Fallacy

Paul Snively http://psnively.github.io/blog/2015/01/22/Fallacy/#comment-1881981921 asks:

 

“I think Christian’s response #4 is no response to the critique on his work, because it responds to a non-issue. So it may have some interesting content but it is besides the point: whether or not Bell was wrong? I think you don’t understand Bell’s theorem well enough hence you think that Christian came up with a smart answer to his critics.”

You don’t have to “understand” Bell’s theorem, as such, at all to demonstrate how it fails. I’m willing to break it down, step by step, if you are. Disqus hasn’t uttered a peep about our extravagant consumption of their resources—yet. :-) I will ask one yes-or-no question per comment. You may answer with either “yes” or “no.” If you wish to elaborate, please do so on your own blog or other freely publicly accessible forum of your choice and provide a link along with your “yes” or “no.” Let’s begin.

Do you understand that Dr. Christian’s work consists of two arguments:

1. Bell’s theorem fails to be a no-go theorem.
2. Given 1), here is a locally realistic model that predicts what we find in quantum mechanical experiments.

Yes or no?

My answer was no and I gave three links: http://arxiv.org/abs/1207.5103, http://arxiv.org/abs/1203.1504, http://arxiv.org/abs/1412.2677

However three links was too many so now I only give one, to this blog.

Actually first of all I answered yes but with a proviso, because I do understand, of course, that this is the intended content of Christian’s work, but on the other hand, both the two arguments are actually wrong. Paul’s question was “ill-formed”. A bit like “when did you stop beating your wife?” (you must reply by giving a date, otherwise I will delete your answer from my blog).

(1) is wrong: see J.O. Weatherall (2013). The Scope and Generality of Bell’s Theorem.
Found. Phys. 43, 1153–1169.

(2) is wrong: Christian’s “model” is logically flawed. Simulate the model in the one-page paper and it certainly doesn’t reproduce the singlet correlations. But anyway, who cares: Bell’s theorem proves it is impossible. It is impossible to simulate the singlet correlations by a LHV model in the rigorous constraints of a decent Bell-CHSH type experiment. Bell’s theorem can be seen as a theorem about distributed (classical) computing. What can be computed, what can’t. It tells us, in view of the fact that apparently nature can violate Bell inequalities (ie. according to quantum mechanics) that nature cannot be understood, even approximately, as a discrete stochastic classical automaton.

The difficulty with answering yes/no questions is that both answers can often be wrong. Either answer, without proviso or explanation, can be misleading. “Ask a stupid question, get a stupid answer”. This is why when you ask a Zen master a question he answers yes but nods his head for no (or vice versa).

Next time I’ll answer “yes and no”.

Paul Snively shows by his questions, so far, that he doesn’t know what he’s talking about. Which in itself is interesting, of course. And nothing to be ashamed or, either.

The UK Lucia

Today saw the publication of a paper entitled “In Search of the ‘Angels of Death’: Conceptualising the Contemporary Nurse Healthcare Serial Killer” by Elisabeth Yardley and David Wilson (Birmingham City University, Centre for Applied Criminology). It appears (2014) in Wiley’s “Journal of Investigative Psychology and Offender Profiling” (J. Investig. Psych. Offender Profil.) http://onlinelibrary.wiley.com/doi/10.1002/jip.1434/abstract

I have many concerns about this paper but my first is the following: it is based on 16 past cases, at least 2 of which seem almost certain to have been rather serious miscarriages of justice. In fact an earlier version of the methodology proposed by the authors was used against Lucia de Berk – the prosecution brought in an FBI expert who has contributed to one of the first publications on HCSK’s (Health Care Serial Killers) … and in which he made use of his inside knowledge of the Lucia case.

So looking at this sample, N = 16 with at least 2 false positives (Colin Norris, Ben Geen), what can be concluded? (Probably at least 3 false positives: the Finnish case of Aino Nykopp-Koski  is highly controversial). The checklist of “red flags” certainly identifies those who are going to be successfully prosecuted as an HCSK. In fact the reason they are being prosecuted is precisely because of some of the red flags. Others are subjective evaluations by fellow nurses and doctors and are contributed to the prosecution case *after* it has started. The most pernicious point is that an investigation into a suspected HCSK starts internally in a hospital and is carried out by doctors who themselves, formally, should be seen as suspects (their own patients are involved…). How objective will they be? What we know from the cases of Lucia, Colin Norris and Ben Geen, is that in this situation, doctors trawl for recent “suspicious” cases, re-writing or at least re-explaining medical events, which till then had not been considered suspicious at all. This results in a compelling dossier which goes to the police … media, police and courts do the rest.

The appearance of this landmark paper was reported in the Guardian newspaper. Ben Geen got some free publicity, the newspaper even included his photograph, taken during trial or appeal. Don’t expect a photograph of a young man in the middle of an appalling nightmare to present a “picture of innocence”. http://www.theguardian.com/uk-news/2014/nov/22/study-identified-key-traits-serial-killer-nurses

Here’s the checklist. Lucia had an enormous high score according to the evidence presented to the court. But not if the evidence is “corrected” with a view to correspondence to truth.

1. Moves from one hospital to another

2. Secretive/difficult personal relationships

3. History of mental instability/depression

4. Predicts when someone will die

5. Makes odd comments/claims to be ‘jinxed’

6. Likes to talk about death/odd behaviours when someone dies

7. Higher incidences of death on his/her shift

8. Seems inordinately enthused about his/her skills

9. Makes inconsistent statements when challenged about deaths

10. Prefers nightshifts—fewer colleagues about

11. Associated with incidents at other hospitals

12. Been involved with other criminal activities

13. Makes colleagues anxious/suspicious

14. Craves attention

15. Tries to prevent others checking on his/her patients

16. Hangs around during investigations of deaths

17. In possession of drugs at home/in locker

18. Lied about personal information

19. In possession of books about poison/serial murder

20. Has had disciplinary problems

21. Appears to have a personality disorder

22. Has a substance abuse problem

I emailed Prof. David Wilson in order to start an academic discussion with him about his research methodology. He said that I couldn’t criticise the paper because it has been published in a peer reviewed journal. And now he has blocked my emails. I have had the same experience when attempting to start a conversation with earlier researchers in this field. It seems that criminology is a field of science with impeccable research methodology and therefore above any criticism from other scientists.

Inside the Mind of HCSK Professionals

De Ene Zijn Dood is de Ander Zijn Brood, or One Man’s Death is Another Man’s Living. The public has a great fascination with murder, carried out by people whom we trust in situations in which we are helpless. Here are two books which capitalize on this: Inside the Minds of Health-Care Serial Killers by Katherine Ramsland; and Engelen des doods (Angels of Death), by Paula Lampe.
Continue reading

Justice in the Netherlands: Guilty until proven Innocent

Kevin Sweeney is serving a life sentence for murder of his wife by arson.

Here is a link to his own site http://www.justiceforkevinsweeney.com. You can read the conclusion of the court under LJN number AB0493 at www.rechtspraak.nl. Investigative reporter Peter R. de Vries reports on the case in his dossier Suzanne Davis. Curiously, Peter de Vries admits that there is absolutely no proof of arson and murder yet he appears to have a lot of sympathy for the victim of the fire, and none for the person who is accused of starting it. He has no interest now in supporting Sweeney’s claim that he is innocent and the fire was an accident (probably started by smoking in bed).

Negative publicity about Sweeney, demonstrably untrue and much of it spread by certain of the close relatives of the victim, has poisoned a lot of the news items on his case, see for instance the story on Sweeney at www.expatica.com, a web-site for Brits in the Netherlands. However independent lawyers and scientists who have become involved in his case are convinced that the fire was an accident, see for instance Fair Trial International’s dossier on the case http://www.fairtrials.net/index.php/cases/spotlight/kevin_sweeney.

On studying the scientific evidence which secured the conviction (results of experiments by TNO on reproducing the fire damage, and the pathology evidence concerning the cause and time of death), I can only agree that the prosecution’s story is totally in contradiction with all known facts, while that of the defense is totally in agreement with them.

It seems that Sweeney’s charming personality, high intelligence (his IQ of 144 is one of the legally established facts supporting his conviction) and spirited and elaborate defense of himself convinced Dutch judges that he was evil and manipulative. The fact that the fire started about the same time Sweeney arrived in Brussels one hundred kilometers away only confirms these judgements. The TNO experiments showed that the fire could NOT have been caused by a naked flame applied to 8 litres of fuel, and did nothing to disprove that it was caused by a burning cigarette fallen on bed-linen or whatever. Police investigators stated that the idea a fire could be caused by smoking in bed “belongs in the realm of fables” (a turn of phrase much admired by the judge, who quotes it in the summing up) whereas this is one of the most common types of fires, and the most common cause of fire deaths, the world over. Statistical evidence that such fires also occur in the Netherlands was not admitted by the judge who preferred to believe in the word of a policeman, out to secure a conviction. The TNO experiments were so complex and expensive that they had to incriminate the suspect, whatever the outcome, despite the fact that they were spectacularly ill-designed and inconclusive.

I notice the following similarities with the case of Lucia de Berk: the suspect is intelligent, a strong personality, a sympathetic but not run-of-the-mill person, with an unusual (complex) background and personal history involving much time spent abroad; the suspect never stops asserting their innocence; the crimes are so perfect that they are actually impossible; the case involves a huge amount of complex multidisciplinary scientific evidence. Scientists from any particular field know that the evidence of their colleagues in their own field was worthless but don’t find it necessary to protest, since everyone knows that the suspect was a bad person who probably did kill his wife. The police suppressed evidence supporting the defence case, manipulated forensic evidence, got witnesses to change their statements and to lie. The prosecution spread slander and gossip about the suspect and paints a beautiful picture of the victim which was greedily repeated by the media, known by scientific experts giving witness in the case, and embellished by the judges in their conclusions.

At the first hearing, the case collapsed; the prosecution appealed and spent three years on the TNO fire experiments. Fully documented support of all the statements I have just made were available to the Appeal Court (which found Sweeney guilty), to the Supreme Court, and to the European Court of Human Rights. All these courts have ignored them totally. Justice by gossip, with science as a willing accessory

Here is a report I’m writing for Sweeney’s lawyers.

 

Continue reading

Little Baba Must be Hung

“Barbertje moet hangen” (Little barber must be hung). This will require some explanation, and not just for the non-Dutch reader. Yesterday I asked a class if anyone had read Max Havelaar by Multatuli. 7 had not, one had read about half. And thrown the book away in disgust because it was all just the same as ever. Those who forget their history are condemned to repeat it. It seems the Dutch have forgotten their literature and hence their history and are repeating it. I must write more on this elsewhere and add the good links. In the meantime try Googling some of the unknown words and phrases here. 


Continue reading

Lies, damned lies and amateur statistics

Lies, damned lies and amateur statistics by Piet Groeneboom 
Piet’s blog actually has a different title, but same theme. Here’s my take on the matter. 
It has been known for a long time that “careless statistics costs lives” and I am referring specifically to bad statistics in medical research. 15 years ago about 90% of statistics in medical journals was wrong, things have improved, now it’s only 50%. Consistently across journals, across sub-fields. The most frequent error is the misunderstanding of p-values and a common recommendation is to have them banned. All this does literally cost lives: the good treatments are not discovered, time and resources … and hence lives … are lost following up “spurious correlations” (often discovered during fishing expeditions and/or using inappropriate statistical methods). Sally Clarke is another example of a life lost to amateur statistics (amateur statistics of an arrogant and self-satisfied medical specialist who transferred his “scientific conclusions” into legal brains with ease). For a good laugh (but perhaps the laugh of a farmer with toothache, as we say here in the polder) enjoy Peter Donnelly’s TED lecture. One of the many scientific papers carefully analysing abuse of statistics remarked how strange it is that we insist on getting brain surgery from a professional brain-surgeon, but are happy to have our statistics done by an amateur. Well, people who rely on amateur statistics, or worse still, are proud of their own, ought to go and see a brain-surgeon (for my very special friends: sshhh, I know this great Polish plumber … ).


Continue reading

Sticky Balls


Sticky Balls

Today (26 May) the new notion of sticky balls statistics (kleverige knikkers, in Dutch) was born. Well, the notion already existed: overdispersion due to the confounding factor time. But the new name should make it easier to get the idea across to sharp legal minds. The prosecution in the Lucia de Berk case only knew about the statistics of pulling nice new shiny balls out of a vase. The latest research shows that the balls were sticky. (Actually, the nice hospital administrators and policemen had a bit sticky fingers too, but that’s another issue). It was chance all right. Just a little bit of bad luck. Bad statistics. That was very bad luck. Doing Tarot cards was not a good idea either (especially in combination with keeping an odd diary and having overdue Stephen King books from the library in your bookshelf). Illustrations: the Chinese delicacy “sticky balls” eaten at the Spring Festival, and aptly symbolizing “reunion”: fromChengdu (the white ones), and Suzhou (green).


Continue reading

How to Lie with Statistics – the Trilogy


How to Lie with Statistics – the Trilogy

Latest news from Cannes: Darrell Bluff announces the long awaited “prequels” (with George Clooney as the Good Statistician, Björk as Lucia, and Donald Sutherland as Judge Death) to his cult movie “How to Lie with Statistics”. Parts –2 and –1 of the already celebrated trilogy will be called How to Die with Statistics, and How to Lie with Legal Facts. Asked whether the cliff-hanging conclusion of the trilogy is going to be resolved by a true sequel, and if so how, Bluff refused to comment, though recent sightings of him enjoying numerous Grimbergen Triples (fine Belgian Trappist beer) with Clint Eastwood are fueling intense speculation among insiders here in Cannes-sur-Veluwe.

 


Continue reading

Lies, damned lies, and legal truths


Lies, damned lies, and legal truthsPortrait of George Canning

If there’s one thing which makes a statistician sick, it’s having people always quote “Lies, damned lies and statistics” at them, and looking expectantly (or sheepishly, or arrogantly) at you as they await wise or witty repartee. Sometimes they refer to Mark Twain, sometimes to Disraeli. Well, Disraeli is apocryphal; Twain was actually misquoting the famous British Prime Minister George Canning who about 1820 said “you can prove everything with statistics except the truth”. (Canning was probably the best PM Britain ever had, but his term was one of the Book Covershortest, since he died “young” from a cold caught while at a funeral in the rain). I have been looking for a long time for a counter “bon mot”. And I think I have found it.


Continue reading