Continuation - Discussion of the Amanda Knox case

[Chris_Halkides]

Nature article previously cited

I've quoted this myself, and personally I believe I was the first to dig out the CPS quote back when we were all talking about the open letter. It doesn't say that if you can't repeat the test you throw the results in the bin, or at least that isn't my reading.

shuttlt,

"Bruce Budowle, a forensic and investigative geneticist at the University of North Texas in Fort Worth and a former scientist at the Federal Bureau of Investigation (FBI), gave evidence criticizing low-copy-number analysis at the Reeds' appeal trial. He argues that the methods for determining whether a signal is true in low-copy-number analysis haven't been validated.

In low-copy-number profiling, forensic scientists generally split their limited amount of DNA into two or three samples and run analyses on two of them. The third, if available, is reserved for the defence. The results of analyses aren't completely reproducible, profiles often won't match and the scientists generally accept as true those STR signals that show up in both runs. The practice is worrying, says Budowle: 'The logic of this approach makes some sense, but the confidence in it has not been assessed.'"

Running a test once is clearly inferior to running it multiple times. I read the CPS standards as meaning that nothing can be said about a test done only once. The defense should also have the opportunity to retest a sample in a laboratory of their choice.
There is not one scintilla of evidence that ILE invested in a state of the art facility for low template work. Beyond that, there are the handling techniques in Dr. Stefanoni's lab, which are not good by traditional profiling standards. There are also some good quotes from Dan Krane and Allan Jamison in the article from which I took the quote above.

 

[RoseMontague]

The special procedures which Stefanoni didn't follow have been listed numerous times already. They include physical separation of the LCN testing area from the rest of the DNA lab; special air-handling precautions including positive-pressure ventilation; extreme sterilisation of all equipment, including UV sterilisation; additional protective clothing including IIRC breathing filtering apparatus and total facial covering; and of course the replication aspects of the LCN analysis.

I believe she is looking for a document similar to the one I mentioned on the FBI site:

http://www.fbi.gov/about-us/lab/codis/swgdam-interpretation-guidelines

That's a lot of guidelines just for regular testing and interpretation and it makes it clear it does not apply to LCN (they would be more stringent is the implication).

This link might be more relevant than the interpretation guidelines

http://www.cstl.nist.gov/strbase/QAS/Final-FBI-Director-Forensic-Standards.pdf

 

[LondonJohn]

How do you interpret the last sentence? For me, I see it as saying they don't calculate the odds of the match.

By the way, how do you interpret the sentence after where you stopped quoting where it says that even after failing to reproduce a result, it is still disclosed?

Yes - it basically means that only the repeated peaks are used to construct the profile against which the reference profile is measured, and from which the odds of a random match are then calculated. Peaks which only show up on one of the two tests are excluded.

The sentence after my quote is this:

If the result is not reproducible or is a complex mixture then it is disclosed but no calculation is carried out.

I would take this to clearly mean that the results of the tests (the numerical and graphical data) are provided to investigators, but that no calculation regarding a match to the reference sample is possible. In other words, it's useless as potential evidence against the person whose reference sample is being compared.

 

[Chris_Halkides]

no control

Not that it's central to the issue, but generally these lists seem to be focused on contamination making the LCN analysis useless. Facial covering for example is about preventing the testers DNA getting into the sample rather than some randomly matched suspects DNA. It's much harder to find discussions of what steps are necessary to prevent the kind of contamination we are talking about here.

shuttlt,

You are ignoring the fact that the lab would have had much of Meredith's DNA from all of the evidence samples they collected. Chemical and UV light treatments are for destroying all extraneous DNA, not just the DNA from the technicians. In addition, Peter Gill mentioned the importance of negative controls in the Nature article. These negative controls would have to be performed and interpreted in the same way as the knife profile. As with so many other things, the defense did not get the actual negative controls in electronic form.

 

[shuttlt]

Well, he is a journalist and journalists do at times get facts wrong or jumbled up.

Is he a journalist? Perhaps in a new media sense he is. As you say, people and journalists make errors. This seems like a significant error, if it is an error, though in that he is claiming to have personally heard a bunch of stuff that he didn't.

I do think he is probably the most reliable journalist of fact on this case, however, he has a different way of expressing facts than the mainstream media does.

He certainly has access to a lot of facts that others don't. Equally he comes with a point of view and his style is hardly striving for objectivity any more than Dr Waterbury's, or TrueJustice.

 

[LondonJohn]

The "information area" of PMF has come up with an interesting translation of an article from the Umbria Journal, discussing the views of a chemistry professor at Meredith's former university, Leeds (thanks to Jools for the identification and translation of the article):

LEEDS TEACHER: "JUSTICE BUT NOT WITH INNOCENTS JAILED"

Meredith Kercher “should have justice, but not at the expense of keeping two innocent persons in jail” quoted, as told in an interview with the weekly magazine OGGI, by a chemistry teacher from Leeds University where the English student murdered in Perugia studied, Michael Krom explains he’s is come to be an ardent follower of the case, having thoroughly studied it and become an “innocentista” regarding Amanda Knox and Raffaele Sollecito. “The thing that I’m most anxious - the teacher tells OGGI- is that Meredith Kercher gets justice. Condemning two innocent people breaks two lives but it also creates a third victim: the memory of Meredith." In the opinion of professor Krom there are no evidence to catch the two former sweethearts. “Let's start with Amanda – he adds -.

There is let’s say an anthropological objection. The cases of women participating in a rape and then kill are very rare. In the UK there have been two in 50 years. Circumstances different from those of Perugia, it was of women subjugated by psychopathic husbands that had childhoods full of violence and deprivation and the evidence against them were overwhelming. What is there against Knox? What is it of hers in the crime house? Nothing, zero… And Raffaeles’s position is even more absurd. Just a fragment of DNA on the hook ‘gets’ him. Here we are in my expertise field. And I say: in the case of the hook suspicions of contamination are enormous. It was recovered a month and a half after the murder. It was found in a different place from the original. As for the knife alleged as the murder weapon, the DNA traces are too low, too scarce to give reliable results in tests. It’s one of the first things I teach my students; when the traces are too low testing is not even done. On site we all agree: Raf (this is how he calls him -Ed.-) is collateral damage -concludes Krom- in a war that does not concern him.”

http://www.umbriajournal.com/mediac...ocente-leeds-giustizia-ma-no-a-innocenti.html

 

[shuttlt]

shuttlt,
Running a test once is clearly inferior to running it multiple times. I read the CPS standards as meaning that nothing can be said about a test done only once. The defense should also have the opportunity to retest a sample in a laboratory of their choice.

I don't see how you get this from them saying a non-reproducible result should be disclosed. It seems pretty clear from the article that they include results that retests fail to reproduce.

 

[Chris_Halkides]

halides 1: "Neither Amanda nor Raffaele have ever used the word amnesia, nor do they claim to have used cannabis heavily (to the best of my understanding)."

Nov 7th Sollecito:

"I'm accountable that if we all ended up in jail it is also the fault of my light regard to the facts of that evening and also that we smoked (Amanda and I) several joints. "

Sounds like he's blaming the mess on drugs, to me.

colonelhall,

I think Amanda said that she smoked one joint, but I could be wrong. Raffaele also said something along the lines that he was in a holiday mood, meaning (I suppose) not taking careful note of the time, for example.

 

[platonov]

Where was Sollecito when he made his appearance before Judge Matteini?

You have a really odd definition of the term "in court".

My definition of 'fatuous arguments' is spot on however.

If the answer to the Q posed in the original post below is not 'NO' - lets hear it.

Otherwise - feel free to have a debate with Rose Montague on this; you already seem to be in agreement

We have AK's own testimony as her last words on this alibi - do we have anything from RS since his lawyers went to court [pretrial] to 'move' that the 'evidence' against AK shouldn't also be held against him.
Doesn't sound like a strong alibi claim - and no, the fatuous arguments that this meant he couldn't know where she was when he was asleep don't carry any weight.

Did he for example ever make a spontaneous statement in court that AK was with him all night - or from 9.10 till 1.00 (bedtime) if you prefer.

I don't think so.

He obviously never took the stand or AFAIK answered Q's after Nov 8 on this (or any) issue.

 

[christianahannah]

I believe she is looking for a document similar to the one I mentioned on the FBI site:

http://www.fbi.gov/about-us/lab/codis/swgdam-interpretation-guidelines

That's a lot of guidelines just for regular testing and interpretation and it makes it clear it does not apply to LCN (they would be more stringent is the implication).

This link might be more relevant than the interpretation guidelines

http://www.cstl.nist.gov/strbase/QAS/Final-FBI-Director-Forensic-Standards.pdf

Thank you Rose.

I guess a list of specifics would be nice: what the protocols/procedures to be followed concerning LCN DNA testing and what it was specifically that Stefanoni did not follow.

 

[Chris_Halkides]

does it clear the bar

I don't see how you get this from them saying a non-reproducible result should be disclosed. It seems pretty clear from the article that they include results that retests fail to reproduce.

From my earlier comment:
"and the scientists generally accept as true those STR signals that show up in both runs."

Disclosing a result is one thing. Clearing the bar for being admissible in a court is not the same thing at all. Offhand, I might be tempted to suggest that a partial profile of three or four alleles should be treated in a similar manner.

 

[Justinian2]

It's a good point. The problem comes with lying to police. You get caught in a lie and then you have to justify the lie with another and so on. Raffaele did the correct thing to admit his was a load of rubbish and also to remain silent after that, in my opinion. It was already "proven" that either his 5 November statement was wrong or the statements before and the one after are wrong.

I believe Amanda tried to explain her false accusation overly much. Again, same issues once the statement contradicting her earlier statements were made. Neither could change the fact that they were now shown to have given false information. Those judging the veracity of their statements now have a choice of believing either or, the cat is already out of the bag and the moving finger has already writ. The hard part now becomes of why they made false statements and which ones are the false ones.

It isn't really news that both liars and honest people are be suceptable to brainwashing, statements of the police, statements of other witnesses, statements of the media, statements of the court, the instructions of their lawyers, the influence of drugs, dreams, and being tired.

Makes a mess when trying to straighten it all out.

Also, both liars and honest people change their statements.

The differentiation is that honest people change their statements to INCLUDE more detail. Dishonest people change their statements because they realize that their justification doesn't make sense.

Amanda usually changed her statements to include more detail. That is a characteristic of honesty.
My theory is that what I have just discussed and other concepts to determine when someone is lying (such as body language), profiling, and LCN DNA should NEVER be used in a court of law. These techniques should be used only by dectective teams to narrow down the search for the proper suspect or evidence.

 

[shuttlt]

I would take this to clearly mean that the results of the tests (the numerical and graphical data) are provided to investigators, but that no calculation regarding a match to the reference sample is possible. In other words, it's useless as potential evidence against the person whose reference sample is being compared.

That is not how the word 'calculation' is used in the rest of the CPS page.

Only those DNA components that are seen twice are included in any calculation, to show that the result is reproducible.
If the result is not reproducible or is a complex mixture then it is disclosed but no calculation is carried out.
<snip.
The statistical calculation, as for all DNA results, compares the likelihood of obtaining that result given the two competing accounts of the defence and prosecution about the person or people who may have left the DNA.

The third mention of "calculation" is clearly talking about calculating the odds of the prosecution and defence explanations of the DNA findings. The previous uses don't in and of themselves make it clear what they mean by calculation.

 

[Chris_Halkides]

Do you have any actual evidence that it happened in this case.

What would constitute evidence in your mind?

 

[colonelhall]

colonelhall,

I think Amanda said that she smoked one joint, but I could be wrong. Raffaele also said something along the lines that he was in a holiday mood, meaning (I suppose) not taking careful note of the time, for example.

So you choose to ignore his written comments. I see.

 

[Chris C]

Rose: "Wasn't the previous statement he is referring to as rubbish the one where he said Amanda was out?"

I don't think so.

As far as I can see, his final statement is the one where he said:

"Amanda and I went into town at around 6pm, but I don’t remember what we did. We stayed there until around 8.30 or 9pm.

At 9pm I went home alone and Amanda said that she was going to Le Chic because she wanted to meet some friends. We said goodbye. I went home, I rolled myself a spliff and made some dinner.”

Which day are you talking about.

 

[shuttlt]

Disclosing a result is one thing. Clearing the bar for being admissible in a court is not the same thing at all.

I agree. Admissibility is a separate question. The lab tech wearing a face mask doesn't seem to me to be an admissibility issue though either.

 

[Dan O.]

Do you have a list of the precautions that the Rome lab did not take? It appears the lab was applying for certification to do LCN DNA testing during the case. A list of what procedures a lab should do during LCN DNA testing and what Stefanoni did not do would be helpful. I have tried to search for ISO 9001 and 17025 but they are publications which need to be purchased (from what I have searched so far).

The precautions have been listed in this thread numerous times. Little things like separation of work spaces, positive airflow, UV lighting, chemical cleaning. But one of the biggest elements is monitoring to know the actual frequency of contamination so you know that the lab design is working and can apply a mathematical probability to the validity of the labs results.

Because of Stefanoni's claim that her lab had never had a case of contamination in 7 years, I questioned whether the lab even ran control samples when conducting DNA tests. I asked very early in this thread if the lab was certified to perform DNA testing and was assured repeatedly that it was. I asked if there was any proof of this certification and pointed to the only certifying agency in Italy and their list of certified labs and asked where the Rome lab was on that list. It turns out that Stefanoni's lab could not be on this list or any other list of certified labs because they only started the process of getting certifies after this case came up. Why did somebody decide that it was necessary to deliberately lie about the labs certification?

We know that the lab now has a 9001 certification that the prosecutor so triumphantly presented to the court 2 years after the lab did the testing for the case. But that is only for management and procedural issues and doesn't address DNA testing. It's 17025 that has sub-certifications for DNA processing including LCN. The Rome lab cannot begin to apply of LCN certification because they don't have a proper facility.


Why don't you start here to read up on ISO 9001: http://www.iso.org/iso/iso_catalogu...tandards/quality_management/iso_9001_2008.htm to get a feel for what ISO 9001 is all about.



ETA: The somebody is no longer a member of this board so I expect to see fewer deliberate lies in the future.

 

[LondonJohn]

That is not how the word 'calculation' is used in the rest of the CPS page.


The third mention of "calculation" is clearly talking about calculating the odds of the prosecution and defence explanations of the DNA findings. The previous uses don't in and of themselves make it clear what they mean by calculation.

But this just appears to be another way of saying what I was saying. If contamination has been (to all intents) ruled out, the situation is almost always that the prosecution says the defendant was at the crime scene and the defence says he wasn't (by definition - otherwise, proving that it's the defendant's DNA would be of no use to the prosecution).

The calculation is therefore this: what is the probability that the DNA profile found from the tested sample is from another unidentified member of the population (the defence's position), versus it being that of the defendant? Clearly, the more complete the match, the more remote would be the calculated possibility of the DNA coming from another person.

 

[platonov]

What would constitute evidence in your mind?

So on being asked to back up your claim...

I think that one reason that Amanda and Raffaele displayed confusion in the days after they were incarcerated was that their interrogators claimed that they had evidence to the effect that their memories were faulty.

you had no evidence other than another spin on The Cartwheel of Analogy and this having been disqualified are responding with requests for definitions.

No it is then.