Letters to the Editor
zzz05
Published Letters: 413 Editor's Choice: 9
-
new iraq mortality survey
[Read the article: Traditional values? Hastert flip-flops on Foley, again]
[Read more letters about this article: Here]There's a standing joke among statisticians; the skeptic who proclaims "Why, this estimate is based on only a sample! And a random one at that!" The fact is, that it can be shown mathematically that the increase in accuracy from increasing sample size virtually levels off after a certain point, and is not related at all to the size of the population sampled from. Which is something that everybody actually knows without knowing they know it. After all, whether testing a drinking bottle or a swimming pool for chlorine, you only take a little test tube sample; you don't have to haul out a bucket full from the swimming pool because it's so much bigger. The underlying sampling principle is exactly the same.
-
Saddam's body count
[Read the article: Study: 655,000 Iraqis have died as a result of war]
[Read more letters about this article: Here]"Can they extrapolate how many would have been killed and still under the heel of Saddam if we hadn't gone in?" and similar questions:
The estimates from the Hopkins type studies are EXCESS mortality. I.e., they are what you get, ***after subtracting the death rate previous to the war***. (That's a pretty obvious requirement for any kind of epidemiological study, perhaps too obvious in that it isn't made abundantly clear to the general media and readers).
-
Proactive survey defense
[Read the article: Study: 655,000 Iraqis have died as a result of war]
[Read more letters about this article: Here]Regarding the "small size of the sample" question which came up with the last Hopkins survey:
There's a standing joke among statisticians; the skeptic who proclaims "Why, this estimate is based on only a sample! And a random one at that!" The fact is, that it can be shown mathematically that the increase in accuracy from increasing sample size virtually levels off after a certain point, and is not related at all to the size of the population sampled from. Which is something that everybody actually knows without knowing they know it. After all, whether testing a drinking bottle or a swimming pool for chlorine, you only take a little test tube sample; you don't have to haul out a bucket full from the swimming pool because it's so much bigger. The underlying sampling principle is exactly the same.
-
let a thousand powerplants bloom
[Read the article: Calculating the global warming catastrophe]
[Read more letters about this article: Here]The answer is diversification. Since we can't store electricity, the electrical grid has to be scaled so as to supply the biggest load that routinely occurs. The biggest load on the electrical grid is summer air conditioning. What a happy coincidence then that that's exactly when solar power is most abundant. Stick solar cells on every rooftop to help power the AC and the grid can relax; that negates the unfortunate fact that solar power is way more expensive per watt than any other source. Similarly, where there are winds and nobody to NIMBY, put windmills; where there are tides put tidal generators; where there are geysers put geothermal generators; where there are mill ponds put hydro generators; etc. Where there are existing nuke plants without obvious flaws (like straddling a fault line) bring them up to current standards, but don't build any new ones, that route is finanacially and environmentally a loser. Keep the new and clean fossil fuel plants running to fill in the gaps.
One main reason this isn't done is that it goes in the opposite direction from the economic trend to centralize power generation and distribution. But technically, digital technology should be adquate to coordinate and control the system. Consider the radio-telephones of decades ago, with a big radio transmitter/receiver attached to a huge antenna that could handle a few dozen mobile phones over a far-flung area, versus the current cell phone system with a myriad of small local systems coordinated by a competent digital network, handling millions of phones across the nation.
-
steady reliable power
[Read the article: Calculating the global warming catastrophe]
[Read more letters about this article: Here]"The main dilemma with generating solar and wind power is they are not constant or predictable"
Between scheduled and unscheduled downtime, a nuclear power plant isn't the simple reliable 24-7 maintenance free gadget it's been advertised as. If it's up 85% of the time, you're doing real good. There's lots of places where the sun or wind are more reliable than that.
Imagine that a circuit breaker might explode in a nuclear power plant, causing a small local fire in the circuit breaker area which shuts off external power input, cutting power to the lubrication system for the turbine. Imagine further: what if the backup power for the turbine lubrication pump doesn't cut in because of a defect in its circuit breaker? That might cause the turbine to grind to a stop so quickly it bends the turbine shaft and breaks the bearings. That might cause the unit to remain out of service for as long as four months, unscheduled, while it's being repaired. Yeah, I know, that's a lot of "what if"s. So imagine how unprepared the folks who get their power from San Onofre Nuclear Generating Station were when it happened on February 3, 2001 (http://www.nrc.gov/reading-rm/doc-collections/gen-comm/info-notices/2002/in02001.html).
That's without even mentioning predictable and potentially avoidable human error, like how, while repairing the damage, the operators managed to drop a 40 ton crane 40 feet into the turbine building. Not drop something from the crane, drop the actual crane 40 feet. But at least that might have been somewhat avoidable in the unlikely event that power company management would get serious about something other than cutting corners to meet quarterly profits; not like the completely random and chaotic chains of events and failures that occur in complex mechanical gadgets, from nuclear power plants to space shuttles. As we all know, this is hardly the only time a surprise took a nuclear plant offline for a chunk of time.
Meanwhile, the sun continued to shine and the wind continued to blow.