We have a poll out showing the National party 8 points ahead - National party supporters dance on what they presume is Labour's grave.

Now a poll out showing Labour 9 points ahead - has the public really changed their mind by 17 points in less than a week, or are both polls "rogue"?

WHo do you think is leading amongst the major parties, and do you think either poll is accurate?

My pick is that Labour is still leading, but only slightly - the best thing for Labour is that it has a coalition option, while ACT are goneburgers and Winston will almost certainly crash and burn now.

Will United Future and Nationla together have a good impact on the polling being conducted now, and will the revelation that Don Brash knew about the "Green Delusion" pamphlet and who was distributing it (The VERY VERY VERY nutty Exclusive brethren, who also are helping out National in other ways) damage his personal credibility?

("I'd work with the devil himself to get rid of this Government..." - Brash, 2004)

Take the average of several polls and see what you get.

I really do hope NZ First crash and burn. I also agree that Labour are ahead, but the problem with these polls is they are only a small sample. Anyone who has taken a statistics paper will know what shit ALL of stats is, its like economics all based on assumptions

"I eat babies and sacrifice virgins to appease my dark gods" - Helen Clark, 2004

The difference being, Brash actually said what I quoted him as saying, Clark did not say what you have her quoted as saying.

(At least not in public, that I saw)

I hear her say it with my own ears. Mr T was there as well, he'll confirm it.

:update: :update: :update: :update: :update: :update: :update: :update:

Oh ho! Now I've figured out your secret Identity Oob :oob: - your Peter Doone aren't you?

Baseless accusations let loose!!! :ninja:




:D

I am the one and only... God. And I saw what you did in the shower last night... you broke at least 3 commandments.

:lol:

Seriously though, the only poll you can trust is the one on September 17. Consider yourselves fortunate - people in America don't even have that!

:hmm:


:oob: + :fap: = :evil8:




Voyeurism is not Godliness....

:lol:

good work guys

good to see we no longer take politics seriously, Im so sick to death of it right now.

Exactly. All the other national supporters were out working for their money, so the telemarketers go house wives instead. *sighs*

Election year is like a buffet for a troll like myself. Everyone with their views and their arguments, all ready to get their teeth into it.

Why? Go plant a tree bro

Indeed, plant a tree and get down to the earth





(MY PUNS ARE AWESOME. Owned)

I hate those smug bastard trees, why would I want more of them?

Forestry is the new housing boom, oob...

Does that mean we can force possums to pay rent?

hmmm now there's an idea...if people have banana gaurds and dog hotels then i suppose charging possums rent could work...

Well, eviction has already been well worked out, but it'll sure be a bitch cutting free food out of the equation.

It's amazing where thread drift can lead us, eh?

to counter the free food you could just include it in the contract...something like a charge that they have to pay on top of the rent to cover any of the food they'll likely eat...

and yeah it is weird how these threads can turn into talk about just about anything...

Why not plant trees with leaves that become toxic at certain levels of consumption - so if they trash the place they die, and you can recoup the cost with their possum fur or just use them as fertilizer?

:rant:

Umm, if I can rudely return to the subject of this thread, my vote is that the polls are particularly useless. I generally defend the validity of polls, despite my scepticism about the role of media and the use of language in the network of power relationships that is a political society. I think that, over time, polls provide at least a rough statistical gauge of long-term trends in public opinion. However, these ones that came out over the weekend are a joke. I firmly believe that the media has a responsiblity to uphold the integrity of a democracy (that's why they're called the 4th estate).

Snap polls of four people conducted in a Dargaville nursing home in the 30 seconds between Close Up with Susan Wood and the opening strains of the theme song to Coronation St. are, in my opinion, not a sound basis on which to run an election campaign, and make promises about the future of this country. Same deal with surveys of 'morality', in which the sample is self-selecting, and massively over-represented by people with fundamentalist, minority beliefs. Obviously, one doesn't want to legislate the running of the media, but couldn't broadcasting standards/print responsiblity rules be expanded to include publishing statistics that are so dodgy as to be meaningless? Can't someone say, 'if youse are gonna publish polls, you've got to do them properly'?

And bollocks to that 'it's up to the reader to decide on the crediblity of the poll' crap. It's up to the fucking media to make them credible.

/end :rant:

hmm an interesting idea...but then what possum would want to relax in a tree covered in its cousins skin...

nice Maus, I agree.

Using the margin of error, most of these polls come quite close to each other.

Do they? I'm not sure how the margin of error works. I thought it was something that could be distributed across the sets, not all on one set. But I haven't done any maths since 6th form. Anyone care to enlighten me?

provided im not wrong like usual the margin of error is what you can add or take off any of results or all of them not something that gets divided. ie polls with labour at 44 national at 38 nz first at 2 greens at 8 etc and a margin of error of say 2% could be interpreted as labout are between 42 and 46 national are between 36 and 40...but i may be wrong...

I think thats basically right, although its usually something like "there's a 90% chance that Labour are between 42 and 46"

I've often wondered this myself. And having a stats background I'm still confused.

The margin of error surely must relate to the % of the %.

I mean if in the above example, with a margin of error of 2%, if say ACT was on 1.5% the above arguement would say they are in the realms of -0.5% to 3.5% which is obviously ludicrous

Surely its 2% of the % so margin of error would be 1.497% to 1.503%.

However, this took is madness, as fewer people would be saying they would vote ACT and thus the margin of error on the smaller parties would in fact be bigger than for the two main parties.

They really should explain what analysis they use.

Plus they should tell us how many people are undecided and how many people are planning to vote at all.

Polls are totally screwed.

Thanks for that, that's very enlightening. I knew that the margin of error indicated a range, and that it was more complicated than it appeared (as things like this always are). As for undecideds and not planning to votes, I worked for Colmar Brunton for a while, in an election year at that, so made a few calls for the One News poll.

Undecideds do get recorded, and usually appear as that on polls that are published. Also, they ask if you're likely to vote, and if not, then they thank you for your time and hang up. So unless you demonstrate at least a minor level of political engagement by considering voting (whether you do or not is a different story), your opinion is not recorded. In that case, your exclusion wouldn't affect the integrity of the survey.

Unsuprisingly, political polls were the easiest calls to get respondents for, as they were quick, and almost everyone wants their opinion heard. You could spend four hours not getting a single respondent to a survey on Metrowater, and in the same time get 35 respondents to the political poll. I recall one woman, though, who, after agreeing to respond to the poll, refused to say who she voted for last time, and, more importantly, who she would vote for next time. Okay, I respect your privacy, but why respond to the poll?

Weirdo.

The other thing that I noticed, although it was a while ago, and its relevance may have dissipated through natural attrition, was that elderly people were far more likely to vote for their electorate candidate based on the qualities of the candidate, as opposed to people below the age of sixty, who were primarily concerned with the party to which the candidate belonged. No end of cajoling was required to get the oldies to say which party their prefered candidate was more likely to belong to.

The irony of using anecdotal evidence isn't lost on me, but I do wonder if there's a real generational difference to be seen there, in terms of party affiliations...

thats what they mean when they say a party is polling below the margin of error.

polls can be quite accurate in predicting the outcomes of elections, but it is the way the media report on them that often causes problems.

The margain of error is not the % of the %.

If Labour are polling 40% of the vote and the margain of error is 5%, then the TRUE value of labours % vote is likely (usually 95% likely) to be between 35% and 45%.

There is a reason why parties can poll below the margain of error, its a relatively simple reason I think, but I cant remember it off the top of my head..

do you make use of exit polls here?

in the UK they are a pretty accurate guess of the outcome in my limited experience.

I don't really see the point of exit polls seeing as the actual result is going to be known within a few hours anyway :sausage::sausage::sausage:

^^ left bowie is voting for progressives, by the way

Maus, that's particularly interesting what you pointed out about inter-generational differences in methods of determining support for candidates. In my opinion, that is probably to do with the greater homogenisation of party messages (overbearing nationwide campaigns that make sure local candidates don't show any colour) and the weakening role of local MPs in actually helping their electorates (which used to be a lot more common - now they just swan around meetings spouting nice and warm bullshit). Renegade MPs no longer seem to get popular support (c.f. John Tamihere, Maurice Williamson) compared to the old days when MPs like Michael Laws, Marilyn Waring and Mike Minogue had huge personal popularity through their roguish stances.

Bugger the polls, as Jim Bolger once said. go out and vote for left wing parties and prove that national isn't getting close to government.

Amen.