cost, sure......

... and hopefully a power analysis at least :)

Re: Do the phone polls capture the young vote?
 

"thing2" <thing2@nowhere.commy> wrote in message
news:mrd9u5-hd7.ln1@news.vuw.ac.nz...
> Your Name wrote:
> > "thing2" <thing2@nowhere.commy> wrote in message
> > news:g4a9u5-hpv.ln1@news.vuw.ac.nz...
> >> Just looking at the USA polling and there seems to be a clear % of <30
> >> who are democrat. Translating that here and that our polling does not
do
> >> mobile phones and I just wonder what or if correction is needed...
> >>
> >> regards
> >>
> >> Thing
> >
> > All polls / surveys / "studies" are ridiculous and useless, including th
> > idiotic Neilsen ratings and the box office numbers for movies. Most
polls /
> > surveys ask about 50 people out of a population of 4 million and yet
people
> > blindly expect realistic results?!
> >
> >
>
>
> Oops wrong group....
>
> I think the numbers are 800~1000 people....they must have some logical
> basis on looking for a sample that size, its going to be a balance of
> numbers, time taken and cost, sure......
>
> regards
>
> Thing

Unless you survey EVERYBODY (and everybody actually gives a truthful
answer!) the results are just mathemacially manipulated guesswork, and
therefore not accurate and are usually misleadingly reported ... and far too
many people actually believe such garbage results.

Re: Do the phone polls capture the young vote?
 

In message <getv3r$80f$1@lust.ihug.co.nz>, Your Name wrote:

> Unless you survey EVERYBODY (and everybody actually gives a truthful
> answer!) the results are just mathemacially manipulated guesswork, and
> therefore not accurate ...

Nonsense. They can actually be more accurate than a full census/vote of
everybody.

How come?

Because it's actually impossible to do a census/vote of everybody. There is
no such thing as 100% voting in an election (not even in Australia).

Re: Do the phone polls capture the young vote?
 

"Lawrence D'Oliveiro" <ldo@geek-central.gen.new_zealand> wrote in message
news:geu026$8c2$1@lust.ihug.co.nz...
> In message <getv3r$80f$1@lust.ihug.co.nz>, Your Name wrote:
>
> > Unless you survey EVERYBODY (and everybody actually gives a truthful
> > answer!) the results are just mathemacially manipulated guesswork, and
> > therefore not accurate ...
>
> Nonsense. They can actually be more accurate than a full census/vote of
> everybody.
>
> How come?
>
> Because it's actually impossible to do a census/vote of everybody. There
is
> no such thing as 100% voting in an election (not even in Australia).

I never said it was possible. I said it's the ONLY way to get accurate
results. Anything elese is pure guesswork and must be viewed as such, but
unfortunately far too many people don't understand that and fully believe
the silly survey results are completely truthful, rather than the
mathematically manipulated nonsense that they are. That's why "results" are
always hidden behind percentages rather than telling you the actual numbers,
ie. "50% of people" rather than "5 out of the 10 people we bothered to ask".

So called "medical studies" are even worse since they usually "study" a
ridiculously small number of people and then claim the results are "proof"
.... for that week, and then another silly "study" changes that idea the
following week.

Re: Do the phone polls capture the young vote?
 

"Your Name" <your.name@isp.com> wrote in message
news:getv3r$80f$1@lust.ihug.co.nz...
>
> Unless you survey EVERYBODY (and everybody actually gives a truthful
> answer!) the results are just mathemacially manipulated guesswork, and
> therefore not accurate and are usually misleadingly reported ... and far
> too
> many people actually believe such garbage results.

Not accurate or not 100% accurate? 90% and upwards accurate, which is what
many surveys/polls etc. manage is a hell of a lot better than just guessing,
wouldn't you say? I'd rather know something to within that margin of error
than have no idea whatsoever... why so hostile? Was your family murdered by
a telemarketer or something?

Re: Do the phone polls capture the young vote?
 

In message <geu1pp$9id$1@lust.ihug.co.nz>, Your Name wrote:

> "Lawrence D'Oliveiro" <ldo@geek-central.gen.new_zealand> wrote in message
> news:geu026$8c2$1@lust.ihug.co.nz...
>
>> In message <getv3r$80f$1@lust.ihug.co.nz>, Your Name wrote:
>>
>> > Unless you survey EVERYBODY (and everybody actually gives a truthful
>> > answer!) the results are just mathemacially manipulated guesswork, and
>> > therefore not accurate ...
>>
>> Nonsense. They can actually be more accurate than a full census/vote of
>> everybody.
>>
>> How come?
>>
>> Because it's actually impossible to do a census/vote of everybody. There
>> is no such thing as 100% voting in an election (not even in Australia).
>
> I never said it was possible. I said it's the ONLY way to get accurate
> results.

If it's not possible, then it cannot be a way, can it?

Re: Do the phone polls capture the young vote?
 

In message <pan.2008.11.06.09.04.42@carnations.com>, Carnations wrote:

> Also, why do you think that Statistics cannot be used to accurately
> predict the view of an entire population of votes cast to within a
> percentage point or two?

In fact, it's well known that censuses tend to miss people belonging to
particular social groups (low-income, living in rental accommodation or
homeless, that kind of thing). And since censuses are commonly used for
such things as setting electoral boundaries, this can influence election
outcomes in undesirable ways.

For instance, back in the 1990s (or was it 1980s?), there was a proposal in
the US to amend its census procedure to get around this problem, by using
statistical extrapolation to fill in the gaps. The Republicans, who had the
most to lose from the kind of voters who would be picked up by such a
proposal, managed to block it as "unconstitutional".

Re: Do the phone polls capture the young vote?
 

"Nik Coughlin" <nrkn.com@gmail.com> wrote in message
news:geu605$58b$1@registered.motzarella.org...
> "Your Name" <your.name@isp.com> wrote in message
> news:getv3r$80f$1@lust.ihug.co.nz...
> >
> > Unless you survey EVERYBODY (and everybody actually gives a truthful
> > answer!) the results are just mathemacially manipulated guesswork, and
> > therefore not accurate and are usually misleadingly reported ... and far
> > too
> > many people actually believe such garbage results.
>
> Not accurate or not 100% accurate? 90% and upwards accurate, which is
what
> many surveys/polls etc. manage is a hell of a lot better than just
guessing,
> wouldn't you say? I'd rather know something to within that margin of
error
> than have no idea whatsoever... why so hostile? Was your family murdered
by
> a telemarketer or something?

You must be an engineer: "close enough is good enough". :-\

I'd rather have actual results that mean something are are not guesswork.
Results that are published saying they are either guesswork or accurate only
to those few people they bothered to survey.

As I said, far too many people do not understand how these things work and
take them as being 110% truthful, which in the case of silly "medical
studies" that are announced on TV News can be dangerous.