Posts by Mr Mark
Last ←Newer Page 1 2 3 4 5 Older→ First
-
Hard News: Decidedly Undecided, in reply to steve black,
I'll just add the following:
- Although it hasn't been widely publicised, Fairfax-Ipsos provide firm evidence that the Undecideds are disproportionately people who voted Labour in 2011. Thus reinforcing the likelihood that the Undecideds do, indeed, favour a change of Government.
- When I look back at the NZES data, I see that (collectively) respondents get it pretty much bang on when they're asked to place the various political parties on a 0 - 10 Left/Right ideological spectrum. That reinforces my feeling that the vast majority of voters aren't quite as ignorant or simple-minded about national politics as some would argue. And that, in turn, tells me that poll respondents are likely to understand exactly what's implied when Fairfax-Ipsos ask them whether or not they desire a change of government.
-
Hard News: Decidedly Undecided, in reply to steve black,
"I simply maintain that you can't exclude my models on the grounds that they don't fit the data..."
Hence, my suggestion that pollsters ask the more specific Do you favour a Labour-led or National-led Government ? question.
I think it's good that pollsters take a very cautious, social science-based approach that prevents them from leaping to conclusions. However, I would also argue that your alternative models border on the pedantic. I mean, anything is possible when it comes to respondents motivations. I prefer to take a common-sense approach that highlights the most likely scenario (once again, in the context of an MSM that doesn't much care for nuance or caution when it "analyses" poll results - with its usual Labour is Dog-Tucker / National is Soaring hyperbolic frenzy).
-
I do, of course, acknowledge Andrew's argument about the whole rationale for conducting political opinion polls (the brief being to determine as precisely as possible the outcome of an election held at the time of the poll - and therefore the need to exclude likely non-voters and the Undecided), but there are much wider implications involving public perceptions of the general political mood of the Country, which in turn of course, are inextricably intertwined with turnout.
-
Hard News: Decidedly Undecided, in reply to steve black,
Part 4 - The Finale
.......But I just don't think there are multiple interpretations or overly-complex nuances here (particularly in the context of the liberties - bordering at times on gross misinterpretation - taken with Party Support data by journos in their Front Page and Op-Ed poll analyses).
Need to acknowledge the massive disconnect between a relentless media narrative of overwhelming support for the Key Government and what the Change of Government data is actually telling us. The thing is: the MSM don't do nuance when it comes to interpreting the Party Support results, so I really don't think we should be treating the Mood for a Change of Government measure as some great mystery that we need to be exceedingly cautious about interpreting.
-
Hard News: Decidedly Undecided, in reply to steve black,
Part 3
I really can't see multiple interpretations of this. Except to suggest that possibly - just possibly - a small minority of those favouring change wanted a National-led Government, but one involving NZ First in some capacity (whether in a formal Coalition or other arrangement). But you'd have to guess that (assuming they exist at all) they'd comprise a pretty small sub-group of the Yes, Change Government response.
Yep, I'd certainly like the question to be a little more explicit - along the lines of asking the entire sample: "Do you favour a Labour-led or National-led Government ?" (Or a more neutrally-expressed version of this). And, in an ideal world, all of the five public polling companies would ask it - not just Fairfax-Ipsos .
-
Hard News: Decidedly Undecided, in reply to steve black,
Part 2
(My Lap-top suddenly can't handle posting longer comments so, unfortunately, I'll have to break it up into different parts. This is Part Two: The Sequel)
If you take October 2013 ( Fairfax-Ipsos ) as an illustrative example, those favouring a Change of Government outnumbered respondents happy with the status-quo by a clear 7 point margin (50/43). One can only infer that the broad mood was for a change of government. And yet in the same poll, the Nats led Labour by a hefty 16 points and the Right led the Left by a clear 7 points in the one measurement that receives all the MSM attention and associated ( National is soaring/Labour is Dog-Tucker ) hyperbole. Look at the comments on Puddleglum's site, for instance, and you'll see people are shocked to discover that National is not, in fact, currently favoured by well over half the Country. That, indeed, despite the 57% rating in the latest (obviously rogue) Fairfax-Ipsos , National's support among the entire sample (read: a rough approximation of all adult New Zealanders) is in fact slightly lower than it was in 2012 (when they were scoring less than 50%).
-
Hard News: Decidedly Undecided, in reply to steve black,
Part 1
"There are too many different causal models which fit the correlations observed from outside the system. We can't rule out some models based on the evidence we have, so we will be left with multiple interpretations"I think that's more a criticism of Puddleglum's analysis on The Political Scientist (essentially along the lines of Thomas Lumley's and KiwiPollGuy's critique). I'll leave Puddleglum to respond to those arguments.
I haven't focussed on correlation analysis at all. And I haven't been particularly interested in swings in and out of the Undecided category (although the evidence is clear that there has, indeed, been a swing from Labour into Undecided territory over the last 6 months).
I've simply argued that, throughout the last 18 months of Fairfax-Ipsos Polls, we have a striking contrast between the climate of opinion suggested by (1) the Party Support results (based solely, as they are, on the Decided and those likely to vote) and (2) the 'Do you favour a Change of Government' measurement (based on the entire sample). A striking contrast all but ignored by Fairfax journos and the wider MSM (and, therefore, hidden from the wider public).
-
I see my first link didn't work.
Let's try it again...http://sub-z-p.blogspot.co.nz/2014/06/fairfax-ipsos-poll-february-2014.html (If it doesn't work this time, can I just say Bugger ! - thank you).
-
Well, as the bloke who kicked the whole thing off with this (admittedly very brief) analysis of the Left-leaning (or,at the very least, anti-National Government) proclivities of the Undecideds and those unlikely to vote... http://sub-z-p.blogspot.co.nz/2014/06/fairfax-ipsos-poll-february-2014.html, I'll be replying to Andrew ( and Thomas and KiwiPollGuy ) in an up-coming post (couple of weeks time at most). Not that I'll be spending too much time arguing about the correlation stats (I'm sure Puddleglum can argue his or her own corner on that score - indeed, I think they've already responded briefly to Thomas in their comments section). I'm more interested in the wider implications.
In the mean time, here are the Fairfax-Ipsos Preferred PM breakdowns by Party Support...http://sub-z-p.blogspot.co.nz/
-
Hard News: The Letter, in reply to Chris Waugh,
I hope you all saw this interesting post about that mysterious undecided tribe at The Political Scientist
Hate to be self-promotional here (well, no, actually that's not true - I quite like it), but can I just point out that - as Puddleglum at The Political Scientist very generously made clear - his/her post was inspired by a post of mine here...http://sub-z-p.blogspot.co.nz/ as well as by one or two comments I made on The Standard a few weeks ago. Have a look at some of the stats I set-out in the top post at Sub-Zero politics ( A Leap into the Unknown ) but also, in particular, click on the link I provide near the top of that post (the link is in blue immediately below: Fairfax-Ipsos Poll - February, May and June 2014 in red). (sorry, due to probs with my laptop - I can't get on to blogger at the moment, so can't link directly to the page).
It became clear to me a few months ago that while the Fairfax/Stuff-Ipsos Polls consistently put the Nats and the broader Right well ahead of the Left Bloc poll after poll (with all the associated 'Labour is Dog-Tucker' hyperbole from Fairfax journos that we've come to expect over the years), the elephant in the room was the Mood for a Change of Government measurement. In February, according to this Fairfax-Ipsos measurement, the Country was essentially evenly split on desire for a change of government (Yes, Change Government 47%, No, Keep Present Government 48%). Since then, the proportion supporting the status quo has remained the same (48%), but there's been a slight move from 'Yes" (down 3 points to 44%) into 'Don't Know' (up 3 to 8%).
But that's still a reasonably evenly split Country in a poll that made headlines (and generated all kinds of heated analysis) for suggesting National was soaring to unprecedented heights, with Labour supposedly plummeting to an all-time low.
Do the math and you'll find that, among Undecideds and Likely Non-Voters (both excluded from party support stats), those favouring a change of government massively out-number those supporting the status- quo. That's certainly true for the Fairfax-Ipsos at least.
Unfortunately, Undecideds disproportionately become Non-Voters. Vicious Circle.