Posts by BenWilson
Last ←Newer Page 1 2 3 4 5 Older→ First
-
Hard News: The flagging referendum, in reply to linger,
in that failure to vote has no single assigned meaning, and no direct count; but “informal votes” are tallied, and can be interpreted less ambiguously as protest rather than apathy or laziness.
There is a direct count. The number people on the electoral roll less those who voted, informally or not. I don't really see how the meaning is less ambiguous than an informal vote. Except in so far as it even more clearly signals disengagement. Which is the signal I want to send. You could call it laziness or apathy if you like. Or you could say that it clearly signals that the whole process is so far below the threshold of worthwhile activity that I won't even open the mail relating to it. I'm as "lazy" or "apathetic" towards it as I am to circular mail that comes to my box. I could show great energy in reading an entire catalog of carpet prices so that I can come to a more full and informed decision that I don't want to buy a carpet, and I could signal it unambiguously by giving the carpet shop a phone call to tell them that I don't want one of their carpets. Or I could toss it in the recycling, signalling the same thing without wasting anyone's time at all.
-
Hard News: The flagging referendum, in reply to linger,
If you don’t care about either choice (you’re equally ‘meh’ about either the current or the proposed new flag), but you want to protest against the process, it would be better to submit an informal vote recording that [e.g. handwritten “(3) Neither"] than not to be counted at all.
Better for what? I'm still counted as not counting, either way. As in, a number will appear somewhere that x people didn't even vote (and it will be a large value of x). But one of the ways I had to make an effort to not count.
And risk ending up with the tea towel flag?
Well, I do use tea towels. Flags, I never use.
-
The only way I can find to protest this process is to not take part in it at all. No other signal can convey what I feel about something so underwhelming as the choice we're presented here. Do I want coke or pepsi? No. Right, so we choose for you? Fine, I'm still not going to be drinking it.
-
Polity: Poll Soup, in reply to linger,
That's a problem with any survey at all, really. Either it's completely anonymous so you can't even check whether people are deliberately messing with it (as can happen with internet polls), or it's not and people might feel afraid in some way to answer honestly. I don't know how you can possibly control those things. Instead you caveat your knowledge to that limitation: "Given that the respondents answered honestly, to the best of their ability". Apart from that, presumably most of it rides on the surveyors stating that the identities of the respondents is not made public. I'd be surprised if 30% of the population is that distrustful of an academic study they agreed to, to the extent they'd deliberately muddle their responses.
In other words, surely the null hypothesis is just that they forgot (we can't know which time).
-
Polity: Poll Soup, in reply to David Hood,
I might also investigate, although not any time soon.
I wouldn't expect it to be "uniform" on jinterest! That was just a suggestion of one of the many variables that could contribute to political forgetfulness, that might have correlation. As I said, could be fun to model.
-
70%?? Wow! That's shocking. I wonder if forgetting is correlated to any of the other variables. Low interest in politics, perhaps (jinterest). That would be a fun one to model.
-
Polity: Poll Soup, in reply to David Hood,
What struck me was the frailty of human memory-
Yes, an actual study that asked them closer to the time is more reliable than getting them to dredge back. How much more reliable? One to look into, considering that a longitudinal study is more expensive and also the opportunity has simply passed to collect information that you didn't think you might want at the time. I guess it would be possible to model the memory curve (particularly in light of having an actual longitudinal study as a reference point) and thus put some kind of confidence numbers around a study done right now which asks the candidates to please think back and state who they voted for as far back as they can remember.
If they're given as much time as they like to answer (and I think they mostly are, it being a paper and internet form based survey) it could still be very useful for insights into how people change their voting. Quite aside from being an interesting study into political memory.
Not that we're designing the NZES.
-
OnPoint: Yeah nah, but what *do* we…, in reply to Rich Lock,
Now all sir has to choose is
Arrrg! The choosing! It burns!!
-
Polity: Poll Soup, in reply to linger,
Wouldn’t “No-Vote” also include those under 18 at the time of the earlier election? If so, those numbers would swamp the signal you’re seeking.
I'd have to look a bit harder to see if they're No-Vote or just NA (R-style), but yes, they should be accounted for. They could be excluded reasonably easily, as the survey has their birth year and month (jyear and jmonth variables). I get a rough count of 100 of the respondents being in this age range out of the 3101 respondents.
-
Polity: Poll Soup, in reply to David Hood,
Yes, there was a one off (not in 11) big move from Labour to National.
Also looks like quite a big move from NZF to National.
Just some thoughts, David - it could also be handy to make a net flow tally which has only the upper triangle, because we have to subtract National->Labour from Labour->National to see the total. Also, it would be handy to see No-Vote in the columns, so we can see how many went from not-voting to voting which parties.
But good stuff nonetheless.
Last ←Newer Page 1 … 109 110 111 112 113 … 1066 Older→ First