Posts by Tze Ming Mok
Last ←Newer Page 1 2 3 4 5 Older→ First
-
Speaker: Identification strategy: Now…, in reply to WH,
Feel free to have this discussion here, I'd just appreciate if you avoid directly implying that I personally owe the Labour Party anything as a member of the tribal left (which I am, but you know, I have this other tribe too that I don't have a choice about).
-
Speaker: Identification strategy: Now…, in reply to WH,
It’s the kind of thing people are going to have views about.
Yes, and in terms of my own personal interests in my ethnic community’s self-respect and safety, there will be two kinds of views.
a) non-racist ones.
b) racist ones.I do Phil Twyford and the Labour Party a disservice? I actually now believe that, as Keith pointed out earlier on Twitter, this is not some gaffe, this is what they wanted. They are happy to make this about race, to get votes from racists. Maybe it’ll work.
Think about how much would it have cost them to *not* make this about race. Not anything. And how much did it cost *you* for them to have, as they did, make it about race? Again, not anything. How much did it cost me? My family? Those 126,000 Chinese Aucklanders? A lot. Like I said, this is personal, and it looks more and more like it was a calculated public sacrifice. I understand that there are those on the Left who wish this would just go away and me and Keith and all those other people would just take a deep breathe and focus on the important issues. I understand that. But maybe you should go and tell someone else that. Go tell someone who isn’t Chinese. It’s not really the right moment.
-
Speaker: Identification strategy: Now…, in reply to Kumara Republic,
Historical footnote: In 2005, 47% of the 'Asian' people polled by National in a private study were going for Labour, 40% for National. No sub-group breakdown available as far as I know.
-
Speaker: Identification strategy: Now…, in reply to Danyl Mclauchlan,
But of course, there is a lot more in this dataset that I am not going to get to before bedtime...
-
Speaker: Identification strategy: Now…, in reply to Danyl Mclauchlan,
Surprise, Asians are significantly more sympathetic to immigration!
And there were only 17 NZ-born Asians in the sample, so not a very good basis for subgroup analysis there.
We may also expect that those Asians who were least likely to participate in the survey may have skewed towards the FOB side, which we could correlate with non-voting behaviour... or not.
Again, if this is the best we have, it shouldn’t give anyone much confidence that they could get away with throwing this particular population under a bus to get elected. -
Speaker: Identification strategy: Now…, in reply to Danyl Mclauchlan,
There’s a weight column which makes the data a bit more robust. But yeah, Asian and Pacifica New Zealanders are way younger than other demographics, which gives them lower response rates to surveys like these, so the data isn’t as robust as you’d like. On the other hand, it’s all there is.
I got no hate for the NZES, but from my incredibly brief glimpse just now, not a particularly sound basis for any political party to strategically write off the NZ Chinese vote…
-
Speaker: Identification strategy: Now…, in reply to Danyl Mclauchlan,
Based on the 2011 NZES survey, not many. That’s a National or Did-Not-Vote bunch of voters.
Okay, five minutes later, you're right: if by 'not many', you mean there are 'not many' Chinese people *in* the NZES 2011. There are 103 'Chinese, Indian or other Asian' respondents in the dataset. That's about 3% of the dataset, meaning that Asians altogether are under-represented in the survey - about three times less than there should be. And of course, we don't know their actual ethnic group as the top code for Asians is basically 'Asian' ('Chinese, Indian or other Asian'). We can only guess the rough proportions of ethnicities of those 103 people going by our general 'Census proportion' habit. I guess a third of these are Chinese, a third Indian, and a third 'other Asian'. Which means that even if we knew they were Chinese, there would barely be enough sample for healthy statistical inference. Nerd out.
-
Speaker: Identification strategy: Now…, in reply to Danyl Mclauchlan,
Based on the 2011 NZES survey, not many. That’s a National or Did-Not-Vote bunch of voters.
Damn you McLaughlan, now you've got me downloading microdata past my bedtime. But I also just had a peep at the questionnaire and the ethnicity categories only reach to "Chinese, Indian or other Asian".
-
Speaker: Identification strategy: Now…, in reply to 81stcolumn,
I repeat: Hahahaha I instigated a Bayesians vs Frequentists comments battle!
-
Speaker: Identification strategy: Now…, in reply to Tze Ming Mok,
Though of course if you did a reliable sample survey, then still had to compare it against a real estate register based only on last names, there would still be serious problem with how racist it is. Because it’s impossible in such an approach to differentiate white, Maori, or any number of other ethnic groups formerly colonised by white people, from white people in data based on last names, meaning that the only defined groups that can be reliably identified by last name are Chinese people and Muslims (not an ethnic group), which means that it’s unavoidably racist to do it at all.