|
|
|
|
|
|
|
|
|
With Colin Blakemore
Colin Blakemore : Life's a risky
business,storms,road accidents,even being struck by lightning,the worst
can happening many ways.We continually weight benefits against the likelihood
of disaster.But how do scientists assess the risks of new technologies like
GM food [Ref: Miller] and
mobile phones .How dangerous is
BSE ,and how safe is vaccination? To discuss the
scientific risks we face and how we manage them,I'm joined by Sir Colin
Berry,Professor of Pathology at the Royal London Hospital,Kenneth
Macrae,Professor of Medical Statistics at the University of Surrey,Lynn Frewer
who's head of Consumer Science at the Institute of Food Research in Norwich
and one of our regular contributors Mike Maier,psychiatrist and senior lecturer
at Imperial College London. Colin,what is risk?
How do we define it?
Colin Berry : Well,it's perhaps easier to begin by saying what it
isn't.It isn't hazard,which a lot of people confuse it with.Mountains are
dangerous and the sea's dangerous,but none of us here in the studio are likely
to be bothered by mountains or the sea at the moment.If we went swimming
in the sea,or if we tried to climb the mountain,we expose ourselves to a
risk,if we climb the mountain in the winter,the risk will be different from
the summer.If we go swimming when we're drunk the risk will be different
and so on.
[First point : Comparative risk is obviously an existent phenomena,and Colin
has characterised the intuitive notion of it.No doubt everyone would admit
it's more dangerous to swim when drunk,so how do you measure this variation?
Let's see.... -LB]
So risk defines a particular circumstance.If you climb the North face of
the Eiger a thousand times,what's your chance of falling off one of those
times? The hazard is the mountain,the risk is the statistic if you like.
[News flash: An ITV reporter from Westminster says regarding paedophiles
and the Sarah Payne case that the government wish to "assess the risk" .The
BBC report said similarly that the data had to be "risk assessed" -LB]
Colin Blakemore : And is it really...is it really correct...I mean
who was it? Ulricht Bech [see Bornholdt- Beck] ,a sociologist said that we
are a "risk society", that risks are increasing
and we're more aware of them,are there more risks in life?
|
SIR - Beck-Bornholdt and Dubben (Nature 381, 730;
1996) describe a common misinterpretation of the P-value of a classical
statistical test using the following example. The chance that a randomly
chosen human being is the Pope is about 1 in 6 billion. John Paul II is the
Pope. What are the chances that John Paul II is human? By analogy to syllogistic
reasoning, Beck-Bornholdt and Dubben suggest 1 in 6 billion but point out
that this is ''obviously not sensible".
It certainly isn't.The probability of data given a hypothesis, P(D |H), is
not the same as the probability of the hypothesis given the data, P(H |D).
This is an elementary error regardless of one's preferred statistical approach.
Bayesian
[Ref: Video: N30 The Numbers Game {Bayesian Statistics};Red File1:
Newsci1.wri] statistical inference, which includes syllogistic reasoning
as a special case, is particularly well-suited for avoiding this sort of
mistake (H.Jeffreys, Theory of Probability, Oxford University Press, 1939)
.
A similar pitfall is the infamous 'prosecutor's fallacy', in which a
probability that a DNA fingerprint match would occur in someone other than
the true criminal - P(match | innocent) - is used incorrectly as the probability
that a suspect is innocent - P(innocent | match).
In a city of ten million people, a one-in-a-million DNA fingerprint match
will give ten other people the same fingerprint as the true criminal. In
the absence of other evidence, the odds that the suspect is innocent are
better than 90%, not one in a million. Let H represent the class (hypothesis)
of humanness, A represent the class (hypothesis) of alienness, and J represent
the observation (data) that a randomly chosen individual is Pope John Paul
II.
Bayes's
theorem tells us how to infer the probability that the Pope is human,
P(H | J): P(H | J)= P(J | H)P(H)
P(J | H)P(H) + P(J | A)P(A)
Thus, to infer the probability that the Pope is human, P(H
| J), we have to have two more numbers in addition to the probability P(J
| H) of drawing the Pope at random from the class of humans:
(1) P(J | A), the probability of choosing the Pope at random from the class
of aliens.
(2) P(A), the a priori probability that a randomly chosen individual is an
alien instead of a human.
P(H) is just 1 - P(A), if only these two hypotheses are considered. Presumably
the probability P(J | A) of choosing the Pope from the class of aliens is
infinitesimal. The prior probability of choosing an alien as opposed to a
human, P(A), is also expected to be quite small,except perhaps near secret
US Air Force bases. As either P(J | A) or P(A) approaches zero,the
probability that the pope is human approaches one.
It is a shame that Bayesian methods are not part of all introductory
statistics classes. In this case, they quickly reassure us that the Pope
is (probably) not an alien.
Sean R. Eddy Department of Genetics, Washington University School
of Medicine, St Louis, Missouri 63110, USA
David J. C. MacKay Cavendish Laboratory, University of Cambridge,
Madingley Road, Cambridge CB3 0HE, UK
Sir - In their syllogistic reasoning to present a papal paradox, surely
Beck-Bornholdt and Dubben's sequence is incorrect. It should be as follows:
(1) If an individual is human s/he is probably not the Pope (premise);
(2) John Paul II is human (premise);
(3) therefore John Paul II is probably not the Pope (conclusion).
In fact, of course, he is the Pope, but probability allows for this possibility
(it was just extremely unlikely) and the logical progression holds. Their
sequence made the classic mistake of assuming that "if A, then not B" implies
"if B, then not A" , which can be disproved by many examples more trivial
than the papal paradox.
Indeed, if applied to their first example, it would imply that all mortals
are human, a hypothesis easily disproved by considering any other living
being. If the logical sequence they describe for statistical hypothesis testing
is actually the one in use, then it too makes the same mistake, and is therefore
in error, but not for the reason they suggest. I think the conclusion in
the sequence should be " (3) therefore the null hypothesis is PROBABLY wrong",
with probability theory providing the degree of uncertainty, leading to
conclusions such as "it is extremely unlikely (5% probability) that this
result arose by chance" - but I'm probably wrong!
Stephen P. Gosden Chaussee de Wavre, B-116O Brussels, Belgium e-mail:
[email protected]
Sir - The argument of Beck-Bornholdt and Dubben can be paraphrased
as follows.We have a human-looking organism before us and wish to test the
hypothesis that it is in fact human. We know that if it is human,it is very
unlikely to be the Pope. We find out that it is indeed John Paul II, and
have to conclude that it is unlikely to be human.
In claiming that this conclusion (and with it, statistical hypothesis testing)
is absurd,Beck-Bornholdt and Dubben not only contradict one of their own
assumptions but also describe bad scientific practice. We use statistics
to test hypotheses because we know that the same set of data could arise
whether or not a given hypothesis is true. [See Ian Stewart in
chess-Falsifiability]
In making the comparison with statistical hypothesis testing, Beck-Bornholdt
and Dubben assume at the outset that a non-human could be the Pope.They cannot
therefore claim that the conclusion of their argument is ridiculous, merely
that it is factually wrong. Furthermore, this incorrect conclusion arises
only if the randomly picked human-looking organism happens to be the Pope.On
the many other occasions when we perform the experiment, we will quite reasonably
fail to reject the hypothesis that our subject is human. The same applies
to statistical reasoning: we will occasionally observe highly improbable
data, and incorrectly reject the null hypothesis. In their statement of their
argument,Beck-Bornholdt and Dubben have implicitly allowed themselves the
luxury of observing subjects until they find one that is the Pope. This procedure
is equivalent to a scientist, anxious to reject a null hypothesis, repeating
an experiment until a suitable result turns up.
B. J. Craven Department of Psychology, University of Stirling,
Stirling FK9 4LA, UK L.
S. Craven 49 Broompark East, Menstrie, Clackmannanshire FK11 7AN,
UK
Sir - Beck-Bornholdt and Dubben have shown at most that the Pope is
not a random sample. Their 'syllogism':
Premise: "If.. we randomly pick a human being, the probability that it
is the Holy Father is extremely low...."
Conclusion: "Therefore, if an individual is human, it is probably not
the Pope"
A syllogism states that an attribute shared by all members of a class
is possessed by each member. Their conclusion ignores the fact that the attribute
of probable nonidentity with the Pope was limited explicitly to the class
of randomly chosen human beings. If the sampling method is not part of the
class specification, why not just give as the premise "The Pope is, with
high probability, not the Holy Father," thus suggesting an even wittier
title for the published letter?
Statistical inference about membership in a hypothesized distribution applies
to data randomly sampled from the population of inference. The failure of
inference here results from no philosophical paradox about probability, but
only from verbal shiftiness about the population and the sampling method.
James C. Nelson Department of Plant Breeding and Biometry, Cornell
University, Ithaca, New York 14853, USA e-mail
: [email protected]
NATURE VOL 382 . 8 AUGUST 1996
|
Colin Berry : I think in practical terms there are very many less.We're
all living longer and we're healthier while we're alive.If we were having
this discussion two generations ago,many women would have lost a child ....they'd
have had more children and it wouldn't have been exceptional for anyone have
them to have lost a child while that child was under five say.That doesn't
happen almost now.
Colin Blakemore : Ken,is it that the risks are increasing,I mean there
are more hazards in life as technology changes and so on,but we're just better
at avoiding them,that's why we live longer.
[I doubt there are more hazards in the sense Colin has defined them,sure
we have cars and nuclear power stations,but we're unlikely to be attacked
by a wild predator,all that's happened is which threats are uppermost has
changed,and our ability to counter others has increased.Typhoons and "Acts
of God" have always been a threat,but meteorological science and super computers
have even lessened the chances of being killed by a storm or a an
avalanche [Ref:
Horizon].
The idea that the past was safer or that halcyon days existed in the past
is a fabrication,people were more likely to die of pestilence and disease
in the past than now.If you check the average age of a person a couple of
centuries ago they were dying at 30-45,now we can expect to live to 90+.
Rick Wakeman has just related the origin of the phrase "Don't throw the baby
out with the bath water". In the 15th century people mostly got married in
June as they were still vaguely clean from their yearly bath in May.The hierarchy
of using the same bath water started with men,then young men, women and girls
finishing with babies who were washed in the dirtiest water.It's no wonder
they caught diseases and had a high mortality rate.How true this is I don't
know,but I wouldn't like to live in those times -LB]
Kenneth Macrae : I think we're much more aware of risks now,because
there are so many more people professionally employed to study risks.The
New York Times for example is almost like an epidemiology journal,they like
nothing better than a story about a new terrible thing that might happen.
[So we don't live in a more dangerous society,we just make more of the risks
that do exist. The emphasis on doom-laden scenarios,especially in the US,perhaps
stems from apocalyptic (see Frayling doc) thinking
borne of Revelations,and our history of distrust of
science (see world30) means that we're likely
to be pessimistic as opposed to optimistic -LB]
|
1. Something Went Wrong in Jet Crash, Expert Says
2. Police Begin Campaign to Run Down Jaywalkers
3. Safety Experts Say School Bus Passengers Should Be Belted
4. Drunk Gets Nine Months in Violin Case
5. Survivor of Siamese Twins Joins Parents
6. Farmer Bill Dies in House
7. Iraqi Head Seeks Arms
8. Is There a Ring of Debris around Uranus?
9. Stud Tires Out
10. Prostitutes Appeal to Pope
11. Panda Mating Fails; Veterinarian Takes Over
12. Soviet Virgin Lands Short of Goal Again
13. British Left Waffles on Falkland Islands
14. Lung Cancer in Women Mushrooms
15. Eye Drops off Shelf
16. Teacher Strikes Idle Kids
17. Reagan Wins on Budget, But More Lies Ahead
18. Squad Helps Dog Bite Victim
19. Shot Off Woman's Leg Helps Nicklaus to 66
20. Enraged Cow Injures Farmer with Ax
21. Plane Too Close to ground, Crash Probe Told
22. Miners Refuse to Work after Death
23. Juvenile Court to Try Shooting Defendant
24. Stolen Painting Found by Tree
25. Two Soviet Ships Collide, One Dies
26. Two Sisters Reunited after 18 Years in Checkout Counter
27. Killer Sentenced to Die for Second Time in 10 Years
28. Never Withhold Herpes Infection from Loved One
29. Drunken Drivers Paid $1000 in '84
30. War Dims Hope for Peace
31. If Strike isn't Settled Quickly, It May Last a While
32. Cold Wave Linked to Temperatures
33. Enfields Couple Slain; Police Suspect Homicide
34. Red Tape Holds Up New Bridge
35. Deer Kill 17,000
36. Typhoon Rips Through Cemetery; Hundreds Dead
37. Man Struck by Lightning Faces Battery Charge
38. New Study of Obesity Looks for Larger Test Group
39. Astronaut Takes Blame for Gas in Spacecraft
40. Kids Make Nutritious Snacks
41. Chef Throws His Heart into Helping Feed Needy
42. Arson Suspect is Held in Massachusetts Fire
43. British Union Finds Dwarfs in Short Supply
44. Ban On Soliciting Dead in Trotwood
45. Lansing Residents Can Drop Off Trees
46. Local High School Dropouts Cut in Half
47. New Vaccine May Contain Rabies
48. Man Minus Ear Waives Hearing
49. Deaf College Opens Doors to Hearing
50. Air Head Fired
|
You know,Claude Colburn,when he was on The Times,in his autobiography
mentioned that the journalists had a competition to write the most boring
headline of the week, such as "Small earthquake in Chile,not many killed"
,and what people like reading about are catastrophes,disasters bad news,they
sell newspapers.
[Do they? Or is that all newspapers are prepared to print as news? -LB]
Colin Blakemore : Lynn,do you agree with that? People are fascinated
with the macabre?
[I think it's fair to say that the head turning that goes on with motorway
pileups and the crowd that gathers round a killing and the Hollywood movies
that graphically portray death,make us seem obsessed with with sex,death,money
and goods. When something good happens I guess this is subject to the notion
that if things are okay then no intervention is necessary,it's only when
circumstances change for the worst that we need to change our behaviour.
Perhaps its also because of the rarity of failures that we are intensely
curious about them when they do happen,and that's why in the case of the
concorde disaster there was such coverage. It may also be that tabloid
journalists are obsessed with gutter level navel contemplation,and a disaster
gives them a respite from their usual waffling about nothing - LB]
Lynn Frewer : They're very interested in these kinds of issues,but
I would dispute the fact that the media is creating this risk society.I mean
people are intelligent,
[Like Lisa Jardine in world30,Lynn is crediting
the populace with an intelligence that they don't appear to have (see
sci-mat).I think people in general do not understand
risk and this is brought out in their inability to compute probability as
shown by Sue Blackmore (see sci-mat3). Those
interested in paranormal phenomena are especially unable to do this and that's
why they see coincidences and miscompute likelihoods, giving them the idea
that things are less or more probable,depending on their view.For example
creationists think the the suppose improbability of the universe (see
trans -Hugh Ross doc) means that it requires a
creator,because they cannot conceive that an extreme improbability nevertheless
CAN happen.Oddly those same people play the lottery and expect to win -LB]
they look at the way risk is reported in the press,and it's a conflict between
different actors in the debate.All these different individuals who're providing
different messages.But I think if people have a fixed idea about a hazard
they'll just filter the information from the media so that it agrees with
what they already believe,and I don't think we can always accuse the media
of creating this....these risk problems,that we're discussing.
[Maybe not,but the media DO exploit innumeracy by emphasising a volume measure
over an area or linear dimension to make it seem worse than it is,or contrariwise
using linear measure to diminish the apparent size of something to play on
our fears. A 100 centimetres SOUNDS bigger than 1 metre and scale can make
something seem worse.Percentages can mislead by not mentioning absolute
numbers,or vice versa using large absolute numbers instead of percentages
or rates. Numeracy skills allow one to filter data effectively and see how
one is being conned in the news (see JA Paulos "A
Mathematician Reads the Newspaper") -LB]
Mike Maier : I think it is a construct that the media brings about
because as Ken said "bad news is good news.You don't hear about all the
successful things that happen,you do hear about the 20 food poisoning cases,you
don't hear about all the people who successfully had dinner without being
food poisoned,
[This is possibly another reason for the public fear of science.Biased reporting
makes it seem like nothing but failures happen. In all the recent food scares
the number of deaths is negligible, and yet the papers make it seem like
a scene from "Outbreak".When science facilitates a benefit it goes unreported
except in the scientific journals,which as it said in world30.html,are beyond
the lay language of the common person,though I doubt "New Scientist"
(www.newscientist.com) is above the comprehension of most people.Given that
journalists are invariably scientifically illiterate (excepting the specialist
writers) possibly the journalists own misinformed fears get passed to the
public.The journalists are also victims of the "I will never need maths when
I grow up" syndrome,and make fundamental errors in their recounting of scientific
data (see JA Paulos) -LB]
and generally I think we have a sophisticated society now,that is preoccupied
with trying to maximise their quality of life and their length of life,and
if there are risks and dangers...
Colin Berry : I think that's very important,I mean infant mortality
now in this country is 6 per thousand births.
[This is an example of what I mean. 0.6% sounds less than 6 per thousand
-LB]
Now that in the 1950s,would probably have been considered an unobtainable
figure.Now it's very....it's a remarkable change in a comparatively recent
time.But you know nobody produces articles saying "Isn't it marvellous how
low the infant mortality rate has become?".It is marvellous,but we tend to
disregard benefits,and I think one of the things that's missing in all the
discussion of risk is benefit,we never do "what's the benefit of a
pesticide,what's the risk of a pesticide?",we never do "what's the benefit
of changes in cars?" as opposed to "what's the risk of change?" and so on.
Colin Blakemore : Well,we all get bombarded with news about risks,what
do we actually make of all that information? Our reporter Dan Rowland went
out to find out.
Dan Rowland : I don't know about you,but I'm a bit confused about
what's good for me and what isn't.There seems to be a new health scare every
day in the newspapers,be it on beef or GM foods, mobile telephones or even
driving your own car.Well I've come down to London's busy Covent Garden to
see if the people here actually heed any of the numerous health warnings.
(Cellphone tune sounds) Are you worried about using your mobile phone?
1st Man : Sometimes yeah.After I've used my phone I can feel my ear
like hot,but I just don't think about it,'coz I need the phone.
[Your ear will heat up ANYWAY when pressed next to the earpiece of a phone,as
your own body heat fails to escape from your skin's surface and is insulated
by the surface of the phone,this will happen with conventional phones as
well -LB]
Dan Rowland : So why are you smoking?
2nd : Er,,just to relieve stress,that kind of thing.I've just finished
my exams.
Dan Rowland : What if the statistics suddenly said that,you know 1
in 2 cycle couriers in London got killed riding a bike,would you stop then.
[See www.statistics.gov.uk -LB]
3rd Man : Definitely yeah. There was woman killed last week on a bicycle.
Dan Rowland : Didn't that put you off?
3rd Man : It puts you off,it knocks you for the day,but it does happen.
1st Man : I listen to myself.If my ear...if I woke up and my ear was
really throbbing and I had bad headaches,then I suppose I'd cut down on using
it.
[This is idiotic.The whole idea of risk assessment is to avoid damage BEFORE
it happens.If this man is feeling physical symptoms he's probably already
damaged his ear and it's tool late to do anything about it. This line of
thinking is like saying " I'll wait and see whether an avalanche happens
and smashes the village before I decide not to live there". Notably politicians
like the "wait and see" approach as they don't wish to panic the public
unnecessarily. Seismic scientists and CDC scientists (such as Dustin Hoffman's
character in "Outbreak") are similarly in a position of being wary of making
wild claims.But as in "Outbreak" the bias should be towards saving life,which
means that if there is doubt it's best to play it safe.This alludes to the
"precautionary principle" described later.Politicians don't wish to take
notice of alarms until the very last minute in case they spend taxpayers
money unnecessarily.But history has shown that when they haven't heeded warnings
and played the "wait and see" game,they've had to pay out more in repairing
damage done that they might have prevented had they assessed risk properly
-LB]
Dan Rowland : Do you like beef?
1st Woman : I love beef! (she and her friend giggle)
Dan Rowland : Did you eat it all the way through the BSE crisis?
1st Woman : Yep.
Dan Rowland : Why?
1st Woman : 'Coz I like it.
Dan Rowland : Were you worried about CJD at all?
2nd Woman : No,I didn't even really think about it.
Dan Rowland : You didn't think about it at all?
[These are the people (who are NOT THINKING) credited as intelligent by Lynn
and Lisa -LB]
4th Man : I think that big business has too much say in policy making
quite frankly.
Dan Rowland : Under what circumstances would you not eat genetically
modified food? When would the risk factor be too large for you to want to
eat it?
5th Man : It's a very difficult question,because unless you're
a scientist,you don't understand what the risk factor is.
[Here we have an answer to the question posed in world30.html - "Should science
be left in the hands of the scientists?".This man recognises that there are
smarter people than he who understand such things and that people "don't
understand" risk factor.That's a bit odd coming from an "intelligent" person,but
then perhaps his intelligence is recognising that trained people are best
placed to advise on things that require specialist knowledge -LB]
3rd Woman : I don't like anything,I'm natural,I don't like messing
around with nature,but more than anything I don't like something that's tried
out on the public before it's been proven.
[This is a chronically ignorant view (and is indicative of the "intelligent
public). First of all there isn't anything that's not natural (see
natural.html),unless the suggestion is that man- made is something that couldn't
exist of its own volition.One cannot mess around with nature,and science
certainly is not "messing".As Channel 4 pointed out in "Secret History" the
myths of tribal cultures such as American Indians being at one with the Earth
and not manipulating nature is a far cry from what they actually did.In those
terms modernity is actually better than some of those ways.Certain tribes
of red Indian in that programme asked why other people had any right to stop
them killing whales,and furthermore they WERE killing whales contrary to
what Greenpeace would have us believe. Secondly I'd like to know what this
woman's idea of proven is if we can't do anything "unnatural" or use the
public to test things out on.As is shown in sci-mat1.html Michael Baum has
to receive assent from patients to test new procedures out on them in double
blind trials to see if they work. If we left it to this woman we'd do nothing
at all not be able to prove anything. Proving necessitates public participation
and doing "unnatural" things.She has contradicted herself -LB]
5th Man : If you believed everything that you read or heard in the
media,you'd be...you couldn't live your life you'd be so scared to go out
the door,you'd wrap yourself in cotton wool and never get out of bed.
6th Man : I'd stop using a phone if I could get work without having
a mobile phone.
Dan Rowland : As ever it seems to be risk versus convenience,and certainly
for most people we've spoken to here today,the old adage "you can't teach
an old dog -or even a new puppy - new tricks" seems to ring true. As for
me,well I could get run over by a bus tomorrow couldn't I? Excuse me (Dan
picks up his cellphone) Hi mum,yeah,yeah.
[The point is that the odds of being hit by a bus on any given day are quite
low -LB]
Colin Blakemore : Ken, what did that reveal? A certain amount of risk
fatigue really didn't it?
Kenneth Macrae : Oh absolutely.I think first of all there is a general
air of cynicism because scientists say one week that something is dangerous
and a few weeks later they contradict themselves
[That's not quite true.What you hear is various views from different scientists
who may have different agendas or different data. The poll later shows that
the public trust environmental scientists as if they were impartial or looking
after the public's concerns.In actual fact they have perhaps more of a political
agenda than anyone else.In "Secret History" it became clear that environmental
groups have seconded the mythical notion of tribal peoples as being
environmentally friendly by virtue of their lack of technology as a means
to selling their ideas to the public,and that the view of a tribal person
as somehow better for the Earth is misinformed. If such people have less
impact,it's not through choice or belief,but through being limited in terms
of their technology. The media polarise the views that are put forward into
safe/not safe,and this gives the impression that scientists haven't got a
consistent view.Given that scientists can represent industry,the government
or be independent or represent a concern such as the environment,there will
be a cross section of views representing those interests.Also the nature
of the inquiry means that something is never "proven" in the way that the
woman above wished it to be.All that can happen is we become more sure than
not depending on context,and over time what was true in one situation might
not be in another.
The gross over simplification for the benefit of the lay person does not
give them due credit as "intelligent",and part of the problem is that people
do not understand complex issues because they don't understand the science
behind it,nor do they speak the lingo (see world30.html),and so a dumbed
down version is proffered so that it IS understood.This can overlook the
subtleties in issues to the detriment of consistency,and thus it appears
that there is no definitive answer,which usually there isn't anyway,that's
why "probability" is used,because no one can say for certain and "prove"
it one way or another incontrovertibly,it can only be "proven beyond a reasonable
doubt" as in court. Oddly Mo Mowlam is on Room 101 and has just referred
to "Red wine being bad for you one week and bad the next",as a government
member,she's obviously subject to the same oversimplification as the public,and
thus we can't presume that government members are any wiser than the public
-LB]
Also I think how people interpret risks is sometimes very unsophisticated,and
it's not necessarily their fault.Risk is actually confusing to people,because
it's complicated.It's not because they are in any way unintelligent.It's
just that they get mixed messages and it's not an easy concept to explain
in a straightforward way.
[I've now heard Lisa Jardine ( world30),Peter Dale
( toynbee8),and now Lynn Frewer and Richard Shepard
( sci-mat3) all attest to the public's ability
to make informed decisions,and yet the vox populi interviews shows they are
thick as too short planks. Perhaps professional people daren't call a spade
a spade for fear of alienating the public and make them feel inadequate enough
not to listen.Whereas I as a lay person, myself,can tell it like it is,without
being bothered by public sensitivities to their own inadequacies,the fact
is the public are largely innumerate,and do not like maths,and this is a
source of their inability to discriminate when faced with conflicting views
put by scientists with possibly vested interests and agenda.had they not
said "I'll never need maths when I grow up" ,they wouldn't be in a pickle
now.From Ken's comment we can deduce that the lay person is uncomplicated
and easily confused.He then makes apologies for them and tries to make out
that they can handle complexity.If they were intelligent,complexity would
not phase them.Note that Ken says "it's not an easy concept to explain in
a straightforward way" contrary to Neil Chalmers assertion that plain English
suffices ( world30). The real explanation involves
complex maths,but your average lay person cannot handle that,so they get
the Chinese whisper version,that does not clue them in completely,they are
thus confused,because they don't have all the information.That information
is kept from them,because the smart guys know that the public finds numbers
and algebra unpalatable.If we were only honest and accepted that innumeracy
was the source of the problem and that the idea that maths is boring is in
our culture,then we could do something about.As it is this is going to continue
indefinitely - LB]
Colin Blakemore : Are there attempts to devise new ways of expressing
risks,that are more easily understood and are appreciated by the public?
[Note that instead of educating the public the solution is to accommodate
the ignorance and find a "more easily understood" message.If the public were
so intelligent then the complex message would be able to be accommodated
-LB]
Kenneth Macrae : Well there's a method which is technically known
as "bootstrapping"
[Not to be confused with the computer term "bootstrapping" -LB]
,which is to compare a risk of something new with the risk of something old
that people have accepted for a long time.So for example,you can compare
the risk of spending 2 days in New York with the risk of eating 40 table
spoons of peanut butter,they both actually have the same risk of death,which
is 1 in a million! (some giggles from around the table)
[Note that who you are makes a difference.If you're peanut sensitive,then
this might not be the case -LB]
And on the same sort of basis,some that I remember - smoking 1.4 cigarettes
or drinking 0.7 of a litre of wine or driving 150 miles in a motor car,or
spending 6 minutes in a canoe,these are all ways of sort of sort of comparing
a risk of one thing with risks of other things you can understand.
[Note that Ken has mixed metric and imperial measure a definite no-no in
science,and for those clock turners who think imperial measure is viable
(see weight.html),they might not know how much a litre is,and thus the comparison
would be lost on them through not being smart enough to be educated in metric
measure.Further most people don't understand what is meant by 1.7 cigarettes
since cigarettes come in whole numbers.The comparison with something that
does not exist thus becomes incomprehensible to them.This goes to show that
something as simple as decimal notation is sufficient to confuse your average
lay person let alone complex probability problems -LB]
And that is probably the most straightforward way of looking at risk,because
people can relate to travelling 150 miles in a motor car and say "Well that
doesn't feel very dangerous to me",of ...depends who's driving of course.
[Besides who is driving,the conditions of the road matter amongst other
things.But the main point here, is that subjective judgement has entered
into it,and that should have nothing to do with it.Driving 150 miles might
feel very safe to an individual ,but cars have been built to give the illusion
of safety,and what someone thinks is safe is not how risk is assessed.What
matters is how many people have ACTUALLY been injured or killed in cars on
journeys of 150 miles.Of course speed also matters,and risk might increase
disproportionately with speed,or given the laws of physics the chances of
being killed maybe correlated with how much force is generated by a force
which accelerates (F=ma).A lethal risk can be generated at 30mph which is
why there is a government campaign to "kill your speed". Intuitively most
people would think a car could kill,but it might not be obvious that the
relationship between mass and speed,means that a large slow thing can be
as lethal as a fast light thing.This is why small pieces of space junk can
be lethal to space ships even though only the size of a grain of sand.Travelling
fast enough the effect is amplified.Similarly one can drown in much less
water than might be imagined so leaving risk assessment to how someone intuits
it is not a good idea.If they are misinformed of the first thing comparing
it only compound the problem -LB]
Colin Blakemore : Risk also has to be weighted,doesn't it? By the
susceptibility of individuals,the risk of a particular activity is not the
same for very person.
Colin Berry : It's interesting that as the genome becomes unravelled,
[Colin makes it sound like a woolly jumper -LB]
we're going to have more and more information that will enable us to define
risky groups if you like,groups who are at risk I should say.There already
is a considerable amount of data in the toxicological literature showing
how people handle compounds better than others,and it's true of course to
say that a lot of people who smoke won't get cancer of the lung,they'll get
a lot of the other damage due to smoking,I'm not suggesting it's a good thing,but
there a are some of them that simply will not get cancer of the lung,because
they handle the compounds differently.
[Right.So there's enough natural variation to make smoking risky to some
and not to others,which means that medicines tested on some people and found
safe MIGHT be dangerous to other people IN PRINCIPLE,and thus nothing can
be proven safe.Furthermore upon this premise how likely is it that Mr Blakemore's
favourite pastime of testing upon animals is going to yield helpful data,when
animals are more different from us than we are to each other.Animal testing
is thus more dangerous than testing on other humans,since their bodies are
less likely to give accurate indication of how drug will act upon us.Case
in point Thalidomide,chiral molecule (as is DNA see
symmetry.html
) Thus animal testing
is opening up human beings to a greater risk,so even if it weren't unethical,such
experiments increase the danger of a problem for human beings,even after
being shown "safe" via animal testing- as happened with Thalidomide.Animal
testing is thus no means to securing safety -LB]
So all of these risks - I think we're getting better and better at defining,that
will generate other problems about how you advise people about what they
should do for a living,for example.But nevertheless it's a real advance in
risk management I suspect,or will be.
Lynn Frewer : I think the problem here is that we're missing the point
that people don't think about risk in terms of probabilities, I mean some
hazards are associated with ethical concerns,GM Foods would be an example
of that kind of hazard.If risk benefit communication focuses on risk and
probability then people won't be interested because the information is not
addressing the concerns which is important to them.
[That assumes that public concerns are valid concerns.In the case of the
petrol blockades unilateral action that did not see the wider picture was
taken,similarly with the Sarah Payne case.The blockaders were heard to say
things like "my business is suffering",and took selfish undemocratic decisions
to cut fuel off from the rest of us.I didn't vote for them I voted for the
government.I didn't vote for the News of the World I voted for the
government.Minority dictatorships (as Peter Hitchens and Ms Heseltine Observed
on Kilroy) are tantamount to terrorism and insurrection. The uninformed minority
of the electorate should not be able to dictate to a government with a
majority.Greenpeace had "ethical concerns" -broke the law and trashed crops.The
ALF break the law and save animals.Ethical concerns should not be allowed
to ignore actual risk factors. I sympathise with the ALF because the vast
majority of people are speciesist and they would not tolerate racism,I don't
advocate breaking the law,but I can see how one reaches that point.I don't
think it was justified over fuel or over GM crops. This perhaps identifies
why the Two Cultures exists even today.Romantics
work on emotion and rationalists use their brain (the act that is supposed
to elevate us above animals).I'm not hard nosed and cold,because I think
Mr Blakemore was wrong to advocate vivisection,but I had REASONS why he was
wrong,I just didn't go of emotions and hand-wringing.Similarly ethical concerns
seem to engender "worries" that one can't pin down,or gut feelings that something
is awry.What is important to people might not be what is important because
they may have a very narrow set of concerns - "my business is failing" -
it's a free market that you voted for-businesses fail from time to time.It's
dog eat dog and survival of the fittest -except for farmers who we keep afloat
by subsidy. Contrariwise,the government should not be allowed to dictate
to the populace or ignore legitimate concerns,as it did over the poll tax.There
is suppose to be a dialogue,the hauliers claim that their 2 year dialogue
fell on deaf ears,and that's probably the case with anti-hunt protesters
and the ALF,but where a risk issue is concerned such as radiation the figures
should count. There is though,little point in telling someone in Sellafield
that the risk of cancer from radiation is only the same or a little above
the national average (it maybe above average elsewhere where there isn't
a nuclear power station) because one of the mistakes BNFL made when speaking
on this (as did biologist Lois Wolpert in sci-mat.html) is to fail to take
into account the consequences of something that may have low risk of happening.
Thus the chance of a Chernobyl type accident at Sellafield maybe 1 in a million
years,that doesn't mean it can't happen this year,it just rates how unlikely
it is.Nevertheless if it DID happen the consequences might be devastating.
This is probably what worries Greenpeace and other activists over GM.But
I think they are overestimating the capacity of a gene to go walkabout,after
all our genes don't get out of our body and go wandering about until we pass
them on purpose via sex.The passing of a trait issue is similar to the "can
CJD enter the human food chain?" problem,and in both cases we'd like a yes/no
answer.What we know for sure,(ironically) is that certainty is not likely
to be delivered,and that the public are not experts on genes and CJD,and
thus are not well placed to say what should and shouldn't happen unless they
educate themselves as Augusto Odone did,to take issue with the experts. Where
a gene company makes sterile seeds and forces people to buy them rather than
harvest their own ,their maybe moral or ethical considerations,but that's
a different issue from GM in principle,and is not a risk issue. In essence,you
can't have uninformed people dictating to informed people about their personal
concerns which may have no bearing on the issue at hand,even though their
narrow view makes it look like they do -LB]
Long term unknown effects,the fact that a risk is unknown to science,but
also those people who are exposed to it is important.I mean they are all
increasing the fright factor if you like.
Colin Blakemore : Who do people look to for information about risk?
Actually we've got some data about this,the people who are trusted by the
public when ...what we're looking at here is the level of public confidence
here in different sources of information about risk. Government statements,the
media,scientists all grouped together -different types of scientists,and
green organisations,and finally the family. Now,do you find it curious that
people should put much more trust in what members of their own family say
about risk.
["Does science matter?" found that the greens
were the most trusted.The main point there is that TRUST should not be the
means by which one decides what to accept,and neither should that acceptance
be a belief borne of how convincing someone is. Con men are convincing and
if one trusted them,one could lose your livelihood.One should not be intuiting
here or going off gut instincts and feelings.One should KNOW what is being
said - UNDERSTAND it,and draw a conclusion based on FACTS.This requires hard
thinking and bothering to listen,instead of superficially passing judgement
upon how convincing or heartfelt a view is. Green views sound heartfelt but
maybe based in myth or upon the emotional reactions of the spokesperson,and
I say this as once supporter of Greenpeace.As James le Fanu points out in
sci-mat3 a small amount of something in a large
amount of water might not be of concern.Similarly if someone says these
cellphones emit 200% more radiation than this one that DOESN'T mean it's
dangerous,because the two values involved might be .0000001 and .0000002
and be even incapable of causing damage even if they are above legal safety
limits.Those limits maybe set low to give a margin of error.An example -
I once heard someone talking about engineering bridges - He said "Engineers
calculate the load bearing capacity of a bridge,and then double it,just to
be on the safe side." Note that such precaution costs money in labour and
building materials,and so one should do "cost benefit analysis".When this
is done sometimes what seem like unethical results can occur, such as the
kind of seemingly inhuman statements heard from hospital officials - "We
had to turn out all those patients who were close to death as we could not
afford to keep them in the beds".Such statements arouse people's emotions
and in that state they are quick to judge and slow to think -LB]
Lynn Frewer : I think what's happening here is that people perceive
that their family members are honest,but might not necessarily be conveying
accurate information,and I think we have to distinguish between perceptions
of honesty - a source is telling the truth as far as it knows - and credibility
- whether a source is perhaps distorting information to protect a vested
interest,or some.....
|
From Mr John Hicks:-
Sir, You report today that, in a test devised by a management expert, anyone
who answers "true" to the question "I have never unknowingly told a lie -
true or false?" is placed under suspicion. That is unfortunate, because to
anyone who understands the ordinary use of the English language there can
be no other answer. We all often unknowingly say things which are not true,
but that is not lying. A false statement is a lie only if intentional. To
lie unknowingly is a contradiction.
Yours faithfully,
JOHN HICKS,
17 Montagu Square, W1H 1RD.
[email protected]
January 7.
From the Reverend Dr Peter Cameron:-
Sir, The only appropriate answer to the question "Have you ever unknowingly
told a lie" is the one Mr Gromyko is reputed to have given to a journalist
who asked him at a summit conference if he had had a good breakfast: "Possibly."
Yours faithfully,
PETER CAMERON,
St Mary's Rectory, Birnam, Dunkeld, Perthshire PH8 0BJ. January 8.
[This puts paid to the idea that speaking plain English in any way aids passing
scientific information to the public (see world30
-Lisa Jardine).If they can't even understand their own language what chance
have they of following mathematics? And in this case the mathematics is embedded
IN the language,showing that they aren't two separate entities.I note here
also that a Rev can perfectly well deal with logic and reasoning in case
anyone thought I didn't think that they could -LB]
|
Mike Maier : There must obviously be a sort of component of conflict
of interest,because the classic paranoia is the sort of the industrial complex
against the individual,and you know when you see industrial scientists give
advice people,nobody believes them,because they feel that they are in the
pocket of industry.
[Notice once again that they are going of feelings and not facts -LB]
Colin Blakemore : That's certainly borne out by an analysis of trust
within the group of scientists.I think the polls show very clearly that
scientists associated with industry are trusted much less than academic
scientists,with government scientists in the middle somewhere,so the level
of trust is weighted by this vested interest.
Lynn Frewer : There's also an element of increasing distrust in
science,scientific institutions which as been going on since the 1950s.
Colin Berry : I mean there are plenty of scientists that see research
opportunities in investigating particular risks and then are happy to contribute
I think to the media,their view of this being a major public problem,I mean
I think that can be a difficult area.
Colin Blakemore : Isn't it surprising that trust in the government
is so low,and that trust in scientists,at least some scientists is relatively
high.
[I don't think so.Who do politicians ask to inform them about what to do.When
Gummer was pumping a burger into a child's mouth,he was doing so because
he had been told that the risk from BSE was very low. The source of information
is science,and people know that politicians will say anything to stay in
power for another term,whereas scientists like other public servants aren't
voted for and have no real reason to warp the results in order to get voted.They
may have other agenda but that's not one of them.Perhaps this is one case
where the public have risk assessed and deduced that (Tory) politicians are
a bad bet if you want the truth,considering how many times they've been shown
up as liars.Scientists on the other hand say X and time usually shows that
X was the truth. The X-files are in scientists hands so to speak! -LB]
When of course the government should be advised,and is advised by scientists....
Colin Berry : And of course Colin that....there's a piece of nonsense
there.The committee of safety of medicines for example,has looked at every
drug we've been given,and has decided whether it's safe to give people.Now
you may say there have been some wrong decisions in that,but of course that's
partly because most drugs have only been tested on say, what? 3- 5000 people
before they're given to a population,so it's not surprising that other things
turn up.
[This is one way that it could look as if scientists have got it wrong,when
in fact there was a given likelihood that a "safe" drug as proven in
trials,nevertheless shows up an effect in one of those people for whom smoking
might not posit a cancer risk.This does not mean that the testing method
is faulty,it's just that nothing can be proved 100% safe -LB]
But the whole of our pharmacopoeia has been tested by government agencies
in effect, though advised as you say by independent scientists.So there's
not a very close,sort of logical consideration of what are the issues taking
place here.It goes back to what Lynn said earlier,about how people's emotional
reactions...
Colin Blakemore : Yes,yes.
Lynn Frewer : Could I just perhaps quickly add,that perhaps a little
distrust in regulatory bodies and regulatory practices is rather
healthy.(Agreement from the table)
[You can,because it is. No person should accept what anyone says verbatim
as true,which is why mindless unquestioning allegiance to biblical or textual
authority is unhealthy.If people are going to take issue with science,they
should (and sometimes because of biblical or textual precedent) do so from
a position of scepticism commensurate with its track record at delivering
benefits. Similarly,the same degree of scepticism (if not more) should be
applied to those authorities which have failed to produce similar results.
The bible fails to deliver miracles,praying and chants and other texts fail
to deliver.Science has systematically delivered all but the most perverse
and extremely wild claims that it said it would.If it were a religion there
would be no reason not to have faith in it.But it ISN'T a religion and it
is NOT a belief system.One is not supposed to functioning after a process
of trust and belief,one should understand why something is so and accept
it because one cannot see any other way that it could be -LB]
Colin Blakemore : Well of course quite a lot of the concern recently
has been about food. Everyone wants to believe that what they eat is safe.But
in recent years that confidence has been severely tested.When BSE first emerged
in the 1980s,scientists began to suspect the source of the infection was
cattle feed,containing offal from sheep.
[Our country "guardians" the farmers who know much more than us "townies"
(see waugh1.html and rspca.html) about animal husbandry and how to run a
business,showed us once and for all how truly ignorant we are about country
ways,and crippled our agriculture business so that now we're having to pay
them to make a loss.Nice going guys- I wish I was as informed as you are
about nature and the biology of animals so that I could be such a great
success.Next thing,you'll be telling us that foxes need shooting because
they are so much a pest that their numbers need controlling by ripping them
to shred with hounds showing us again how much you understand the environment
and the "ethical" concerns that should be taken into account when trying
to maximise profit margins. It's quite obvious that you guys know all about
bull shit,you have your own local source in close proximity to your body.
The fuel blockades showed clearly how much your knowledge of straw surpasses
mine. Perhaps this is also because you have your own local source between
your ears,the latter, like as not being made of corn.But this could be just
a straw man argument -LB]
The feeding method was banned,and although any risk to human health was deemed
"highly unlikely",the government still took drastic steps to try to allay
public suspicion,with limited success.While BSE was headline news here, Americans
were in uproar over apples and apple juice. A pesticide called Alar had been
linked to childhood cancer.
[And who linked it? A lay person? A mystic? A politician? A farmer? -LB]
The doses being talked about for any serious effects were enormous,but the
manufacturer soon announced it would abandon Alar production.
[So the link which may have been speculative,and which would possibly have
only have produced less cancers than people get through self inflicted
smoking,and which may have yielded a benefit was curtailed because it posed
a "risk".How much benefit is smoking? And how risky is driving a car?-LB]
In the mid 1990s,the government announced a probable link between new cases
of human new -variant CJD and BSE infected beef.The renewed scare set the
scene of another test of public faith in science - genetically modified food.Some
now feel that concerns over the risks of GM technology are drowning out pleas
for rational argument over its possible benefits. Lynn why this obsession
with the safety of food?
Lynn Frewer : People have to eat,I mean they have no choice.
[They do-it's just Hobson's - LB]
So they can't avoid risks associated with food,but beyond that food embodies
people's culture,the kind of social norms which they use to regulate their
lives.Food taboos are very important,there are religious links.
[Recently I witnessed a debate about whether soap operas affect out lives.On
it,a person who considered themselves "black" argued that she was not represented
properly on TV dramas claiming her food preparation techniques were peculiarly
cultural.She then went on to say that she washed her food before seasoning
it,something that I guess most people do. People might not understand that
a lot of rituals are based on facts that they do not know,for instance the
removal of blood from organisms that many faiths subscribe to is not done
purely because God said so,this is probably an observed wisdom picked up
fro the fact that those who ingested blood were given to pick up diseases
from the corpse. Today we know that there IS risk associated with that,and
as Anthony Giddens explains in reith992.html the concept of risk is relatively
new,so religious texts may contain old wives tales or pieces of basic wisdom
-this does not lend them credence as virtuous authorities,it just means that
some things our ancestors got right by accident or by empirical
observation.Similarly, some of those things can be catastrophically wrong,because
something is old does not add value (No matter what Antiques Roadshow says),and
given that there have been myths as ludicrous as thunder being created by
a god or the sun being a god in all probability such authorities are highly
ignorant-containing all the inadequacies of mankind without the benefit of
modern technology.I dare say there've been Cadfael characters in the past,but
that doesn't means that such people could outwit a modern toxicologist who
understands molecular structure.I note here also that in world30.html Lisa
Jardine made up a 1/4 of the speakers in terms of gender,and here Lynn is
1/5 of the speakers as the only female.In both cases it is the female that
asserts the "emotion/romantic/intuition" or people orientated type approach,and
even though in these cases it isn't extreme (Lynn said "If you use probability
people won't be interested") it nevertheless does nothing to dispel the idea
that females are hand-wringing emotionally distraught anxiety mongerers who
believe in 6th senses,and supports my (and the guy on"Soul of the UK") idea
that post radical feministic Germaine Greer ranting has produced a female
populace that is essentially anti-science by virtue of being anti-male
establishment.If true, it would be a shame to lose a powerful tool that worked
merely because females were attacking anything thought to be male and
establishment. Science should not suffer merely because females are retaliating
for being subjugated,after all there are now and have been female scientists,and
for the benefit of those PC idiots that think everything has to be reflected
by ratio.There were no "black" people on either panel ,partly because there's
no such thing as "black" person,contrary to what Jim Davidson continues to
say,and it is not necessary to have 50% women on a panel at every opportunity,or
N% Irish Scots or Welsh people,because everyone is an individual not a group
heading -LB]
|
Sue Birchmore on the problems in interpreting statistics
Why can't a woman be more like a man? Two
scientific studies . . . say it's impossible. ''Women, apparently, are inherently
worse at driving and map-reading (inferior spatial skills) but better at
talking and listening (superior verbal skills). It appears that such differences
turn up even among rats; male rats are better at getting out of mazes than
female rats. My apologies if I am misrepresenting Glenn Wilson , author of
The Great Sex Divide (Peter Owen, October 1989); I'm quoting from
reports on his book in the popular press, because my moan is against the
reporting, not the research.
I don't doubt the excellence of the science, but
I'm desperately tired of the debates reports of this kind induce. Thanks
to the rats, we are in for another flurry of fruitless, angry arguments,
during which a lot of men will be accused of chauvinism, a lot of women will
be accused of hysteria, and when the dust finally settles we will all be
more or less where we were before except that my temper will be a little
more frayed.
The trouble has all to do with statistics. An ex-colleague of mine used to
describe statistics as "the science which tells you that if you lie with
your head in the oven and your feet in the fridge, on average you'll be
comfortably warm".
Don't get me wrong: I'm not anti-statistics.I don't at all subscribe to the
"Lies, damned lies and..." school of thought.
The discipline of statistics is a marvellously useful tool. In fact, I
think an intensive course on statistics should be a compulsory component
of every schoolchild's education.Then, perhaps, we might breed a nation which
actually understands how to use them.
I became very fond of statistics while working in quality control. Once we
had acquired some computing power to do the number crunching, it became magically
easy to cut through all the heated arguments as to whether a suspect batch
of components was or was not out of specification. And statistical process
control put paid to the old machine-setting routine, where the setter would
make adjustments on the basis of a single measurement and wonder why the
rest of the batch didn't come out the same. Wonderful. But dangerous.
Statistics are tricky things,to be handled with care. The new-found enthusiasm
for statistics of one of our customers was very nearly extremely expensive
for us. Testing to destruction of components for car seat belts had produced
a large scatter: none of the parts actually tested had failed below the specified
limit, but the capability study predicted that a large percentage of the
batch would. I steamed down the motorway without delay to investigate this
disturbing finding.
Closer examination of the results showed that most of the failure loads were
around about the expected level, with a few "flyers" much higher. The
distribution was, in fact, not normal at all but highly skewed. Analysing
it on the assumption of a nice, symmetrical normal distribution predicted
that there would be a corresponding set of failures at loads which were lower
than the average. But there wasn't.
Statistics can be dangerously mesmerising if not interpreted with a large
dose of common sense. In a similar way , I was able to still the worries
of one of our suppliers, who had analysed tensile-test results from samples
of wire and come up with a spread that was well out of limits. A little probing
revealed that the test results weren't all from the same batch. It was actually
a double distribution.--- two intermixed normal distributions with different
means. [See JA Paulos - paulos2.html] Combining
the two resulted in a "mean" somewhere between the two peaks, and a calculated
spread that went right off the graph , but had no relation to real life.
The most common statistical fallacy, though, is to argue from individual
cases to whole populations and vice versa. The machine setter used to make
this error when he measured a single component-which
Murphy's Law
ensured would be somewhere from the tail end of the distribution.---and assumed
that it was typical of the batch. When it comes to interpreting measurements
concerning people, we tend to make a more subtle error. The initial statistics
may be impeccable; a large, well-randomised sample, yielding a statistically
significant result, showing that, on average, men have better spatial reasoning
abilities than women-scientifically proven! Yes, and on average men are taller
than women. But, even though I'm only a woman of average height, I've met
several men who are shorter than me. The distributions
overlap--but the digested versions of the scientific results in the popular
press don't tell you that. Instead, they present the case as "men are better
than women" without even the qualifying "on average".
The readers go away with the fuzzy notion that it means all men are better
than any woman, which is manifest nonsense if you stop to think about it
clearly. I'm quite prepared to believe the results of research which say
that-on average- males are better at mathematics than females. However,
that hasn't prevented a series of brilliant female mathematical minds
from Hypatia of Alexandria right down to the youthful
prodigy Ruth Lawrence outdoing their male contemporaries. The distributions
overlap.
My contention, which I tend to propound at excessive length when drawn into
men- versus-women arguments, is that the variation between individuals is
far more significant than small statistical differences between large groups,
whether they're divided by sex, race, hair colour or shoe size. Be all that
as it may, I can make two sure-fire predictions about the outcome of the
present debate without the aid of statistics.
One: those who always did believe than men are, in some mystical way
, superior to women will go on believing that.
Two: those who didn't still won't.
Sue Birchmore is a technical writer living in Birmingham.
In matters of sex,we're stuck in a time warp
By PATRICIA ROBERTS
We're used to women at the top. After all, female prime Ministers
are no longer a novelty, and "the girls" have breached most male bastions.
So its sad, as well as sick, isn't it, that when a woman tried to join the
boys-only fire brigade at Manchester Airport, the reaction of the lads was
allegedly to have a sweepstake on whether any of them could have sex with
her. Now a furious row has erupted about the whole affair and there is to
be an independent investigation which will cover several issues, including
why the woman did not pass the various tests to become the airport's first
female firefighter.
We're just a garter-stretch from the 21st century, yet when it comes to sex
equality. we still seem stuck in a time warp. Britain's firefighting service
must be the best in the world It has a fine reputation for top class
professionalism arid dedication to the job. But Ringway's firefighters, reputed
to be among the highest paid in the country, now seem to want to short- change
the opposite sex with the lowest common denominator response to a female
colleague joining them, and the allegation, if it proves to be true, does
them no credit, and lets down all of their colleagues. Wearying, really,
isn't it? Childish even.
But should we really be surprised? Men still tend to guard their male preserves
jealously... after all, the 15th hole of any golf club is still a man's domain
and gentlemen's clubs in London are still just that - clubs for men only.
Common sense has dawned up north, however, where only this week a gents'
club in Liverpool bent to pressure to open its portals to women. The thing
is lads, we women can run countries. or companies, at the same time as doing
the washing, stuffing a mushroom, doing the cleaning and the ironing and
mending a fuse (only don't ask us to set the video PLEASE).
Millions of working women also get few thanks and only poor pay for juggling
work and home, for they still make up the bulk of the lowest paid of the
workforce. Most of the time the fellas are happy to let us slave away, because
it's the chaps that usually benefit. But do us one favour please,leave the
smutty and demeaning remarks in the dark ages - where they belong.
Manchester Evening News Apr22 1996
|
Lynn Frewer : Food permeates the whole of people's lives and defines
culture.So to start linking food consumption with risk is affecting not only
human health but all these other elements that are important to people.
Colin Blakemore : But it also generates all sorts of myths,doesn't
it?
Colin Berry : The Alar's a very good example,you'd had have to...first
of all the study that caused concern was a mouse study,where rather odd tumours
developed.It was probably due to a hydrazine in the compound that was being
tested. Now the...you pointed out the enormous dose that was necessary,and
I calculated at that time,there were 28,000 apples a day is what you'd have
had to eat in order to get the same dose that the mice got,even assuming
that the mice are relevant in this context. Now there's much more hydrazine
in mushrooms.One sort of serving of mushrooms would...might well do the same
sort of thing.So there's no rationality in this.
Mike Maier : Why wasn't it sufficient to give an explanation as you've
just given to the American public,why was it necessary to ban it then?
Colin Berry : Well first of all of course,if you generate a story
of this kind,it takes time....I mean not....the experiments were done,it
would have taken a week to do the calculations and work out,and that's too
long.Once a risk is established in this sort of way,and obtains a
momentum,denying it a week or a month later,or merely pointing out that it's
daft doesn't work.
[So now we see how the confusing "One week red wine is good the next week
it isn't" stories come about. First of all a risk is speculated upon doses
to mice that far exceed anything a human is likely to ingest,and the mice
get cancers (when humans might not through being not like mice),the media
then report this as a risk and set the agenda.The scientists then has to
back-peddle and allay fears because innumerate journalists have failed to
explain that no one is likely to eat 28,000 apples a day.This has all happened
because animals are given excessive doses "just to be on the safe side" in
the way that an engineer doubles a bridges load carrying capacity to ensure
the public isn't put in danger.Perhaps the answer is not to be over excessive
in doses used on animals,or better still use people who volunteer,and have
the media informed properly about test does levels and what the risks actually
are -LB]
Lynn Frewer : The problem we've got though is that the public are
very well able to conceptualise uncertainty,but science and scientific
institutions have been operating to date in such a way as to deny that those
uncertainties exist in risk management,
[Whether or not the public conceptualise it they can't calculate it (See
Sue Blackmore sci- mat3.html) -LB]
,and I think the problem-the reason for this decline and trust in science
- is that the public see uncertainty in risk management - particularly when
scientific knowledge changes - we know more about a particular risk - but
to date - and BSE would be an example - science,scientific institutions have
been denying that that uncertainty exists.
[That's not true.The public demand certainty,and science does it's best to
deliver to the demand.People prefer to gloss over the uncertainty,that's
why when faced with conflicting reports rather than go with what makes rational
sense,they stay with their inherent beliefs,or warp the data to suit what
they would like to believe.Those who smoke shrug the risk off an continue
smoking,no one drives slower if told the risk of death is double at 40kph
than what it is at 30kph because they figure there's enough uncertainty about
it that it might not be true.In fact the uncertainty is overplayed.For things
like global warming there maybe data that shows both an increase or a decrease,or
even no change.But there certainly is an ozone hole,there is no uncertainty
about that.What maybe the dubious in that case is just how profound an effect
it may have,if the diminishing in ozone is only 1 part in a billion it might
not be a threat (it could be a product of sensitive measuring equipment),but
if that deterioration is part of a trend to lose 1 part in a billion compounding
over a series of months or years,and shows a trend,then it could be serious
-LB]
So,it looks as though science is protecting some kind of vested interest.
Mike Maier : Yes I think the lack of clarity is crucial here. The
picture of John Gummer giving his daughter a four year old...er daughter,a
hamburger is going to become a sort of icon of misjudged behaviour,because
that picture probably made people even more concerned about the whole issue
of BSE.
Colin Berry : And yet there's a pressure to indulge in that sort of
thing, I think,
[That's what I mean.If Gummer hadn't done it,people would have had no image
with which to go off to indicate safety. That's the fickle public for you
- you just can't please all of the people all of the time! -LB]
because when estimates of the numbers of people likely to be involve range
from say 200 to 30% of the population...
Colin Blakemore : This is the numbers that are likely to die from
new variant CJD?
Colin Berry : ...exactly,there was enormous criticisms of the scientists
for being so sloppy.What they were actually doing was making a fairly accurate
statement of the uncertainty.
Mike Maier : Yes.
Colin Blakemore : Yes.Yes,so this is one of the problems isn't it
Ken? That the public seems to expect that scientists are infallible.
[That's an indication of just how powerful science has been in delivering
the goods,people now expect it to deliver everything with 100%
capability,including their morals and spiritual views which it isn't supposed
to! And nothing delivers with 100% capability-that's inhuman and science
is done by human people who are fallible. It's because of a "taking fro granted"
or a "resting on laurels" that people act like spoilt brats,they "never had
it so good" for so long that they are used to getting high standards.Now
a small error looks huge in terms of what has been delivered. We make mountains
out of molehills -LB]
They're there to give absolute statements about right or wrong.
[Thus you can see ho people think they are supposed to deliver morality
-LB]
Kenneth Macrae : Well I think we use words to mean different things.I
remember Stephen Dorrell when he was minister for health talking about BSE,in
the house of commons,and he used the word,the was a very small risk that
BSE would cause CJD.What he actually meant was,he thought it was very unlikely
that it did.
[Another example of how using plain English doesn't aid communication (see
world30.html) - LB]
And if it did,the risk quantitatively would be very small.In other words
he was confusing risk and uncertainty.Uncertainty is admitting you don't
know,and I think.....I mean we're guilty as scientists,people I think like
to claim that they have more certainty about the statements they make,than
really often they are justified in claiming.
[The risk is the chance that it will occur. By saying "small risk that BSE
causes CJD,this assumes there is a link,it's just that the BSE is unlikely
to cause CJD.Whereas saying "It's uncertain that BSE causes CJD" means no
link has been established,but one could be forthcoming. It just goes to show
that the English language in the hands of politicians when dealing with technical
points is a source of confusion not information.One would suggest that if
Ken had been delivering this message,the confusion would not arise,since
he knows what he is talking about.So who is best to inform the public -
politicians or scientists? It's best to get the info from the horses mouth
-LB]
Colin Blakemore : I would guess that it...that the BSE episode has
certainly influenced the judgement of people in this country about GM food,about
whether they want it,and certainly the reaction - the hostile reaction -
against GM food was so much earlier and so much stronger here,and that may
well be due to the BSE experience I guess.
Kenneth Macrae : Oh I think that's an example of confusing facts and
values.That people have a value system that natural is either good or that
scientific progress has been wonderful (see
natural and trans
-cblast1.wri) and I think people use reason and evidence to try to justify
their opinions,rather than to form their opinions.
[That's essentially what I've found by talking to people online.People have
a fixed view (some people) and try to shoehorn the facts into an a priori
view.God is a prime example. One cannot assume God's existence and try to
warp the facts to justify God.One must not believe either way and form an
opinion upon the facts,not bend the facts into one's world model. Perhaps
this is the difference that forms the two cultures (see
world30) Science theoretically progresses by having
no preconceptions (beliefs) and does tests to find out. Whereas romantics
(or the public) have preconceptions (beliefs) that they ask to be disproved
by the facts,and of course no amount of data is ever enough to remove the
dogma -LB]
Because,I mean the...we just don't have the information at the moment about
GM food.But some people are frightened by the uncertainty
[Fear of the unknown -LB],and it's meddling with nature
[Which puts man outside nature,when we're not- we're inside it -LB],and other
people,perhaps the ones who know more,but are not trusted by the rest of
us,can see the possible advantages.
Lynn Frewer : "The advantages to who?",is also very very important.
(Murmurs of assent)
Colin Blakemore : Well,that's right balancing benefits against costs
and not just benefit to yourself.
Colin Berry : And which population? I mean the vitamin A enhanced
rice is going to be very much more significant for some populations than
others,and those sorts of things affect things too.
Lynn Frewer : I think perhaps part of the problem was that the European
public saw that the benefits were accruing to industries,producers in different
countries,and being imported into Europe,where they had no choice,because
genetically modified ingredients were not segregated,there was not a concrete
and understandable labelling policy put into place.So...
[Whilst this is true,and I support labelling,to a certain extent this is
like asking for 5mm grains of rice to be separated from 4mm grains of rice
and packets labelled accordingly. A GM food is not necessarily any more of
a threat than a normal "natural" food (which could of course be more dangerous
eg Hyradazine in mushrooms),and perhaps because of this no option was taken
to label.Again what the public perceive or misunderstand should not be
implemented merely because they ask for it,if it costs money does not actually
produce a viable choice.Having said that,the free market works by democratic
choice and defaulting to no choice removes a citizens right to choose.I just
think it would be a somewhat farcical choice,like separating out hard and
soft water so one would choose not to have to de-fur a kettle.It's a choice
which could be implemented,but water is essentially water,and food is essentially
food.The Frankensteinian myths propagated by the press fail to appreciate
some of the subtleties in GM -LB]
Colin Berry : Yeah but tomato paste was accepted without (Colin says
something meaning "dissent")...and still I believe outsells any other sort
of tomato paste.
Lynn Frewer : Which is labelled...
Colin Berry : Sure.
Lynn Frewer : ..which is cheaper,though a consumer benefit...
Colin Berry : So the benefit was the price and so on yeah,yeah.
Lynn Frewer : ...it was identifiable.
Colin Blakemore : I'd like to spend the rest of the programme talking
a little bit about health risks in things like mobile phones.I have to declare
an interest there because I was a member of the independent expert group,that
was chaired by Sir William Stewart,that's recently produced a report,and
our dilemma in considering an area of what is clearly an area of great public
interest,and potentially of great concern - half the population now have,and
use mobile phones - was that there were hints in the research literature
of possible effects of the microwave radiation from mobile phones on the
cells and tissues,but no clear evidence that any of those biological effects
carried any health hazard.How does one deal with that?
[Note that the tests probably involved strapping a cellphone to a mouse and
leaving it on for 6 months non stop (ie dosed it with 28,000 apples worth
of microwaves),and they still didn't find anything.Humans have radiation
all around them at much stronger doses than cellphones, even if proximity
is an issue -LB]
Well what we decided to do was to apply what's called "the precautionary
principle" or a version of it. Lynn what do you think in general of this
approach of using a precautionary approach when you're just not sure of all
the facts.
[It seems eminently viable.It's like treading with care on ground that was
recently seismically disturbed to be wary of falling down a crevasse,it's
the only sane policy -LB]
Lynn Frewer : I think one way to deal with this is to involve the
public more in risk management processes,and I mean directly.
[You mean that "intelligent" public that can't accept risks in terms of
probabilities and adds in its own subjective value judgements? And that will
help? If they are to be involved it has to be at the level Shiela MacKeckney
is,being informed of how science functions.One can't just wander in off the
street and expect to be involved,any more than a someone can walk in of the
street and perform brain surgery.Risk management isn't brain surgery,but
it is a complex task requiring training and skill -LB]
Look at ways to try and understand what's driving public views about risk
and take those views and use them to inform regulatory policy,so that we're
responding to what the public want,rather than getting them to react to a
particular regulatory style.
[Again we're accommodating public ignorance and value judgements,as if the
customer is always right.In many TV discussion shows the kind of nagging
doubts that seems to be indicative of the public mind,have been shown up
to be products of their lack of information about the subject with which
they take issue,by a scientist brought on to allay fears or discuss the issue.We
SHOULD NOT be accommodating misinformed or uninformed views.What should happen
is that the facts should be put forward as they are,and in no uncertain terms
the public should be told that the facts come in a language that requires
them to think and try hard,and if they want answers they have to put in effort.
Accommodating their laziness and preconceptions does nothing to add to
progress.You can't water down complex issues to make them palatable to a
simple audience and expect to have communicated the reasons why what policy
is is not what is the public wish. There seems to be an insatiable urge to
ask "why?" in the public domain that is critical of policies that seem to
be against the public will (ie Tony Blair's stance of fuel tax).Such decisions
aren't done merely to upset and wind up the public,they are done because
they make sense,and however much lack of sense they appear to make to public
scrutiny,this is because public scrutiny is superficial and not penetrating.If
they looked with deeper eyes,and stopped adding in their own preconceptions,then
they'd see the sense.Of course this appears as a very arrogant and supercilious
view,but I don't make excuses for public ignorance -LB]
|
Even a superficial glance at the world around us reveals
the extent and diversity of statistics. Almost daily our newspapers report
public opinion, as defined by the latest poll, or the state of the economy,
as measured by various econometric indicators.Articles, sports reports, and
a great deal of public debate feature an assortment of descriptive statistics.
Concern by eminent statisticians over political interference in the use and
collection of official statistics has recently surfaced. Within industry
and commerce, aspects of insurance, accounting, economic forecasting, and
market research have long relied on statistical techniques, while the current
revolution in quality management stems from a more sophisticated statistical
approach to the variability inherent in production.
Many of the same statistical techniques are now an integral part of research
methods in medicine, applied science and much of social sciences ; they have
also begun to impinge on both the theory and the practice of our judicial
system. At a deeper level, scientific theories of the nature of humankind
and of our Universe explicitly include randomness.
Little wonder that these authors claim that "statistics and probability are
all-pervasive" and that "the empire of chance sprawls over whole conceptual
continents". It was not always thus. Probability theory originated about
300 years ago in the study of gambling problems.
Against the backdrop of a profoundly deterministic [and fatalistic-LB] view
of the world,its relevance was strictly limited to (at best) a statement
about ignorance of underlying causes, or the (current, but presumed not
indefinite) inability to measure with complete precision.
Statistics, in the sense of large sets of data, became available during the
19th century. The collection of techniques we now use to study data have
only recently been developed. Written by historians and philosophers, The
Empire of Chance seeks to trace the development of these ideas and concepts,
and their symbiotic role in shaping scientific thinking. However,its subtitle
"How probability changed science and everyday life'' indicates more about
its content than its title. Rather than describe or delimit chance's current
empire, the perspective is historical.
The authors' treatment of their chosen subject matter is thorough, if somewhat
dry, and no formal background is necessary. The first two chapters detail
the separate rise of probabilistic and then statistical thinking. Real progress
in making inferences from uncertain data had to wait for the pioneering work
in statistics of Ronald Fisher in the 1920s, but this was soon followed by
the controversy between Fisher , Jerzy Neyman and Egon Pearson, and the
Bayesian
school [Ref: Video:N30 The Numbers Game {Bayesian Statistics};Red
File1: Nature1.wri] over the "correct" approach.
Three separate chapters then each focus on a particular discipline. Correlation
and regression arose through Galton's work on inheritance. More generally
chance entered evolutionary theory through the field of Mendelian
segregation (and possibly), random drift and mutation, and variation
within populations provided the raw material for natural selection.
In physics, statistical mechanics (which arose by analogy with the "social
physics", or "statistics" of the early 19th century) ,first challenged
classical determinism. In the 20th century quantum mechanics
explained behaviour at the atomic level in a fundamentally probabilistic
way. In psychology , statistics first became institutionalised into research
methodology in the and statistical ideas were actually providing theories
to explain the manner in which the mind functions, and ultimately "transformed...
the idea of what an explanation is".
The penultimate chapter briefly outlines some contemporary applications and
the final chapter surveys important issues. Although it is an unfair criticism
in view of the book's stated historical aims, I was disappointed that the
authors did not extend their account sufficiently into the late 20th century
to include the exciting modern role of statistics,for example, in image expert
systems, the analysis of DNA sequences or in probabilistic
algorithms. There is similarly only passing reference to CHAOS.It
would also have been of interest to ponder on the future of the subject.
While commendable on one level , the detail and style of The Empire of
Chance may deter casual readers such an account is both valuable and
relevant to the non-historian, not least in adding our insight and highlighting
the themes that continue to recur. These include the contrast between various
deterministic and stochastic world views ,and the meaning of probability.
Of particular irony is our "fascination with the numerical and the longing
for certainty that the numerical symbolises" when statistical experts
may still disagree both on a theoretical and practical level about which
number is the ''correct'' one in any given situation. The vastness of the
empire of chance is well documented. As M. G. Kendall claimed. statisticians
"have already overrun every branch of science with a rapidity of conquest
rivalled only by Attila, Mohommed, and the Colorado beetle" . Society has
certainly institutionalised statistics. It is far from clear, however,
that society understands uncertainty.
New Scientist 27Jan 1990 No1701
p67
What planet are these people from?
by David Manson
They're at it again, the statisticians with nothing better
to than upset us. Just a couple of weeks after claiming that nearly half
the people in Manchester who could work don't,they now assert that Manchester
has the highest shopping prices in Britain. One question. How do all these
people at the bottom of the first pile pay for the goods that put them at
the top of the second?
[It's quite possible for the two figures to be true,just because we have
the highest prices does not mean there aren't lower prices in Manchester
that those who don't work can afford,I should know I am subject to both
categories.An example of a journalist who can't follow figures -LB]
It's possible that we have all found incomes that don't involve work, such
as winning the Lottery or joining the board of a recently- privatised monopoly.
But unlikely.
["Unlikely" presumes he's calculating risk - which itself is "unlikely"
-LB]
The truth is that the prices survey, like the government's work survey three
weeks earlier, is nonsense. It tried to record what families spend on everything
from car insurance, to furniture. And on food, the vast bulk of any family's
bills. Surveying food is notoriously difficult. Europe gave up years ago
after The Great Sausage Debacle, when statisticians agreed that the standard
sausage was a frankfurter. During a subsequent survey in the Midlands, not
one frankfurter was sold, which proved to the statisticians that no sausages
were sold.
[Logically if sausages are defined as frankfurters and the data shows
no frankfurters are sold,then no sausages are sold,this MUST be true. This
does not mean however that sausages not defined as frankfurters continued
to sell -LB]
Equally fatuous were the conclusions of a survey of what we pay for consumer
goods compared to Americans, which found that some items were 50 per cent
dearer here. The Office of Fair Trading got all hot and bothered and began
muttering about cartels and price-fixing. An kinds of legal explanations
were also offered, ranging from higher production costs to VAT. But totally
ignored was the fact that we are prepared to pay more for an item we considered
important.
[I'm sure US citizens are quite prepared to pay more for goods they consider
important too,I don't see what that's got to do with it. In the recent fuel
crisis,one Labour MP explained that our taxes are calculated on a different
system to Europe (thus probably the US too),because that's what we voted
for.Oddly those Euro sceptics who probably include farmers and hauliers,would
not have a price discrepancy that they see with Europe had they been sympathetic
to rules which would have applied to all Euro countries.The UK's opting out
has left us cold shouldered.By being so patriotic and pompous,we shoot ourselves
in the foot and don't gain the benefits of a co-operative buying in bulk
-LB]
The recent shopping survey didn't mean, as the Reward statisticians concluded,
that "Manchester must be a relatively affluent area," just that we know how
to spend what money we've got. We choose to spend on food instead of designer
socks, on lettuce instead of Lacroix. Good news for supermarket chains now
recording record profits (there's nothing wrong with profit, except tycoons
this successful should be running the country instead of messing around with
cornfakes).
Bad news for statistic-lovers who read too much into too little. In fact
the government's Central Statistics
Office has just changed its ground rules when checking prices by including
Knightsbridge and Belgravia as "test" areas and shelving several ailing towns
in the midlands. Having done this, they found that Belgravians don't drink
many pints of mild, an old favourite with statisticians when it comes
to comparing the price of "basics." So instead they've started comparing
such things as private school fees and the prices of aerobics classes.
What planet are these people from?
But the very last word on statistics must go the the Greater Manchester CID
chief, who had to explain on Thursday why we had the country's worst record
for solving burglaries What he said is: "In terms of figures,GMP? has at
least maintained its performance oven the last year and, when you take it
over-all, you see there has been a reduction per thousand head population-wise."
What he meant is: "We may not be too hot at locking up burglars but by golly
we know how to truncheon a bad statistic until it looks good."
Manchester Evening News Apr22 1996 |
Kenneth Macrae : I was curious,because I haven't read - apologies
- I haven't read your writings on mobile phones,how did you apply the
precautionary principle to that issue?
Colin Blakemore : Well I think we should get clear what the precautionary
principle is really.I mean the way we interpreted it was just to be reasonably
cautious where the costs of being cautious are not so high.I mean not so
draconian...
Colin Berry : It's not always defined in that way though Colin! (slight
humourous tumult)
Colin Blakemore : ...well I know,exactly,exactly,no there are lot's
of...
Lynn Frewer : There are other versions.
Colin Blakemore : ...there are other versions.
Kenneth Macrae : But cautious in what way? Cautious about saying "there
are dangers" or cautious about saying "they are safe"?
Colin Blakemore : Cautious,oh,cautious about the possible hazards.What
we said was that while we could not at the moment state clearly that was
any risk to health,there was a background of evidence that phones were causing
biological effects in tissue which might conceivably carry health risks in
the long run.So what we said was to be reasonably cautious until that evidence
was in.
[That's like saying no one really knows if a budgie can peck
you to death,so in the meanwhile keep your budgie caged,because there is
a background of evidence that suggests some birds are capable of inflicting
great damage on human beings -LB]
And one element of being cautious is to make moves to gather more evidence,to
do more research of course.But also informing the public so that they can
make their own choices we thought was terribly important.Providing more
information about the level of emission from phones because they vary
enormously....
[Yes but they come no where near the emissions from a microwave and no one
is carping about microwave ovens,and in terms of power and proximity possibly
ovens are emitting more than phones,why no oven scare? Could it be that the
media/public have no conception of EM waves and are not versed in physics
and maths? Oh come on now - give them credit! -LB]
Kenneth Macrae : So we might get stickers like on cigarette packets?
Colin Blakemore : ...well we won't exactly...
Kenneth Macrae : Low dose/high dose telephones?
Colin Blakemore : ...that's exactly what we suggested?
Kenneth Macrae : Really?
Colin Blakemore : Yes,and that tables should be available to show
people a comparison between different phones.
[Theoretically all electronic equipment should have a power rating and radio
emission data in it's instruction pamphlet.I know for a fact that my electronic
keyboards are rated according to radio emission rules. My Yamaha PSS-480
is registered as a class B computing device compliant with part 15 of the
FCC regulations. If people bothered to check,most likely their phones have
similar data already associated with them,and if they check the FCC regulations
they should be able to find out how strong an emission is,what frequency
it is and what effects it's likely to have.But that requires effort which
has most likely not been put in.The data is there it's just that Joe Public
is too lazy to look. I noted yesterday that our Hoover affected by VDU,and
thus the Hoover is emitting EM radiation and generating magnetic interference,why
no Hoover scare? The comparison will mean low radiation will be a selling
point much as MPG is for cars,but one would not make a selling point out
of how unpowerful or small the engine was,and the radiation in a phone is
a measure of it's ability to transmit a signal.If you buy a low dose phone,that
is a weak phone that will be less able to do it's job! It's swings and
roundabouts if you want convenience you have to accept risk.If you don't
want the risk then stop asking for convenience.Science tries to deliver
convenience with low risk,but 100% convenience with 0% risk is impossible
-LB]
And also we suggested there should be an audit to show the levels of radiation
at different parts of the country so people can see at least where they stand
in the league tables of radiation.
Colin Berry : And other sources presumably?
Colin Blakemore : Well that's right yes.
Mike Maier : But again....
Colin Blakemore : And also to involve the public more in making decisions
about how the phone system should grow.At the moment you don't require planning
permission to put a mast up,if it's below 15 metres in height. 15 metres?
And we suggested that full planning permission should be needed for that.
[Scientists working in the public interests contrary to common
conception.Scientists are people to,not a breed apart,and stand to suffer
from radiation as well.It's in their best interests to minimise possible
problems -LB]
Mike Maier : But you are struggling with...
Lynn Frewer : But the most extreme version of the precautionary principle
would mean that you would ban mobile phones. An how do you......
[Let's ban cars,and TVs and CDs and VCRs and VDUs because after all what
use are they and they are so dangerous aren't they? -LB]
Colin Berry : I have a very grave suspicion about the use of the
precautionary principle,from the point of view of science (murmurs of
assent),because it demands that you attach a different weighting,a different
-what's called an "epistemic warrant" - a different weighting to some kinds
of information versus others,and that's the antithesis of science,and if
you actually applied it say to some parts of pharmacy and pharmaceutical
development,it would lead you down some terribly dangerous paths I think.
[Those paths are more likely to be trod if we allow public value judgements
to affect policy as those weightings will become considered and applied,because
the public aren't scientists and don't understand how their personal feelings
and ideas are irrelevant to how science is carried out.If science worked
by personal whim then the moon might be made out of cheese until the moon
landings showed it weren't.Science's power exists because those personal
subjective whims are outlawed and not taken into account,that's WHY it
works.Accommodating public "I think this should happen because I believe
in God" arguments destroys what science is and undermines our modern
society.People don't realise that they are digging away at the very foundations
of that which supports their standard of living -LB]
Mike Maier : Of course you forget the great benefits the mobile phones
have given to people,I mean there was an article in "The Lancet" a couple
of years ago,from Australia outlining the lives it's saved,you know the...in
remote and far away places,(murmurs of assent) of 300,000 - I'm trying to
remember 300,000 phone calls made by boats,you know from boats in distress
on the sea...
Colin Berry : Even Dartmoor you don't have to go very far!
Mike Maier : ..yes,millions from people having MIs (?).Half a million
from people having had accidents in remote places,it's saved countless
lives,irrespective...
[And how many have died from a cellphone? None -LB]
Lynn Frewer : Isn't that an important point..
Mike Maier : ....of course it is.Yes.
Lynn Frewer : ...in the whole debate about perception? That people
perceive that exposure to a risk is necessary for some other reason.I think
perhaps perceptions of need are even more important than perceptions of direct
personal benefit.
[I don't think that's true. The hauliers and farmers were selfishly concerned
for the well being of themselves and their livelihoods and apparently negated
the risk incurred to everyone else that we might be starved to death.They
said "We are doing it for the country" without ever asking anyone or being
voted for.I don't support farmers I think they are selfish materialist money
grabbers,and what with Tony Martin,probably unstable misfits who have no
idea of the rule of law,and think only of protecting their property and their
interests to the exclusion of all else.The NEED for food did not seem to
be a concern for them in their action -LB]
Colin Blakemore : But one area where people do seem to respond very
positively to medical advice even though it may not actually make sense in
an objective way,is in accepting vaccination of their children.Could you
tell us about that?
[Yes! That IS indicative. A vaccination is a small weakened version of the
disease in question to familiarise the immune system with the potential threat.By
vaccinating the body is put at risk from the disease,and yet people accept
vaccinations almost unquestioningly. Perhaps this is because their track
record has been good in stemming the diseases? It's odd that acceptance of
vaccination happens when issue is taken elsewhere such as with cellphones
when the same policy of risk management governs both! -LB]
Colin Berry : Well I think here,if we're talking about risk
evaluations,things are very clear.Let's take measles for example.A miserable
disease if a child just gets simple measles,a dangerous one if it affects
its lungs and it gets pneumonia and can leave - even if it recovers,which
- some of them unhappily die,and particularly in other countries - may leave
damage in the lungs which leads on to long term problem in the lung in
Bronchiectatis(?),and rarely - as you will know very well - can affect the
central nervous system,producing a very terrible disease.So there I think
if there were a risk attached to the vaccine,and I am not personally convinced
that there is,the risk benefit analysis is very easy for most parents to
see. I think the same is true of mumps,where in a sense,the same sorts of
complications are also true of mumps as a natural disease,
[Be careful Colin- some people think that you scientists manufacture these
diseases! -LB]
and the rubella which is the third part of that thing,the benefit is there
really quite different,it's to the next generation isn't it?
Colin Blakemore : Yes.
Colin Berry : If it's a girl,as we started wrongly in this country
just immunising girls
[In the current climate this might be interpreted as "Male establishment
tests out its scientific poisons on subjugated women,rather as JA Paulos
quotes Mort Sahl (see paulos2) -LB]
the benefit is to their potential children.We started immunising boys,because
we realised that with the herd effect - that's to say if you get a significant
part of the population immunised,it depends a little bit on which disease,but
something like 75-80%,the rest won't get it anyway whether they are immunised
or not,because they are protected by the fact that the disease won't spread.
[Sir Robert May's work adds Chaos Theory into such equations (see
beffect),and in "Outbreak",the idea of not spreading
through the disease say having a short lifetime in the body (or the host
dying and being unable to travel) was thought sufficient (initially) to contain
the spread,this in turn depends on the medium of exchange and incubation
period.AIDS (or HIV) is thought to have a long incubation period,and even
though a host may not know they are a carrier,they nevertheless can be.What
is certain here is that mathematics both applies and can help as a means
to understanding and controlling disease patterns.I recall one programme
that Robert May was on where he said that because of Chaos Theory, vaccination
could actually make the potential to spread worse. Thus those parents doing
"cost benefit analysis" (do they?),have Chaos Theory to consider before
vaccinating a child, but I bet they don't -LB]
Those are three diseases which are terribly common and which in unimmunised
populations produce devastating epidemics,with regularity.For meningitis,an
interesting new vaccine,the chances of your getting meningitis are really
very small as an individual,and I think we are in an uncertainty phase -
we don't know about the vaccine yet,it hasn't been given to enough people
for long enough,though if there are risks they are small.
[I had to recently in lieu of scientist being present say that scientists
were honest enough to say "we don't know" in certain instances.Here Colin
says exactly that.Note that as far as the woman who wished to be natural
and not have untested things go into the public domain is concerned,the
meningitis vaccine would be "unnatural" and "unproved",nevertheless unnatural
things have to be done to stop natural things from killing people,any
intervention by man could be seen as "unnatural" but it isn't.Under that
auspices everything man has ever done is "unnatural". Similarly the vaccine
cannot be proven until it shows that it is impeding meningitis in people
which is the litmus test. So to some extent an "unproved" thing will always
enter the public domain,it's naive to suppose it won't -LB]
Now there's a different kind of risk benefit evaluation to be done,because
of the consequences of you getting meningitis are so catastrophic,
[This is the same point I made earlier over BNFL and Lois Wolpert -LB]
even...I mean firstly a high death rate.Secondly a very high morbidity and
mortality,people lose limbs because of the scepticemia that often accompanies
the common form.All of these mean that the risk benefit analysis,in my
mind,strongly favours immunisation.I've seen fatal cases of whooping cough,and
of measles,in this country,in recent years,I'm talking in the last ten.These
diseases still exist,we have them controlled.If we stop immunising large
segments of the population,they will reappear.That's a certainty,I mean....
Colin Blakemore : Yes.
Kenneth Macrae : We have a very good uptake of vaccinations in the
UK,but we're very proactive,we have you know,health visitors and GPs who
are on incentive bonuses and so on.We go out and vaccinate children.The parents
who I speak to who ask me for advice, actually are not that keen on it,and
it's because of the herd immunity issue,because these illnesses are not epidemic
in this country and so you can get away with not vaccinating your child
actually,but of course it's a high risk thing to do,for the general
population,because if enough children don't get vaccinated,there will suddenly
be epidemics again and all those unvaccinated children will suffer.
[This sudden tipping from one state to another due to slow variation is subject
to Catastrophe Theory,which general model is that of sand grains that suddenly
slide as one by one they move.I'd be interested in whether different cultural
heritages have opposing views to vaccination,and if so how they justify taking
the risk of not being vaccinated upon the potential to affect the rest of
the population who might not share their beliefs.It's one thing for a Jehovah's
Witness not to accept transfusion when only their life is on the line,but
if a belief impacts on the rest of the community who don't share that belief,and
those with the belief aren't vaccinated,that could mean someone dies because
someone of belief didn't vaccinate,the belief thus will bear some of the
responsibility for the incurred death. The sensitivity to initial conditions
and potential to boom and bust are characteristic of Chaos Theory,and it
amazes me that people take unilateral decisions without being informed of
these ideas that fundamentally should be informing them of which decision
to make -LB]
So I think it is a very important issue,and again part of the equation of
this risk assessment that one has to make.
Colin Blakemore : Yes,well risk is not going to go away but unfortunately
we have to,so thanks to Sir Colin Berry,Lynn Frewer,Ken Macrae,Mike Maier
and to you for watching goodbye.
[email protected]
Risk and Bayesian Statistics
Dear Prof,Sorry to bother you twice close together (but things come
in threes if Ch4's maths on buses is to be accepted,so I'm the expected clumping
factor!),and I hope you've eaten the right meal for breakfast , but New Scientist
reports this week on the low risk of Killer Stranglets eating up the planet.My
concern is not the physics of the anti matter,but the assessment of risk.I
have many pages covering this online,because as Sue Blackmore points out,it
is the inability to assess it properly that leads to belief in pseudo sciences
and/or knee-jerk reactions to news reports.The upshot is this: Adrian Kent
argues that because of the possible incurred deaths,the risk upper limit
should be set higher compared to another event,ie the possible outcome is
being taken into consideration.In my own pages,I think Lois Wolpert is quoted
as saying that holes in the road are far more dangerous for cyclists as they
are a greater risk than a nuclear accident.This may mathematically be the
case,but pot holes are unlikely to wipe out large numbers of people in one
catastrophic event.The director of BNFL also once said that a nuclear breakdown
was likely to happen once in a million years.This may indeed be true,but
we have suffered many events in recent history.The risk does not mean that
events can't clump together as far as I understand it.Ie something can happen
more than once in any given set of million times. What bothers me is the
notion of Bayesian Analysis,where I think an extra term is added on for the
possible outcome in calculating probability (I've seen Fisher Dilke talk
on this but I might be a bit shy of quite wholly understanding it).It seems
to me that Adrian Kent is making use of this type of argument and Lois Wolpert
is not,is there any right answer here,or is it dependent on one's view?I
only ask as according to one publication,you are the "most famous mathematician
in the world" and I have to explain these things to those that have a tendency
to the pseudo sciences and fail to understand risk,and if fame is an indication
of authority,I'd like to get the info from the horses mouth,as it were.
Risk and Bayesian Statistics
I wrote a piece on risk for New Scientist, ages back, that made several
points very similar to yours. In particular standard risk analysis assumes
that what matters is the AVERAGE. That is, a one in ten chance of killing
ten people is the same as a one in a million chance of killing a million
people. To my mind, this is nuts--- and it leads to idiotic comparisons between
holes in the road and nuclear accidents.
For each individual, it may be the case that the risk is the same,
assuming all that matters to them is their own life. But a nuclear disaster
would reduce social structure to ruins (look at the effect of a puny fuel
protest) whereas a million cyclists killing themselves in holes (over a period
of, say, a century) would hardly be noticed. The rest of science is going
nonlinear, and I think it's high time risk analysis did the same. Ian
Stewart |
|
|
|
|
|
|
|
|
|
|