I keep hearing about this Amazon thing. Maybe some day I will try it.

bad math from academics lands nurse in jail

Looks like she will be cleared, but it just goes to show you... something.

http://www.guardian.co.uk/commentisfree/2010/apr/10/bad-science-dutch-nurse-case

The story is that in the three years she worked at the hospital, 6 patients on her floor died. They had a mathematician testify that that was impossible and convicted her of being a serial killer, on that, and no other evidence.

The three years before she started working there? Seven died on that floor. Something not mentioned in the trial since it was irrelevant, and also mathematically impossible.
Permalink CC 
April 12th, 2010 2:27am
It is crazy and the law is out of hand.

Why would they prosecute?  Because someone died in a hospital?

Wow.
Permalink Bot Berlin - a stupid f**k troll 
April 12th, 2010 2:29am
The stress from it all caused her to have a stroke and now she is partially paralyzed.

In the US, you have things like IRS Golden Eagle audits where they can convict you of tax fraud based on no evidence, just the fact that something in your life is a statistical anomaly. The NSA FBI CIA are also doing such computer drag nets of data looking for things that are outliers. Being an outlier is now accepted by courts as proof of guilt of a crime, just by itself.

Statistically this is going to lead to lots of false convictions.

But the prison industrial complex will do well. Some are even publicly traded.
Permalink CC 
April 12th, 2010 2:59am
In her diary she had written to have committed the "murders" on the patients.

So the prosecution strongly suspected her, and tried to force evidence.

It has been a high profile case here for a long time, and now a textbook case of criminal court cases gone wrong.
Permalink Attila 
April 12th, 2010 3:33am
Whaaaaaat?

She confessed?

Strange there was no mention of that in this latest article!!! WTf.
Permalink CC 
April 12th, 2010 3:37am
OK, so when she was held in an insane asylum, another patient there who was committed claimed that she told him she confessed. Later, he recanted and said it was one of many stories he makes up.

That's very different from YOUR version of the story, for which I am now going to request from you a citation.
Permalink CC 
April 12th, 2010 3:40am
Statistician reviews the math and says it is bullshit.

http://www.math.leidenuniv.nl/~gill/lucia.html
Permalink CC 
April 12th, 2010 3:42am
Decent summary.

http://en.wikipedia.org/wiki/Lucia_de_Berk
Permalink CC 
April 12th, 2010 3:47am
"The chance of a nurse working at the three hospitals being present at the scene of so many unexplained deaths and resuscitations is one in 342 million." - testimony used to convict her

"the chance of a nurse working at the three hospitals being present at the scene of so many unexplained deaths and resuscitations is one in 25." - analysis by qualified statisticians
Permalink CC 
April 12th, 2010 3:49am
from http://www.math.leidenuniv.nl/~gill/hetero6.pdf
Permalink CC 
April 12th, 2010 3:49am
Page 3 of that pdf:

"All these corrections result in a total of 14 incidents in Lucia’s shifts, instead of 7, out of a grand total of 32 incidents in 1734 shifts. Taking again ρ = 1 we find the probability 0.036429 or one in twenty-five."

"A modest amount of variation makes the chance that an innnocent nurse experiences at least as many incidents as the number Lucia actually did experience, the somewhat unremarkable one in nine. Making some less favourable
choices to her in the data cleaning process, only decreases this chance to one in twenty-five."

One in 9.
Permalink CC 
April 12th, 2010 3:51am
Where is argv? He should come here to prove all I say is bullshit and this lady is guilty.
Permalink CC 
April 12th, 2010 3:51am
Useful reading for jurors.

http://en.wikipedia.org/wiki/Prosecutor%27s_fallacy

"The prosecutor's fallacy is any of several fallacies of statistical reasoning often used in legal arguments."

With several examples of how prosecutors use bullshit math they know is bullshit to secure false prosecutions.
Permalink CC 
April 12th, 2010 3:54am
This lady was convicted of murdering her children because  Sir Meadows, expert math witness, said the chance two children in one family would have cot death was 1 in 76 million.

http://www.guardian.co.uk/society/2007/mar/17/childrensservices.uknews

Since then statisticians have said that the actual chances are 1 in 100.
Permalink CC 
April 12th, 2010 4:02am
Notice that she was released from prison after a couple years, but was so depressed she drank a lot and then died from it.
Permalink CC 
April 12th, 2010 4:03am
Interesting in her case is that the prosecution's pathologist told them that the children had died of infections, not murder. But this did not match the prosecution's agenda of securing a conviction, so that information was not disclosed in court, or to the defense.

To make clear - the prosecutor knew for a fact she was innocent all through the trial.
Permalink CC 
April 12th, 2010 4:12am

This topic is archived. No further replies will be accepted.

Other topics: April, 2010 Other topics: April, 2010 Recent topics Recent topics