Posts by Che Tibby
Last ←Newer Page 1 2 3 4 5 Older→ First
-
'Damian Christie in Secret Love-Child Surprise'
weird... i was in a place called "Post-Deng" on Lt. Burke St a few days back, and I'm sure that was one of the dishes on the menu!
naturally i opted for the bird's nest combination instead.
-
glad to hear that job you spoke about came through </assumptions>
give us a shout when they've finished with you!
oh, and the best advice i ever heard about the better place you're going to?
don't let the baskets get you down
-
[canberra] looks as if Albert Speer had been commissioned to redecorate Ashburton on the cheap
<genuine LOL>
-
she'd say he was not worthy enough to be a cunt
indeed. 'cunt' is a dirty word for a beautiful thing.
veitch is just a dirty word.
-
I can't help feeling sorry for his wife, though - no matter how much she knew, she probably never expected it to become so excruciatingly public.
i'm assuming you're talking about zoe halford. if so, don't feel bad for her.
with a mug like that she'll be on every [woman's] magazine from here to kingdom come in no time. her career is about to take off.
tony's on the other hand should be sleeping with the fishes.
-
Anyone have some definitive information on that?
rb might remember. it was at a great blend at the boat shed. maybe '06?
-
The reason New Zealanders use TradeMe rather than E-Bay is that a killer idea was refined adapted and implemented in a superior way to meet the needs of the local market.
the one time i heard sam m. talk he claimed that trademe pre-dated ebay.
but you're right on the money with meeting the needs of the local market. they sink money into parts of the site that don't turn coin, but do increase user loyalty. smart move there guys.
-
find the application of the word deprived to folks that don't have a car or access to a phone or are forced to live on a benefit, as relative as it might be, rather bizarre.
are you suggesting my big-screen tv, high-meat diet, and flush toilet is over-consumption or greedy?
because those, mr. griggs, are fighting words.
-
Having just checked our stuff, what you do is weight the over-sampled sub-populations with the inverse of the probability of their selection.
and again in non-math-geek english... this is what i learned today.
you sample more of some groups ("over-sample") to ensure you get a decent and comparable number of responses.
there are methods to ensure that you don't skew the data when comparing groups, as hadyn indicates. in plain english you could, for example, look at the percentage of a variable within groups.
</feels very smug for having worked out "maff">
-
You hope... The data may have been weighted at that stage, or they may have done the weighting themselves, I wouldn't be prepared to put money on it though
according to my on-tap stats guru, that would be a fairly fundamental mistake. i.e one you'd not find among decent analysts.
unless, of course, you did it deliberately.