Posts by Another Ben

Ministry of Justice IT bods to strike over outsourcing fears

Re: Just strike on principle!

The "incredible savings" seem to become credible only to the people who have sat through all those boring meetings. To those who rely on the services in question, they remain incredible.

EDF: We'll raise bills 11% - but only 2% is due to energy costs!

How to lie with statistics

If you actually look at the relevant figures, which are not even in the footnotes to the press release but on <a href="http://www.edfenergy.com/products-services/for-your-home/customer-commitments/what-

makes-up-bill.shtml">this web page</a>, which is cited (thought not actually linked) in the footnotes.

The cost of environmental and social schemes may have increased by 50%, but this is from a base of 4% (for gas) or 8% (for electricity) of the cost to the consumer. The transmission costs have apparently risen by 9% on a basis of 25% (for gas) or 22% (for electricity). So the contributions to the overall increase are pretty similar, and appear to leave something of a gap still to be explained.

To get from this to 'The other 9 per cent, the great bulk of the increase, is mostly down to "renewable, energy efficiency and social schemes" and partly from "transmission and distribution charges".' is quite a leap, rather than simply a restatement "in other words".

So we have a self-serving EDF press release reported by a journalist with an axe to grind. Just as well no-one is actually reading it instead of trotting out the same comments they make on every other article on energy or climate change.

UK lays carbon plan before Earth Goddess

Not exactly

The 7% projected reduction is "compared with what they would have paid in the absence of policies", not "compared with current prices".

Global warming much less serious than thought - new science

One small step for science, one giant leap for this reporter

So new research concludes that the range of likely warming (with 66% confidence) is between 1.7 and 2.6 degrees, most likely 2.3, instead of between 2 and 4.5 (most likely 3). It also concludes that warming of over 10 degrees is unlikely. Hurrah! But remember that 2 degrees is a fairly arbitrary limit, a compromise between caution and political achievability, and even if these newly published results are entirely correct there is still a substantial chance of missing even that target.

So where do you get "much less serious than thought" from?

Good on you for linking to the paper though. If only more journalists would do that.

Climategate 2.0: Fresh trove of embarrassing emails

All models are wrong

You seem to have accidentally misquoted Phil Jones. "All models are wrong" is an acknowledged fact about empirical science, and perhaps an allusion to George Box: "all models are wrong, but some are useful". "All our models are wrong" carries a rather different message.

Move along please, no intellectual corruption to be seen here...

Short domain land-rush coming to .uk

No it isn't

RAC, formerly the Royal Automabils Club: www.rac.co.uk

AA, the Automobile Association: www.theaa.com

Fukushima fearmongers are stealing our Jetsons future

Occam's razor

Occam's razor is the principle that one should prefer the simplest hypothesis that agrees with the available data. Its subjectivity is illustrated by the fact that you regard "there is no effect at low doses" as the simplest hypothesis, while I would probably pick "the effect is linear" as simpler given that we know there is an effect at higher doses. Neither is true, of course, and Occam doesn't really help much when it comes to uncovering the real picture, although advances in biology might even if epidemiology can't.

Assumptions

I can see the circularity of the argument you seem to assume I am making, but I'm not sure why you feel the need to choose one assumption or the other. An open mind is quite acceptable in some circumstances. It would seem an amazing coincidence if the effect of low radiation doses were precisely zero. It might even have a net beneficial effect in some cases. My point is that quite significant (in the life-or-death sense) effects do not necessarily result in significant (in the statistical sense) signals, and that glossing over such issues for the sake of a rhetorical point is not a prerogative of the antinuclear fearmongers.

Logic and uncertainty

My point is that there is inevitably a large degree of uncertainty in the numbers of additional deaths, and the fact that epidemiologists have not picked out an actual signal puts only a rather loose upper bound on the size of the effect.

Chernobyl: 4000 "unmeasurable" deaths

Maybe I am using the wrong news sources, but I have not seen any coverage that is as hysterically anti-nuclear as Lewis Page's is hysterically pro-nuclear. I find his dismissal of the consequences of Chernobyl rather glib.

If radioactive caesium from Chernobyl "didn't cause any measurable health consequences" this is largely due to the difficulty in making such measurements. It is difficult to count additional deaths resulting from a relatively small increase in risk across a large population, but that doesn't mean they are not real. The Chernobyl Forum's 2005 report estimated 4000 additional deaths as a long-term consequence of the Chernobyl accident. They deliberately looked only at the most exposed population, in which the radiation doses are in a range that is better understood than the low doses experienced farther afield.

Second explosion rocks Japanese nuke plant

6570 times background radiation

Actually the article says that "hourly radiation levels around the plant have settled back to 75 per cent of normal yearly exposure" so you get your extra 9 months' worth in one hour. Which is not "insta-blast" level but is not great, especially if you are there for significantly longer than that.

Mobile phone users have fewer brain tumours

Work in progress...

Randomly select thousands of people and have the networks share their usage data? Surely that will never happen. Oh wait... http://www.ukcosmos.org/ ... "There are now 30,642 participants registered to take part in the study."

Yes, then there are other risk factors to unpick, but at least this study isn't relying on our dodgy estimates of how long we spend on the phone.