I see your polls TV/Radio/internet but I never get asked, nor does anyone I know

Where do these figures come from?

I've been asked once, via a phone call at home.

I think it greatly depends on what the particular organisation wants the outcome to be.

Yar. Once several years ago. By phone again.

Someone called once and I declined their most generous offer to partake in their pole.

I only partake in phone polls or surveys if they are prepared to come clean as to who commissioned it. It's surprising the number of survey companies that expect you to give out information without knowing what it will be used for.

I was asked once

Slim wrote:I've been asked once, via a phone call at home.

I think it greatly depends on what the particular organisation wants the outcome to be.

I would tend to agree with them trying to get a desired outcome as I have also been quizzed on other topics and then declared as not fitting the target group so apparently not having an opinion worth including hence spoiling the intended result

I have been twice - and while I normally would not have participated, as a lefty in an area of nasty tories I felt it was my duty to try to skew the results...

I partake in an election poll about once every three years - the ones that about 2.5 million others do, with a margin of error of approx 0.0000% (or so we are led to believe )

As for those other polls, nope - never been arksed.

bugle wrote:I see your polls TV/Radio/internet but I never get asked, nor does anyone I know

Where do these figures come from?

The more well known polls use Random Digit Dialing, and then random selection of a potential voter in each household (I think). This approach includes every household in New Zealand that has a landline telephone (including those with unlisted numbers).

This method is by no means perfect because potential voters without a landline have no chance of being selected, phone polls a biased toward those who speak English well, the response rates for phone surveys have been declining, and those from small households have a greater chance of being selected than those from large households. Post survey weighting tries adjust for some of these things, but is not perfect either.

Nationwide door to door surveys would provide better response rates, more representative samples, and would be more robust overall, but a sample of 1,000 would take a couple of months to collect and process and the budget would need to be enormous (and there are only a couple of companies left in NZ who do them). So phone surveys are currently the accepted approach. The alternative of an online political poll is, well, just plain bad.

None till this one.

Only once every 3 years since 1984, usually on a Saturday in November (aprat from 1984 and 1987 when it was in July)

I have been phoned for a few polls. They all seemed pretty straight & well designed for internal consumption by parties though, not for reporting to the public.

You should not necessarily expect to be told in detail who is commissioning the poll before answering questions and especially not before deciding whether to participate, as that can skew the results horribly. I don't see why they shouldn't tell you after the survey.

Just in case you care... here are two comparisons to show which polls were close and which were not. The Horizon poll is an online poll.

http://www.scoop.co.nz/stories/PO1111/S00761/analysis-of-political-polls-election-ratings.htm

http://www.kiwiblog.co.nz/2011/11/how_the_pollsters_did.html