The last two Newspolls have shown a 5% swing to the Coalition in two weeks. Is this the long-awaited Narrowing, or merely a statistical blip? The other three polls – Nielsen, Morgan and Galaxy – have all shown swings to Labor in the last week, so Newspoll seems out of step. Nevertheless, I am among the minority of online commenters who have predicted some narrowing in the final weeks of the campaign. As always, we will have to wait for more polls to confirm or refute Newspoll’s results.
See my poll table here.
November 7, 2007 at 4:42 pm
Is it my imagination or, from looking at your poll table, that it appears that Newspoll actually lags behind any trends that show up at Nielsen or Morgan? I predicted 80 seats to the ALP previously, but noting Possum’s predictions elsewhere this seems a little pessimistic (just hoping I’m not being optimistic!).
Out of curiosity, (and on another topic) are there any updates anywhere regarding the final results of the PNG elections? I keep looking around on the net, but everything PNG linked seems under construction or out of date… Ihad heard there had abeen a PNG Greens candidate elected, but couldn’t see it anywhere, although they may have been running as an independent. The former General Secretary of the PNG Greens ended up running under the PNG Labor Party name (Monica Hasimani)!
November 7, 2007 at 8:25 pm
The recent behaviour of Newspoll is indeed puzzling. I don’t believe there was a 5% swing in two weeks, and I assume the 58% poll was too high. Whether there has been a genuine move back to the Coalition in the past fortnight remains to be seen. The next Nielsen may confirm this trend, or it may not.
I have been sadly neglecting the international section of the website during the Australian election campaign. The last figures for the PNG election I uploaded in September are here, but they are not final figures. I don’t recall seeing any Greens candidates.
November 9, 2007 at 12:20 pm
Further to Newpoll, I saw the Morgan poll now out, and noted that there appears to be an anomoly between the phone & face-to-face polls. Essentially it appears that the combined ALP-Greens vote remains the same, but there is a swapping effect the two parties vote (the phone poll seems to have a higher Greens vote each time, with a commensurately lower ALP vote). Any thoughts?
On PNG – there were 3 official PNG Greens candidates, all in the Western Provincial/Fly River region, but the best they scored was 4%. However, I had a report that a Green had been elected somewhere.
November 10, 2007 at 5:10 am
Adam,
Would it be straining the friendship to ask if you could put up a graph that included maybe the ACN online poll?
Also, can you do it with error bars rather than connecting lines? That actually makes more sense to me.
November 10, 2007 at 8:37 am
I don’t think much of online polls, and my poll table is very cluttered as it is, so I don’t think I will be adding the ACN online polls.
What is an error bar?
November 10, 2007 at 12:52 pm
This being your most recent plog contribution I thought you might like to update the information on the seat of Parkes. Well at least for the poor man listed as unemployed when its unemployment by choice because he held a public service job – a cop. This Daily Liberal article should clear it up.
November 10, 2007 at 2:57 pm
Thanks for reminding me about Horan. I have included his photo and website now. He called himself “unemployed” on his nomination papers, but I have now called him a former police officer.
Is he related to the Toowoomba Horans?
November 10, 2007 at 8:24 pm
Adam, I wasn’t suggesting cluttering the graph more. Rather I was suggesting a separate plot with the more ‘experimental’ polls in there. The ACN online poll does seem to fairly well behaved. Whereas the Morgan F2F is headed to mars.
As far as error bars goes, something like this. I wish I could find a better example right now but this will have to do.
http://reference.wolfram.com/mathematica/ErrorBarPlots/ref/ErrorListPlot.html
I’m used to analysis of gps surveying data and I find it easier to see a trend without the connecting lines.
Another way of ‘joining the dots’ that I have yet to see but I’d like to, is a line that has a gaussian density. Like those shaded ‘MoE’ graphs but with the line shaded more heavily toward the middle to represent higher probability.
November 10, 2007 at 11:46 pm
HCC, you’re talking to someone who failed Year 10 maths here. I think people know that polls have margins of error, and I can’t possibly plot five sets of MOEs on one table.
November 11, 2007 at 12:06 am
Ok coolies. Btw I’ve become addicted to your graph page
November 11, 2007 at 5:56 pm
I do not know if he is related to those Horans. The only reason I gave him a thought was because I remember that he had Peter Andren’s backing in the NSW State election.
I wasn’t aware of the website so thanks for that. Even if you do have an extra .htm rendering it a 404 but http://www.federalelection.com.au has done the same with the only IND for Calare – Gavin Priestley, so it doesn’t really come as a surprise. No big deal and easily fixed.
November 11, 2007 at 5:58 pm
I do not know if he is related to those Horans. The only reason I gave him a thought was because I remember that he had Peter Andren’s backing in the NSW State election.
I wasn’t aware of the website so thanks for that. Even if you do have an extra .htm rendering it a 404 but http://www.federalelection.com.au has done the same with the only IND for Calare – Gavin Priestley, so it doesn’t really come as a surprise. No big deal and easily fixed.
November 11, 2007 at 7:11 pm
No Adam, it’s not *that* Tim Horan.
November 11, 2007 at 10:14 pm
I have fixed the url, thanks.
I know it’s not *that* Tim Horan, but I was wondering if they are related.
Anyone with local knowledge think that either Horan or Haigh can win Parkes? Or that Priestley can win Calare?
November 12, 2007 at 2:17 am
Sorry – couldn’t resist.
No local knowledge sorry, but I’ll chip in my 2c on Parkes anyway. Haigh is a perenial candidate and Horan couldn’t win the state seat of Barwon. Whilst the new Nat candidate would probably have Anderson alongside of him wherever he goes. So you’d have to be sceptical about their chances.
November 14, 2007 at 8:58 am
Personally I’d like Priestley to win Calare, which I think is fairly obvious, but I don’t think he will. I had a quick read of all the online newspapers for the towns in Calare “town.yourguide.com.au” and the only mentions I found of Priestley were in the towns that were in the old Calare – the rest mentioned Lyne. You would’ve expected them to mention Parkes before Lyne.
Also Priestley is bottom of the ticket so can’t rely on the donkey vote. Priestley did have a lot of Andren in his television advertisements but since his passing he’s gone missing. I think he lost a lot of weight (Priestley that is) without that.
I also agree with David Walsh on Parkes. I think Horan will become a perennial candidate too though and that’s how they build their profiles and win. I doubt Haigh’s diplomatic service will serve him well in a rural seat thus could probably be considered a write-off in every election.
November 16, 2007 at 8:19 pm
I’ve got to say, I can’t get enough of this psephology! I heard the Liberals are still hanging out for a 48% two party preferred national vote, hoping for the marginal seats strategy to perhaps pull out another 1998. Can’t see that ‘narrowing’ happening in just one week. The ‘insight’ program this week showed lots and lots of undecideds in marginal though. Haven’t seen the onslaught of Liberal advertisements yet but Labor seem assured on those figures.