Posts by James Green
Last ←Newer Page 1 2 3 4 5 Older→ First
-
Ross. Here's a fatality rate graph from MOH. As you'll see, it's bucked around a bit. The apparent spike in 2005 includes 30% of deaths caused by C-strain meningitis. And the provisional data from 2006 looks like back to 4.2%, the long term average (and still within the +/-2% band for annual fatality rates).
Stephen -- IIRC, there was an outbreak around a high school ball in Owaka or Balclutha, plus perhaps some scarfies. It's certainly not the only time a school ball has been involved (one in the wraps methinks?). Anyway, the median age for a Pakeha case is 16 I think I saw somewhere, compared with sub-5 for other ethnicities. So potentially quite distinct transmission lines.
-
I downloaded the reports from the Ministry of Health, and it was very apparent from the breakdown of area, age, and ethnicity that basically a middle-class Pakeha child at primary school had a very low chance of contracting the disease, at least an order of magnitude less than the equivalent Polynesian child. (If I can dig up the numbers I ran, I'll post them).
The 2004 data does tend to suggest that. However, in 2001 where there were many more cases, Otago and Canterbury were particularly peaky, which somewhat goes against that interpretation.
-
Ross -- that graph in the Scoop article looks impressive, except for one slight ommission. There's no variance indicator. The 150% claimed, is a relative rather than an absolute increase. Those fatality rates are calculated monthly (so fewer than 100 cases), meaning that the margin of error is rather large, like in excess of 5%. In other words, the apparent pattern is most likely noise.
Also, and equally importantly, if the number of cases is waning (and especially if the number of cases are diminishing because of the vaccine), then because the number of cases is a key part of the fatality calculation, you can't compare them like that. An example confound would be that if people with weakened immune systems are not helped by the vaccine, and are over-represented in the number dying, then as those with normal immune systems stop dying, the fatality rate will appear to increase, but only because you're not comparing apples with apples.
Thus, that graph appears completely rotten on 2 key points -- the apparent presence of a serious confound, and no evidence of a significant difference.As to the cost being relatively low. It's presumably relatively low (cost of vaccine divided by cost of prevented cases), despite being absolutely expensive.
-
half of the population could've been given the vaccine and the other not. So we'd have seen whether it worked or not.
Of course they could have done that. However, if it is accepted that there is little risk to the vaccine, and that the relative cost is low enough, then it's not ethically acceptable to do that. You might want to re-read the links I provided to Bradford Hill's stuff (assuming that you've already read them).
-
Ross -- Can you clarify how the epidemic is both waning and the death rate has skyrocketed? (preferably with some numbers).
I'd also be intrigued if you could explain the mechanism whereby a boffin would make a tidy sum. -
I recall seeing a really neat design for an "adkill" TV accessory that relied on the typical compressed audio signal to detec ads. It seemed quite easy to build...
My old TV had a built in solution. When there was too much white space in the picture, the sound turned to white noise. Conincidentally, ads tend to be the most likely source of a mostly white background.
My favourite ripping/encoding tool (grip) has a setting to normalize the volume across tracks which helps with this. Maybe yours does too.
Yes it does. I used to have an ethos whereby I attempted not to modify the original data at all. Which I've since given away for the obvious reason.
-
Because surely the mastering is done, and then that master is pressed to CD and vinyl, so the sound (and therefore dynamic range) is essentially the same? I could be wrong, I'm interested to know.
That is an interesting question. However, I wonder whether the way analogue representation of heavily compressed audio might be vexed. I look at the wave form of the feelers album from noizy's link, and wonder whether you could make such a wave form from a vinyl surface. Any audio geeks got a better theory?
-
I hate audio compression. I especially hate that TV companies claim they are unable to do anything about the volume of advertisements.
It's also frustrating (and very obvious) if (hypothetically of course) a person was to make a mix CD. -
While we're bandying psych theories and rebuttals around, it's also worth nothing that some behaviourists believe that pain and fear are different to other rewards/punishments, and work through more primitive channels. That is, most other stimuli applied in a learning paradigm take multiple exposures to produce learning, but not always so with those 2. Whether that's good or bad is up to your interpretation.
-
Stephen, if you can did out those numbers, I'd be extremely interested. My understanding is/was that MeB was transmitted the old-fashioned way (airborne bacteria/sharing drinks etc.), although I guess that doesn't preclude poor housing. Certainly, there seemed to be enough Pakeha cases iin the media (although perhaps that's skewed reporting!).