Posts by Kyle Matthews
Last ←Newer Page 1 2 3 4 5 Older→ First
-
Not very Friday, but I want to refer people to this email by Duncan Woods sent to Brian Edwards in response to comment he made on the sentence given to the killer of Nayan Woods a few days ago.
A wonderful statement of humanity and parenthood I thought. Recommended reading for a time when being upset is OK!
-
Is it a standard scale or does it go up exponentially like the richter scale?
-
Any in the Narnia books?
-
I hope it starts a craze of otter ownership. Better an otter than a tiger...You'd have to have floorboards, preferably polished, not carpet. Lino would be okay. And a pool.
I would give up vast amounts of money (which I don't actually have, but anyway) to have a pet otter. Even more to have four.
I'm the guy that goes to the zoo and while everyone else wanders around all the animals, I spend two or three hours watching the otters play. Most favouritist animal ever.
-
Also rather cheering to see a shameful Tea Bagger twatcock like Carl Paladino not only got his arse kicked hard, but managed to make his concession even more bizarre and ugly than his campaign:
He'd make an excellent mobster. "Me and this here bat, we'll be back tomorrow, and we want our election OK?"
-
Don Brash disagrees with you
Got relatively close to being PM a while ago. That's a list of new right talking points, scary.
-
It's also interesting to me that these studies always emphasise the demographics of the perpetrators but totally overlook the demographics of the victims. There's a pretty obvious subtext at work there.
Demographics of perps is consistently collected by the police. Victims - I don't know if police collect it at all (victim support probably does?), but if they do, it'd be inconsistent so not much use for a statistical study.
-
Yep, Rand Paul is off to the Senate.
It appears that Stewart and Colbert weren't quite as effective as you'd hope.
-
Kyle: possibly one problem is this. Suppose the risk of problem X happening is tiny, maybe 1 in 10,000. Suppose a certain group "G" has double that risk, so 2 in 10,000.
I'm not sure it's so useful in the abstract, but let's put your doubling figures in the real world.
To take what Bolger was talking about if the likelihood of a person who is unemployed committing a property theft type crime is twice as much as an otherwise similar person who is fully employed, wouldn't that indicate to us another reason that we want to keep unemployment low? Given that there is a reasonable amount of property crime in NZ. And do some other things with unemployed people that make them less likely to commit crimes?
I'm not opposed to what Jackie is saying - at the micro level all generalisations are wrong, and indeed at the macro level a bunch of them are as well. That doesn't make all statistical data gathered and put into groups useless.
-
Not when the boomers went to university, no.
Actually the changes in the mid 1970s were introduced as a reflection in the growth in participation in tertiary education that the baby boom generation caused.
The first ones through missed it, but the baby boomers born in the mid 50s caught it.
And the system before that wasn't income tested on parents either. Your bursary from high school (which stayed a fixed amount for many years, it was hardly worth applying for by the time I went to university) was a half-decent amount of money, students made up the gap with summer savings, scholarships, government grants. It was a lot more generous than the scheme over the past twenty years.