Gibbs entropy and dynamics
G. Piftankin and D. Treschev
Let M be the phase space of a physical system. The dynamics
is determined by the map T : M-->M, preserving the measure
[mu]. Let nu be another measure on M, dnu=rho d[mu]. Gibbs
introduced the quantity s(rho)=[integral]rho log rho d[mu]
as an analog of the thermodynamical entropy. Attempts to
... [Chaos 18, 023116 (2008)] published Tue May 20, 2008.
Not that this is anything unusual.
But who is to judge? And who is
the willing consumer of such quick
"Studies show..." stories?
Still, there could be an accompanying
warning and with computers so ubiquitous
we might have the original data to work
in such stat programmes as R and others,
by ourselves.
But then it is the Mail and who has time
to challenge everything we read, or want to.
McLuhan's the-newspaper-as-a-warm-bath-of-society.
X-URL: http://www.guardian.co.uk/media/2008/jun/02/pressandpublishing.healthandwellbeing
Overhyped health stories? They're all pants
Peter Wilby The Guardian, * Monday June 2 2008
Last Tuesday, the Daily Mail informed readers that cocoa could be good
for the heart; that a diet pill will make you feel full as soon as you
start eating; that diseased gums increase the risk of cancer; and that
a third of babies whose parents smoke at home will end up in hospital.
That was just the news pages. In the health section, it had underpants
that control blood pressure, tree bark that eases arthritis, a herb
that relieves ear infections and peanut butter that stops hiccups.
Whether Mail readers rushed out to buy cocoa, peanut butter and
underpants, and to strip nearby trees, I do not know. But newspapers
believe health coverage attracts readers. We live in a medically
anxious society: surveys suggest one in two Britons worry about their
health against one in 10 in the 1960s. At some level, newspaper
reports must influence eating, drinking and buying habits, and affect
the wellbeing of readers and their families. Yet the press, sceptical
about anything politicians say or do, becomes credulous when faced
with medicine. All research studies are "authoritative", all medical
journals "prestigious", all scientists "experts", all findings
"breakthroughs".
Research findings rarely justify the categorical press reports, as a
new section on the NHS Choices website, Behind the Headlines, shows.
Each day, it analyses two stories, looking at what the papers said,
comparing their accounts with the studies on which they are based, and
assessing the validity of the research itself. The treatment is
factual and unhysterical, and it raises important questions about
journalism.
Take that heart-protecting cocoa. The research, from an American
journal, studied patients with diabetes, as the Mail made clear. The
cocoa used is not available to buy, as the Mail didn't make clear. Nor
did the research cast direct light on cardiovascular risk; it looked
at the effect on the main artery in the upper arm when cocoa is
enriched with flavanols, a type of antioxidant. Fruit and vegetables
also contain flavanols and, unlike most cocoa products, aren't full of
fat and sugar. Most important, neither the Mail nor other papers that
covered the study explained it was funded by Mars.
Another recent story concerned celery which, it was reported, could
help victims of high-speed car crashes and brain diseases such as
Alzheimer's. In fact, according to NHS Choices, the study, carried out
on mice, merely established that luteolin, a compound found in celery,
reduced the production of one molecule associated with brain
inflammation.
A third story was about that familiar bogey, the mobile phone. Mothers
who used handsets during pregnancy, it was reported, risked their
children suffering hyperactivity and emotional difficulties.
There were two big problems, common in reports of science and
medicine. First, the press implied a causal connection. The research
proved no such thing; for example, mothers who use mobile phones a lot
might be mothers who give their babies insufficient attention. Second,
all papers reported a 56% increase in risk for the children. This
sounds big, until you realise hyperactivity isn't that common and,
therefore, 95% of the children whose mothers used mobiles were
entirely unaffected.
Journalists would argue that their stories include caveats, plus the
words "can" and "may". Besides, nobody will come to much harm from
celery or cocoa and, if parents stop using mobiles, who cares? I don't
think this will do. Readers look at headlines and intros and often
don't get to the caveats, and such stories can create hope or anxiety
where neither is justified. The best-known example concerns research,
published in the Lancet, which linked the MMR vaccine with autism.
This study of 12 children got coverage far beyond what it merited, and
led to campaigns in several papers. As Cardiff University's Tammy
Boyce reports in her book Health, Risk and News, the UK child
vaccination rate fell as low as 80% after the coverage peaked. In the
US, where only a handful of stories appeared, the rate never fell
below 90%.
So reporting of medical research can sometimes lead to damaging
consequences. It is not easy, though, to see a solution. The stories
analysed by NHS Choices were not exactly wrong or even distorted.
Rather, they misrepresent the provisional, tentative nature of
research. Science proceeds by gradual accumulation of evidence, not by
the sudden "eureka!" moments implied in the press.
Sir Muir Gray, the NHS chief knowledge officer, told me the medical
journals should share the blame. They too are under commercial
pressures and tend to favour "positive" findings, issuing excitable
press releases about them.
Medical research doesn't fit the conventional news frame, which
demands crisp intros, novelty and immediacy. Many stories analysed by
NHS Choices wouldn't have made the paper if they had been written up
correctly. So should the press leave the public in ignorance? Perhaps
it should. A tale about the supposed effects of celery is neither
informative nor particularly entertaining. As one health journalist
used to say, there are, in any event, only four medical stories: new
treatment gives hope; old treatment causes cancer; patients will
suffer unless doctors get more; and jogger's nipple, the novelty
story.
Health reporters, instead of churning out several inconsequential
variants on these four each day, would be better employed writing
"state of knowledge" features, on what scientists now think about the
treatment of particular complaints. And if they must write spurious
"news", perhaps they should regularly steer their readers to the sober
analysis on NHS Choices.
X-URL: http://newsvote.bbc.co.uk/mpapps/pagetools/print/news.bbc.co.uk/2/hi/business/7432092.stm
BBC NEWS
Biscuits 'key' to clinching business deals
About four out of five UK businesses believe the type of biscuit they
serve to potential clients could clinch the deal or make it crumble, a
survey says.
The outcome of a meeting could be influenced by the range and quality
of biscuits, according to 1,000 business professionals quizzed by
Holiday Inn.
The chocolate digestive was deemed to make the best impression
followed by shortbread and Hob Nobs.
Lawyers were most impressed by good boardroom biccies, the survey
added.
Dunking Do-Nots
Jammie Dodgers and Bourbons were also among the biscuit types thought
to help sweet-talk customers.
However crumbly biscuits are a big no-no in the meeting environment,
the questionnaire found, with 30% frowning on a regular digestive in
the work environment.
And when it comes to helping yourself to biscuits from a communal
plate, the most acceptable number to take is two, the research
concluded.
However more than half of respondents looked down on dunking biscuits
in tea or coffee during a meeting.
A survey released last year, which quizzed 7,000 people, suggested
that the custard cream is the nation's favourite biscuit.
Story from BBC NEWS:
http://news.bbc.co.uk/go/pr/fr/-/2/hi/business/7432092.stm
Published: 2008/06/02 16:41:40 GMT
Language Log, a consortium of
Anglo-phone linguistic scholars
goes off the deep end again over
the rhetoric of science coverage.
Do people even get the later item?
Who members more than the first para?
X-URL: http://languagelog.ldc.upenn.edu/nll/?p=184#more-184
[4]Language Log
[8]Parrots and children: still silly season at the BBC
May 23, 2008 @ 11:02 am? Filed by [9]Geoffrey K. Pullum under
Yesterday the BBC [13]published a story (thanks to Sam Tucker for the
reference) about a stray red-tailed African grey parrot that told
police how to find its owner. ....
The story is probably true in outline. No one disputes that parrots
can be trained to do a pretty good acoustic reproduction of a human
utterance. And they will do just as well on an address as on a line
from a children's song or a few verses from the Kor'an; the content
doesn't matter -- for them, there is no content. But a paragraph at
the end of the story reveals that the BBC still brings out its most
gullible writers (or perhaps its most cynical and dishonest writers)
as soon as anything to do with the cognitive or linguistic sciences
comes on the scene. The last para says this:
The African Grey parrot is considered one of the most intelligent
birds and is said by experts to have the cognitive ability of a
six-year-old.
They mean a human of age six. There are people writing purportedly
serious stories for the British Broadcasting Corporation who think
that a grey parrot has the cognitive ability of a normal six year old
human child. Have these people never met a normal six-year-old human
child?
....
Perhaps, as Mark once suggested, at the BBC [14]they just don't care.
For now, I'm inclined to think that the BBC's coverage of science,
especially cognitive and linguistic science, still deserves the
ridicule that has been heaped on it in the past here on Language Log
([15]this post on Language Log Classic offers a sample, and lists a
few other posts on BBC science reporting).
----- End forwarded message -----
