Written By:
Tane - Date published:
4:04 pm, July 14th, 2008 - 37 comments
Categories: polls -
Tags: nzier, tns
Buried away in the NZIER’s recent TNS survey on the Emissions Trading Scheme was this political poll showing a surprisingly high number of people still undecided on who they’d vote for.
You should be aware it’s based on some pretty shonky methodology (basically it’s a weighted online survey), but it’s interesting to see what happens to the outcome when you include undecided voters in the mix.
The server will be getting hardware changes this evening starting at 10pm NZDT.
The site will be off line for some hours.
Whats the saying used to describe this again? Oh yes, clutching at straws
Nick, you could argue, but I suspect you’re just trolling. A reader sent us this poll and asked us to cover it, so I did. If you’ve got an issue with the content of the post then feel free to elaborate, otherwise leave the discussion to the adults.
Tane,
Correct me if I’m wrong, but I thought you had stopped covering polls in recent weeks. Why did you choose to highlight this one?
Mostly I’ve just been too busy to keep up with the polls. As I said above, a reader sent us this one a couple of days ago and asked us to cover it, so I did.
That’s okay, isn’t it?
I like the won’t vote portion of the pie. That’s my constituency – VTO stands for Vote Them Out. Each member elected does not participate at all in Parliament etc and the seat is effectively cancelled out. Gives voice to the people that want to get rid of the system. Use the system to vote itself out of existence. Democracy – gotta love it.
Tane,
That’s dandy, my friend. Your blog and all that. I just thought it odd to cover a poll with, as you admit, shonky methodology.
Will you be covering the Jimungo poll?
http://www.nzherald.co.nz/section/1/story.cfm?c_id=1&objectid=10520984
Or should I send you an email and ask you to cover it? 😉
Yeah, I was going to write something mocking it (at least this dodgy NZIER poll weighted its results), but in the end laziness won the day.
What interested me about the NZIER poll wasn’t so much the result as the fact they included undecideds, and what a difference that made. It’d be nice to see other polls do the same rather than pretending everyone’s made up their mind.
I was interested to see Tracey Watkins post the don’t know/won’t says for the June Fairfax poll. They were 13% and not included in the final poll. In my opinion this stat and the number of people who were phoned should be included in every poll.
The online poll method can be reasonably credible. I participate in both Neilson and Colmar-Brunton polls online. They e-mail me a link and I go fill in the survey. I must fit some demographic they require and I’m part of the element in the sample.
The only thing not random is my pre-determined willingness to participate.
Does that invalidate the poll? This TNS Conversa poll may not be a shonky as it appears…..though not equivalent to a truly random survey.
I read Idiot / Savant’s objections, then read the survey’s methodology and I’m not sure I see the problem…..
IrishBill: Most polls used to always include the “don’t knows”. They now claim they break out the same as the people who have decided….or don’t vote at all. That may be the case most of the time, but it can’t be considered to be inevitable.
A lot of polls seem to be asking variations on “if the election had been today, how would you have voted?”, which forces the undecideds to make something up “um… ahh…. Winston/that-nice-John-Key/Helen I suppose”.
I wonder how many of those undecided voters will actually vote? I suspect half of them, which would be consistent with the 80-odd percent turnout we tend to get.
So what that leaves roughly 15% of the population “up for grabs” – and internationally undecided voters tend to vote for the opposition. I cannot see much hope for ‘closing the gaps’. on those kind of numbers.
Labour has to ask itself; Does it feel lucky? Does it feel lucky, punk?
IrishBill says: You’ve been banned.
What Steve said at 4:52pm. A couple of years ago one of NZ’s top psephologists told me that online surveys were proving surprisingly accurate and as good as convention telephone polls (which is not that good).
But, where did you find the detailed description of the poll’s methods? I downloaded the NZIER report, but couldn’t see it.
Postscript. Oh hell, I seem to have agreed with DPF…
I wish all polls would include the undecideds. Colmar Brunton always does. Roy Morgan used to but no longer publish them. TV3/TNS have them in their formal report but they are damn hard to track down online.
I do not believe the undecided necessarily vote the same way as the decideds. And for that matter, there is also a difference between hard undecideds and soft undecideds.
As for online polls, they are to some degree the future. If done properly with a large sample base, and weighting they can be reasonably reliable – You Gov in the UK has an excellent accuracy record. But they have the population base to do it.
In NZ it is very hard to get a large enough sample base for online polls. If you want 1,000 responses each time you really want a sample base in the tens of thousands.
DPF,
While you’re passing through…
What do the undecideds and gender split look like from where you’re sitting? They’re both bloody hard to winkle out of the published polls.
Yeah, “undecided” is incredibly important. (so is “no vote”, but for different reasons)
And yeah, undecided voters will usually vote in very different ways, as they tend to disproportionately be character voters, wheras people with hard decisions as to who they’ll vote early on are usually issue voters.
Anita: Traditionally female voters are more likely to be undecided than male voters. However I haven’t seen a significant gender difference in the undecided voters this year.
I think the number of undecided votes is pretty much the key measure at this point in time, considering 1 of the main parties hasnt even released any policy yet.
DPF,
That surprises me, I would expect that women who had voted Labour but are swinging toward National would describe themselves as undecided for far longer than men making the same transition, something about about a sense of loyalty to Clark.
Any idea what explains the lower than expected rates of undecided women?
The last time I saw the Herald-DigiPoll gender splits on National/Labour they perplexed me too.
The undecideds in the public polls tend to be hard undecideds – ie those who are truly unsure how they will vote – even after pushing.
Traditional female Labour voters who have swung towards National may be soft undecideds leaning towards National (which would be recorded as National in most polls) but are not hard undecideds with no preference at all.
Where are the Colmar Brunton results including undecideds?
I can’t find any on the CB website.
The only CB results in the Herald that I’ve found including undecideds is this one in Feb 2004 that totals to 108% (i.e. somebody made a mistake)
http://www.nzherald.co.nz/section/1/story.cfm?c_id=1&objectid=3549474
Does anyone know the “refusal rate”? ie those that are approached but do not wish to participate? I saw somewhere that it was around 70% but this has never been confirmed – David’s earlier comment re requiring a base in the tens of thousands to get 1000 respondents tends to indicate it may be accurate.
If this is the case, then the main polls are only measuring 30% of the pop and this particular poll (including “wont votes”) would appear to the (albeit uneducated) eye to be more credible perhaps…
Yeah, from discussions I’ve had with people involved in polling a refusal rate of 60% is considered good these days, and 70% is pretty standard. Again, David Farrar probably has a better idea considering he does it on a regular basis.
I think the undecided are a lot less this Election there are so many people lining up to get rid of this Labour Government. They dislike them with a vengeance and want a fresh start and vision for New Zealand
Obviously they don’t dislike them with that much of a vengeance Rob or 27% of the population wouldn’t be undecided.
I do want a fresh start for NZ which is why I won’t be voting for National and their 19th century policies.
ak
I’m still waiting for you back on the other thread.
http://www.thestandard.org.nz/?p=2488#comment-71144
There aren’t that many undecided in the major polls you guys are using a very small poll to try and gain some solace for a pretty poor performance in most of the major polls. I believe there is a point when you have to face reality and not be trying to chase butterflies
I believe there is a point when you have to face reality and not be trying to chase butterflies
Yup, that point is the election. And it comes after the point where Key will have to face reality and get specific on policies (and also face some pretty intense media scrutiny, leader’s debates etc).
I’m a realist about the polls, Labour is very much the underdog in this election. But it ain’t over yet!
But it ain’t over yet!
Too right rOB – the current polling is the weirdest situation I’ve seen in decades and from the level of rightie ranting still going on (surely if they were so confident they’d simply sit back smugly), it’s more than a little surreal even to them.
I think there’s no doubt that the media-led (espec Herald, witness the Auck polls as opposed to Wgtn, Chch etc) “Hate-Helen” vibe has taken root in a “hate-ready” minority, but I have great faith that most kiwis will vote on substance and Labour’s exemplary record when it counts.
We’re not idiots: experience and a sound record will always trump slick salesmanship and cunning ruses in the important matters. If the media starts acting responsibly and giving us substance over sensationalism, the decision will be clear to all (or at least enough!)
(and HS: Grow up: you know you lied – repeatedly. It’s on the record for all to see. Childish wriggling and hair-splitting is only further eroding the tiny shred of credibility you cling to)
AK
Damn your eyes!!
Please explain why you keep calling me a liar when I take issue with your comment that NZ is the second best healthcare system in the world based on survey or six countries in which NZ was ranked third equal overall with Australia ?
Please explain why that is as you put it is “childish wriggling and hair-splitting” and is eroding my credibility.
I was prepared to give you the benefit of the doubt that you’d made an honest misinterpretation after a very quick read of a report and were just taking the mickey but you appear to be deliberately attempting to mislead – perhaps a career in parliament awaits ?
HS: OK, slowly this time:
As well you know, those six countries are the USA, Germany, Canada, the UK, Australia and NZ.
And as well you know, and as is clearly reported, NZ scored second in 2006
Knowing the above, you flatly denied this finding two or three times with no justification or explanation whatsoever.
They were the LIES.
Now you are clutching at truisms (more than 6 countries in the world) and by inference belatedly proposing (again sans one iota of evidence or explanation) the existence of some unnamed health system or systems superior to those of the countries above.
That’s the CHILDISH WRIGGLING AND HAIR-SPLITTING
Clear enough HS?
AK
You appear to be demented !
How can you possibly claim that NZ has the 2nd best health system in the world based on a survey of 6 countries in which NZ has gone from number 1 out of 6 in 2004 to number 2 out of 6 in 2006 to third equal out of 6 in 2007 ?
I have denied nothing you buffoon apart from the fact your claim of NZ having the second best health system in the world based on the survey you linked to is an absurd statement to make.
How can you possibly claim that I am lying to say that from the survey you linked to there is no way you can make a case for NZ being the 2nd best health system in the world ?
There’s only one person continuing to spout lies and mislead and it’s not me.
Rob “There aren’t that many undecided in the major polls” and David Farrar “Colmar Brunton always does”, I am still waiting to see the references that show that any polster other than Conversa has shown the figures for “undecideds” in their polls in the last year or so.
The Roy Morgan polls have been accurate within very small percentages for a large number of elections. They picked the 2005 one and the polls well before the 2005 election reflected accurately the election outcome.
http://darrenrickard.blogspot.com/2008/07/political-poll-roy-morgan-july-4-2008.html
This article is a desperate 2 fisted hand job
Viva NZ freedom day, November 2008!!
Darren Rickard, the Roy Morgan polls do not appear to include the percentage of Undecided, and are therefore irrelevant. The whole point of this thread is that the Conversa poll shows Undecided. Please tell us when you find another pollster who does that.
http://www.politicalstockmarket.co.nz/
A prediction market for the election.