do you know science?

We're not talking about swimming pools. We're talking about people who haven't the energy for water pumps, or for transport, people who treasure an old plastic bleach bottle because it enables them to carry water miles to their homes. They have no electric bills, and they have no thermostats.

Think about living on 30 KWH of electrical energy per year.

John

But then we'd have to make our computers out of coconuts.

Thanks, Rich

*see "Gilligan's Island"

Am I allowed to use propane or LNG or kerosene or wood or charcoal for heating and cooking?

Thanks, Rich

In sci.skeptic John Larkin wrote: ...

In some regions there are more problems than a lack of electricity.

But now, a slight digression.

At the start of the 20th century there was a big bang in Siberia. Tunguska was so remote it took a few years for the central govt to send soldiers out there to see what had happened. It seemed a big meteor or comet head had flattened a large number of trees.

Some years later --in the 80s or maybe 90s -- various scientists made their way out to that part of Siberia to study the event and try and figure out in detail what had happened.

They took a lot of instruments and planned to do a lot of digging. But it turned out there were still a very large number of eyewitnesses to the event. I.e. people in their 80s, 90s and 100s that still recalled what they were doing the morning the big bright object made it's way over the horizon and went "ka-bang".

Some doc's made even later showed some of the even then still-living eyewitnesses -- little old people in their 90s, and 100s -- carrying huge bundles of firewood back home, something they did every day or two.

The area was so remote it didn't have hospitals or even regular doctors until the 90s.

It seems life in central Siberia for the past 100 y involved a lot of hard physical work, eating from back yard veggie gardens, and missing out on TV, central heating, and household plumbing and electricity.

But end of diversion.

It's been long known that comparable 1st-world countries with modern energy generation have interesting correlations between energy consumption and longevity.

Admittedly, I am not a fan of "logevitity" calculations, but these correlations exist none-the-less.

Failing having more recent data to hand, I have the consumption in coal equiv per capita for 10 countries as well as their 1st-world longevities in 1995 (Economist pocketbook):

fossil cons combined m&f coal equiv longevity country kg/cap years

australia 294.0 77 us 267.4 76 uk 192.7 76 japan 152.9 79 france 130.8 77 germany 88.2 76 netherlands 77.9 78 spain 70.7 77 canada 62.4 77 italy 33.0 76

An OLS finds the following: y = -0.00364283*x + 77.1833 limits for beta at 90.0% CI tc = 1.89458 at 7 d.f. beta in -0.00364283 +- 0.00301629 = [-0.00665912, -0.000626533] Probabilities: P(beta95% confidence. The country-to-country variation in fossil consumption corresponds with 43% of the variation in life expectance -- i.e. fossil energy consumption is likely 1 of 2 indep variables that explain the variation in life expectancy between 1st-world countries.

--
[Full metal rebuttal:]
Not true.
  -- John Stafford , 08 Dec 2010 10:16:59 -0600

Access to clean drinking water is usually far more important.

This is a pathetic level of correlation and right down in the noise. The observed longevity of a few people who have survived for a lifetime on near subsistence levels ignores their typically very high levels of infant mortality. It is survival of the fittest writ large.

Looking at these figures I am surprised that the Canadian figure for fossil fuel usage is so much lower than either US or UK.

It is worth noting though that seriously reduced calorie intake is strongly correlated with longevity in many species and is worth about

40% increase in lab mice. Whether you want to starve yourself to live longer is another matter and it has yet to be confirmed in humans. It is certain that overeating, obesity and lack of exercise all combine to shorten life and these do correlate strongly with the developed world.

I have to say that I think you are clutching at straws here.

If there is evidence for anything it is that Japan and the Netherlands have better healthcare systems for their entire population than other first world countries. Hence on average they live longer.

Regards, Martin Brown

In sci.skeptic Martin Brown wrote: ...

...

You are -- as always -- free to provide hard evidence for your various speculations. :)

I have pointed out a well-know correlation, not an explanation.

But perhaps -- along the lines you seem to suggest:

It may be those developed countries that have high levels of energy consumption per capita are also those where individuals tend to drive over the street to the supermarket and gobble down double hamburgers wedged between pairs of grilled cheese sandwitches, washed down with a litre of supersaturated sugar solution, then to return to the apt for a 24 hr xbox marathon.

--
[Stage 1 denial:]
There have been floods before and there will be floods again. The
Austrailian floods weren't even record breaking.
[Later turns out Monkey judges all Australian floods by height of Brisbane R].
  -- Monkey Clumps , 30 Jan 2011 15:13 -0800

What a bunch of gobbledy-gook.

The only correlation here is that the inputs (on the left) have no effect at all on the outputs (on the right).

You must be a liberal.

Thanks, Rich

In sci.skeptic Rich Grise wrote: ...

...

So that makes you a creationist?

--
"Denialism"? Is that a word? 
  -- Mickey Langan, Sun 2 Jan 2011 4:50 pm

You're only looking at a tiny part of the available data. Where's Haiti, Bangladesh, Africa, Asia, South America?

The AGW drive to reduce CO2 is killing people worldwide by making energy, and now food, less affordable. Making corn into methanol isn't helping the kids in Benin.

John

Hmm... having done an experiment along these lines, I can tell you that -- at least for me -- well before the house hits 50, while my body can stay plenty warm and I can certainly read or watch TV or whatever, my fingers are cold enough that I lose the dexterity to type particularly well!

--------------------------------------------------------------------- ...

Compare apples and oranges? Why would I do that unless I didn't want to find a clear association?

See the first sentence again. I've underlined it this time.

I'm not denying that electricity has its uses, but at some point energy consumption becomes a proxy for overconsumption.

--
I will not be dictated to by others.
  -- Hosni Mubarak, Egyptian "dictator", Feb 11 2011

I don't imagine you can live on 30KWH per year without a little kerosene to go with it. It's interesting there's about the same number of obese and underfed people in the world.

"In Nauru, 70 percent of the population is classified as clinically obese"

-Bill

It might be "well known" correlation but it is clearly pure bollocks. The data you quoted yields quickly to ANOVA and there is much stronger "evidence" of a square law dependency on fossil fuel useage thanks to Japan with its exceptional longevity being in the middle of the pack.

Sorted into x order and plotting ASCII art to make it easily visible (with 75 years subtracted).

coal equiv fossil cons country kg/cap years italy 33 76 ## canada 62.4 77 #### spain 70.7 77 #### netherlands 77.9 78 ###### germany 88.2 76 ## france 130.8 77 #### japan 152.9 79 ######## uk 192.7 76 ## us 267.4 76 ## australia 294 77 ####

If the Germans were fitters and Australians were as unfit as Americans the fit would be even better.

Formal ANOVA results for Chebyshev fit are Least Squares Chebeshev fit coefficients ANOVA T(x) normalised to -1,1 a b c

9 a 77 8.9 a 76.9 0 0 8.84878 a+bx 76.889 -0.12678 0 7.87461 a+cx^2 76.705 0 0.56106 7.830195 a+bx+c2 76.6989 0.12963 0.61402 Allowing a to change from 77 to 76.9 explaings 0.1 of chisq. Allowing b to vary fits an additional 0.052 Allowing c to vary fits an extra 1.07 that is 20x more significant.

Japanese lifestyle is mostly quite sedentary so I would hazard a guess that most of the damage comes from vast over consumption of junk food. This is becoming a problem in the developing world too.

Regards, Martin Brown

...

I'm sure you can find any kind of polynomial fit that might work better than a corrected LS

But then, you've moved to "exploration" rather than hypothesis testing.

--
Ah yes, a very strong brand of anecdotal evidence. Climate science will
be proud. The science remains settled, the consensus unshaken. 
  -- Mickey Langan, Sun 2 Jan 2011 10:00 am

One gotcha is that the US counts infant mortality differently than most other countries do.

Another problem in the USA, in my opinion, is that we have a large fraction of the population that is descended from folks who are not adapted to the European diet of unlimited meat, wheat, and dairy products, the stuff that really tastes good.

John

In sci.skeptic John Larkin wrote: ...

There are many problems in analysing such cross-national data. Not the least of which is isolating the numerous effects.

If you suspect a priori that one or two data points that may be of interest have quite different characteristics, you are free to leave them out. But if you then start finding subset of points that support your hypothesis best you have ended up "exploring" and not hypothesis testing.

This distinction is usually not covered well in science-oriented stats courses.

An "exploration" of various possibilities -- that maybe ends up finding something ostensibly significant -- needs to adjust the relevant "degress of freedom" to take into account all those data and results that have been ignored before finding the "significant" result. Usually this ends up showing the "signfificant result" is actually not significant. Hence "exploration" and experimental curve-fitting are not typiucally useful methods for determining an underlying model to explain the data, although they can certainly be used to describe the data (of course a completely different thing).

In the present case (as far as I'm concerned, anyway :) the hypothesis that "some range of energy consumption is a proxy for sloth" we have seen there is *some* data that indicates there is a prima facie case. Even for "bollocks" posters the original data showed the same tendency, albeit one ignored in favor of continuing to look for another pattern that was more paletable.

But returning to the the basic principles of science, we should be asking if the thesis is true, it should be robust. Does the same pattern show up in completely different (but logically equivalent) data?

And -- of course -- it does.

Again, my datasets are taken from old pocketbooks rather than the Internet. The fossil fuel burning dates back to 1990 from the Cambridge Factfinder. And the lifex male/female from the Economist pocketbook 1995.

Before I looked at fossil consumption per capita for some developed countries. Naturally, the biggest slice of this corresponds with transport. More transport generally means less exercise, and that should have some health impacts.

But in the next dataset I have fossil-generated electricity per capita.

First, fossil electricity against male life expectancy (this time, not rounded down to whole years like one page of the Economist pocketbook):

DATA:

fossil elec prod kwh/cap fem life ex 1990 1995 switzerland 174.478 80.1 sweden 610 79.7 france 859.33 81.9 spain 1794.99 80 belgium 2722.7 78.8 germany 3567.49 79.5 canada 4187.31 81.3 uk 4281.18 77.4 netherlands 4549.6 79.7 finland 4883.4 80.2 us 8280.94 78.2

Model: y = -0.000251015*x + 80.5286 limits for beta at 90.0% CI tc = 1.83311 at 9 d.f. beta in -0.000251015 +- 0.000290867 = [-0.000541882, 3.98522e-05] P(beta90% and the usual automatic transforms my s/w makes to normalise the dep var didn't succeed.

male life ex 1995 switzerland 174.478 73.4 sweden 610 73.8 france 859.33 73.9 spain 1794.99 73.3 belgium 2722.7 73 germany 3567.49 73 canada 4187.31 75.3 (*) uk 4281.18 71.5 netherlands 4549.6 73.1 finland 4883.4 72.8 us 8280.94 71.1

(*) 1-sigma outlier

Model: y = -0.000268363*x + 73.9852 beta in -0.000268363 +- 0.000240568 = [-0.00050893, -2.77951e-05] P(beta rc (0.534000 2-sided) at 5%; reject independence

Here the Spearman agrees with the T-test and we have some greater confidence there is an actual relationship between fossil electricity generation and male life expectancy.

From social science we might suspect a priori anyway that males life ex would tend to show a greater sensitivity to any kind of health effects of an indep var -- ignorning warning signs and not going to the doctor is almost SOP for western males.

The 2 datasets show a not unsprising agreement -- for each kwh/cap of fossil generated electricity there is an associated reduction of around .0003 years of life expectancy.

This shows reasonable agreement with the original "all fossil energy consumption in TCE" from before, despite the fact electricity is a rather small proportion of all energy consumption in developed countries.

And just as a note, in each dataset the US is not seen to be an outlier. But if it is removed the -ve beta remains but the T-test prob drops below 80%.

--
[for certain range consumption is proxy for sloth]
It might be "well known" correlation but it is clearly pure bollocks. 
The data you quoted yields quickly to ANOVA and there is much stronger 
"evidence" of a square law dependency on fossil fuel useage thanks to [...]
8.84878		a+bx		76.889	-0.12678   0	
			[the -.12 seems to confirm the original thesis]
7.830195	a+bx+c[x]2		76.6989	0.12963	0.61402
  -- Martin Brown , 25 Feb 2011 15:00 +0000

.

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

Surely you are not suggesting that Androcles is a physicist!

Androcles believes with all of his shrunken heart that a particle in uniform circular motion with respect to an inertial frame of reference is undergoing _tangential_ acceleration.

He does not uinderstand a direct demonstration made by taking two derivatives, since he also believes that calculus is not well founded logically, since it requires division by zero.

He has _fixed_ calculus with his "hypothesis" that there is a smallest number h greater than zero. (He means a real number, not an integer, although he could not define "real.")

When presented with a counter-example, such as h/2, he replies that that is not allowed, since it violates his "hypothesis." This, he says, is "proof by contradiction."

Uncle Ben

About fifty years ago, Abraham Robinson put infinitesimals on a sound mathematical footing, which avoids those sorts of problems. It does violate the Archimedian order principle, iirc.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

Well, a boat with a mast (presumably a sailboat) might be 'fast' running before the wind. In which case, it's NOT in a vacuum, and the marble will hit in front of the mast. It seems to me that these confounding factors are given, albeit cryptically, in the statement.