Continuation Part 2 - Discussion of the Amanda Knox case

[shuttlt]

Look, in order to find a meaningful counter-example, we have to find cases of previously well-balanced people engaging in a murderous attack for no reason and with no warning - as Amanda Knox and Raffaele Sollecito are accused of doing. It's a tall order - which is the reason why the Perugia police, prosecutors and pro-guilt bloggers have gone to such lengths to paint a false picture of the 2 of them.

Not for no reason. If AK and RS are the killers, we just can't know what the reason is.

 

[LondonJohn]

A short discussion about the efficacy of offender profiling:

http://www.apa.org/monitor/julaug04/profiling.aspx

The results of the studies seem to suggest that professional profilers tend to get closer to an accurate profile than regular members, and that they are also better than a randomised profile. But, as the study notes, there are very often mistakes of varying proportions.

As I've said previously, the current thinking on offender profiling is that it is very far from invulnerable as far as accuracy is concerned. But it's better than nothing, and can help law enforcement agents narrow down their universe of suspects. Occasionally, of course, inaccurate profiling can hinder an investigation by sending police down the wrong trail - the hunt for the Washington sniper is a very good example of this. But instances such as this are fewer in number than those where a profile is useful in directing police towards a particular subset of the population.

 

[shuttlt]

I find it ironic you bring that up in a case where the head of the Scientific Polizia claimed they could tell Amanda Knox was guilty simply from observing her behavior for a few days and noting a few items that sound ridiculous as indications of a murderer in retrospect; to the extent that they'd mount her head on the wall a year before she'd been charged and they didn't find a trace of her in that room.

I'm not trying to build a case for guilt.

It does not invalidate what John Douglas and the FBI do however. What they use criminal profiling for is to try to use psychology and probability based on crime statistics to get an indication of what to look for in their suspect. Then they use tried and true forensic methods to evaluate physical evidence to confirm or deny whether they are on the right track or have a case against a specific person. Being as often initial clues are sparse, like for the unabomber, the possibility of errors--especially gross errors is high--and the good ones know that.

The research I've been able to find indicate the behavioural patterns the profilers claim to be using don't exist because there are too many variables.

What Giobbi did is laughable, what John Douglas does is use every tool at his disposal to attempt to evaluate a crime and the criminal who committed it. He's damned good at it too, because he is rational enough to be able to falsify hypothesis and change profiles as new information is provided and old information invalidated.

Criminal profiling is a completely unvalidated technique. The articles I've been able to find indicate it fails the same tests Randi uses on astrology

Here's a more balanced article on the field. Incidentally I don't think simply looking at Amanda Knox or her history one can tell whether she committed the crime, that's what you need evidence for. You'll note in the short article that John Douglas also mentions the forensics, you also probably remember Steve Moore does the same thing, they evaluate all the information at their disposal.

What I find odd is those who look at Amanda Knox and her history and see a monster, that's just weird.

From your quote:

He has studied, for example, whether police officers perceive the same profile to be more accurate and useful when they believe it was written by a professional profiler rather than a layperson.

Kocsis agrees that the future of profiling lies in more empirically based research. He also believes, though, that just as some clinicians are better than others, there is also a skill element involved in profiling. Is profiling an art or a science? "Realistically, I think it is probably a bit of both," he says.

I don't see much evidence that profiling is more than hunches and confirmation bias. If you convince me it's a valid technique, I'll reconsider my dismissal of the claim that they could tell the killer was female.

 

[LondonJohn]

Not for no reason. If AK and RS are the killers, we just can't know what the reason is.

This is true. But reasons (and, by extension, motive) are usually fairly easy to find. Nearly all murders occur for reasons of jealousy, power/control, money, pride, self-preservation or lust (or a combination of any of those), or are the product of mental illness on the part of the perpetrator.

In this case, there doesn't appear to be any plausible reason why Knox, Sollecito and Guede would have ganged together to murder Meredith. If one goes down the list of possible reasons, not one of them seems like a credible reason for these three to participate in a group killing.

On the other hand, it's extremely easy to posit a rationale for Guede alone to have killed Meredith: a potent combination of self-preservation, power and lust, which quickly took hold once Guede found himself alone in the cottage with Meredith. This was a mixture of an underlying sexual attraction for Meredith, coupled with a deeper-seated frustration in Guede's mind about his lack of success with women, all wrapped up in a confrontation based upon Guede being inside the cottage for no innocent reason, and the reasonable expectation that nobody else would come back to the cottage for some time.

 

[shuttlt]

In this case, there doesn't appear to be any plausible reason why Knox, Sollecito and Guede would have ganged together to murder Meredith. If one goes down the list of possible reasons, not one of them seems like a credible reason for these three to participate in a group killing.

When we're all finally convinced they're guilty, I'm sure we'll be able to agree on some post-hoc rationalization of why they did it. The second part will be easy once the first part has been achieved.

 

[shuttlt]

A short discussion about the efficacy of offender profiling:

http://www.apa.org/monitor/julaug04/profiling.aspx

In the murder case, however, they fared no better on average than the nonprofilers did.

I'd be interested to find the study though. The design looks iffy, but what do I know. The profilers wrote longer profiles and I'd be interested to know how they decided how good a fit each profile was.

 

[LondonJohn]

I'm not trying to build a case for guilt.


The research I've been able to find indicate the behavioural patterns the profilers claim to be using don't exist because there are too many variables.


Criminal profiling is a completely unvalidated technique. The articles I've been able to find indicate it fails the same tests Randi uses on astrology



From your quote:

I don't see much evidence that profiling is more than hunches and confirmation bias. If you convince me it's a valid technique, I'll reconsider my dismissal of the claim that they could tell the killer was female.

Firstly, I would say that it's very difficult to conduct proper research into the efficacy of offender profiling. Most such research is done on an ex-post basis, where a crime has a known culprit with a specific known profile, and subjects are asked to create a profile (based on the details of the crime) which is then matched against the known profile of the known culprit.

A better (and more realistic) study would examine those crimes where offender profiles were generated during the investigation, and comparing them with the actual profile of the actual culprit (assuming that the culprit was caught). I'm not aware of any such studies having been conducted.

The other important issue in judging the efficacy of profiling is the specificity (or otherwise) of the profile. As has often been noted, some profiles contain nebulous statements such as "the offender will have self-esteem issues" - which could probably apply to well over 95% of the population to some degree. But there are numerous instances of accurate profiles that have contained more specific statements regarding age, marital status, domestic situation, job, hobbies, level of intelligence etc.

Lastly, it's generally accepted that Mignini and Massei were using incorrect reasoning when they interpreted the covering of Meredith's body with the duvet as a trait of a female culprit. In fact, psychological research coupled with analysis of other murders suggests that covering the victim is in fact a trait of an immature culprit, who has probably just killed for the first time. In that respect, the act is attributable just as much to Guede as it is to Knox (or Sollecito).

 

[RWVBWL]

This is true. But in any case there's other evidence that Amanda and Meredith had some sort of friendship that went beyond mere acquaintance. They went together to the classical concert in late October at which Knox first met Sollecito*. And there is a documented series of friendly text messages between the two that was exchanged on Halloween night - the night before the murder.

Knox had told her family and friends that she and Meredith were friendly, and Meredith's English friends in Perugia stated that if Meredith wasn't with them, she hung out with Amanda. I suspect that Meredith's family might also have some evidence of the friendship between the two girls (emails, photos, phone calls), but that they are reluctant to present this evidence...

* although Meredith left half way through the concert, for reasons that have never been explained. I can only imagine the opprobrium that might have been heaped upon Knox if it had been she who had abandoned her friend half way through a concert, but there you go.

PS: following my bath, I am now enjoying a lovely meme of fresh mint tea

Hi LondonJohn,
Enjoy that tea!

I just wanted to take a moment from scarfin' down some insane carne asada, pollo, and carnitas tacos that we just grilled and drinkin' some very cold brewski's on this holiday weekend here in the U.S.A., to chime in with another mostly forgotten tidbit.

On the night before Halloween, Oct. 30, 2007, Meredith, a few of her British friends and Amanda, the small, quiet Amerian girl, drably dressed*, and Raff, her much taller, plain looking Italian boyfriend** enjoyed some wine at a local Perugian joint called La Tana Dell'Orso, this according to according to co-owner Lucy Rigby.

Make of that what you will,
I'm headin' back to taco's and beers,
Peace, RW

* - Page 97, Murder in Italy, C. Demsey.
** - Page 98, Murder in Italy, C. Dempsey

 

[LondonJohn]

When we're all finally convinced they're guilty, I'm sure we'll be able to agree on some post-hoc rationalization of why they did it. The second part will be easy once the first part has been achieved.

An interesting position.....

Anyhow, this whole discussion serves to illustrate quite well why motive is not a requisite part of the burden of proof. Sometimes (but rarely) people do kill for no discernible reason. And sometimes when someone is murdered, the person with seemingly the strongest reason to kill is not the actual culprit.

All that needs to be proven is intent and opportunity. Both of these elements can be proven through evidence and witness testimony. If a court can be convinced that intent and opportunity have both been proven beyond a reasonable doubt, then a guilty verdict should ensue without any need to examine motive. Of course, motive is often introduced in a criminal trial - but it's usually introduced to provide evidence of intent and/or opportunity.

 

[shuttlt]

Firstly, I would say that it's very difficult to conduct proper research into the efficacy of offender profiling. Most such research is done on an ex-post basis, where a crime has a known culprit with a specific known profile, and subjects are asked to create a profile (based on the details of the crime) which is then matched against the known profile of the known culprit.

A better (and more realistic) study would examine those crimes where offender profiles were generated during the investigation, and comparing them with the actual profile of the actual culprit (assuming that the culprit was caught). I'm not aware of any such studies having been conducted.

The problem there is that the quality of the match is subjective. Unless you can find some way around that such a study is no good.

The other important issue in judging the efficacy of profiling is the specificity (or otherwise) of the profile. As has often been noted, some profiles contain nebulous statements such as "the offender will have self-esteem issues" - which could probably apply to well over 95% of the population to some degree. But there are numerous instances of accurate profiles that have contained more specific statements regarding age, marital status, domestic situation, job, hobbies, level of intelligence etc.

But how many wrong details were there as well?

 

[Kaosium]

That's the point though, surely. Offender profiling should only be used as a tool to help direct the investigation. The profile should never be used as any sort of evidence in itself.

What a coincidence! Within minutes we both replied with very similar rebuttals. Now, this could be for various different reasons, which do you think some people might think that would be?

1. We're right, that's more or less what the field entails and how it is applied by competent investigators and just happen to be online at the same time and interested in the same post.

2. We're conspiring together behind the scenes because we're part of the Gogerty-Marriott/FOA conspiracy and sinister cabal of innocentisti 'piling on' poor Shuttlt because we wish to drive him from the board! There is no cabal!

3. I read your article, quickly googled up a confirmation link and I really can type that fast without using the word 'excellent' three times within the space of about two sentences like I did in a recent cringeworthy reply to a post of yours.

4. It's been a ruse the whole time, we're the same person! LondonJohn is not from London, or Texas, but Wisconsin! Thus 'we' typed up two separate articles and posted them, but left a 'clue' to this perfidy by doing so within three minutes. Now we are attempting to desperately cover our mistake with an 'admission' which only proves we must be lying, we must be guilty! That's why Kaosium is constantly making fun of the 'investigative abilities' of those who thought the above, it was all part of the sinister plot!

Now, which conclusion do you suppose those 'intrepid investigators' from the Bunny and Kitten Detective Agency will come to?

“Oh what a tangled web we weave, When first we practice to deceive”

 

[shuttlt]

What a coincidence! Within minutes we both replied with very similar rebuttals. Now, this could be for various different reasons, which do you think some people might think that would be?

What are you rebutting? People seem convinced I'm trying to use this whole criminal profiling thing to make an argument relating to the case. I've already said that I am purely responding to the, in my view dubious, authority of the criminal profiler in the quote. I feel like Amanda... why won't anybody believe me? At this point I'd gladly agree to anything to get out of this loop...

 

[RoseMontague]

What a coincidence! Within minutes we both replied with very similar rebuttals. Now, this could be for various different reasons, which do you think some people might think that would be?

1. We're right, that's more or less what the field entails and how it is applied by competent investigators and just happen to be online at the same time and interested in the same post.

2. We're conspiring together behind the scenes because we're part of the Gogerty-Marriott/FOA conspiracy and sinister cabal of innocentisti 'piling on' poor Shuttlt because we wish to drive him from the board! There is no cabal!

3. I read your article, quickly googled up a confirmation link and I really can type that fast without using the word 'excellent' three times within the space of about two sentences like I did in a recent cringeworthy reply to a post of yours.

4. It's been a ruse the whole time we're the same person. LondonJohn is not from London, or Texas, but Wisconsin! Thus 'we' typed up two separate articles and posted them, but left a 'clue' to this perfidy by doing so within three minutes. Now we are attempting to desperately cover our mistake with an 'admission' which only proves we must be lying, we must be guilty! That's why Kaosium is constantly making fun of the 'investigative abilities' of those who thought the above, it was all part of the sinister plot!

Now, which conclusion do you suppose those 'intrepid investigators' from the Bunny and Kitten Detective Agency will come to?

“Oh what a tangled web we weave, When first we practice to deceive”

Rose agrees with that.

 

[RoseMontague]

What are you rebutting? People seem convinced I'm trying to use this whole criminal profiling thing to make an argument relating to the case. I've already said that I am purely responding to the, in my view dubious, authority of the criminal profiler in the quote. I feel like Amanda... why won't anybody believe me? At this point I'd gladly agree to anything to get out of this loop...

They are saying a professional profiler uses the knowledge simply as a tool in the overall investigation. The investigative experience this person has and his credentials make his standing as an expert legit, imo.

Contrast this with the Statement Analysis Dude, in whose case I would agree more with your premise.

 

[Kaosium]

I'm not trying to build a case for guilt.

I was contrasting the type of 'profiling' done by some, which might be what your initial article was referring to, with the profiling and especially the use of it done by more advanced institutions.

The research I've been able to find indicate the behavioural patterns the profilers claim to be using don't exist because there are too many variables.
Criminal profiling is a completely unvalidated technique. The articles I've been able to find indicate it fails the same tests Randi uses on astrology.

That's because they're trying to measure the wrong thing, the skeptics article you cited didn't understand how profiling is employed--by the competent--neither did the Liverpool paper it linked. These aren't definites. When they say 'the killer is 30-40ish, left-handed, probably has advanced schooling and might have had a troubled childhood' that's simply what the available information would suggest. Of course it can't 'tell' those things with certainty.

From your quote:

I don't see much evidence that profiling is more than hunches and confirmation bias. If you convince me it's a valid technique, I'll reconsider my dismissal of the claim that they could tell the killer was female.

That was silly to begin with. They can't tell that sort of thing, just what seems to be more likely based on past cases and an understanding of the psychology of murderers. As for that one, as I recall no one but the ones who said it had ever heard of it.

 

[shuttlt]

They are saying a professional profiler uses the knowledge simply as a tool in the overall investigation. The investigative experience this person has and his credentials make his standing as an expert legit, imo.

I think his standing is undermined if his primary expertise is in what may be a pseudoscience. Sure, he's seen a lot of cases and I'm sure he's knowledgable and so forth.

Contrast this with the Statement Analysis Dude, in whose case I would agree more with your premise.

If statement analysis dude had worked on a lot of cases, would it make a difference?

 

[Kaosium]

What are you rebutting? People seem convinced I'm trying to use this whole criminal profiling thing to make an argument relating to the case. I've already said that I am purely responding to the, in my view dubious, authority of the criminal profiler in the quote. I feel like Amanda... why won't anybody believe me? At this point I'd gladly agree to anything to get out of this loop...

It had nothing to do with you, Shuttlt. It was a bit of whimsy employing one of the oldest jokes on the net, and past behavior of other posters in the debate here and elsewhere.

 

[shuttlt]

That's because they're trying to measure the wrong thing, the skeptics article you cited didn't understand how profiling is employed--by the competent--neither did the Liverpool paper it linked. These aren't definites. When they say 'the killer is 30-40ish, left-handed, probably has advanced schooling and might have had a troubled childhood' that's simply what the available information would suggest. Of course it can't 'tell' those things with certainty.

Where are you getting this "definites" thing from. I posted this quote from the skeptic dictionary article above and I'll post it again:

the Liverpool group selected a hundred stranger rapes in the United Kingdom, classifying them according to twenty-eight variables, such as whether a disguise was worn, whether compliments were given, whether there was binding, gagging, or blindfolding, whether there was apologizing or the theft of personal property, and so on. They then looked at whether the patterns in the crimes corresponded to attributes of the criminals—like age, type of employment, ethnicity, level of education, marital status, number of prior convictions, type of prior convictions, and drug use. Were rapists who bind, gag, and blindfold more like one another than they were like rapists who, say, compliment and apologize? The answer is no—not even slightly.*

 

[RoseMontague]

I think his standing is undermined if his primary expertise is in what may be a pseudoscience. Sure, he's seen a lot of cases and I'm sure he's knowledgable and so forth.


If statement analysis dude had worked on a lot of cases, would it make a difference?

His credentials are an online course and he read the book (three times). How does he compare with the FBI profiler Dude?

 

[RWVBWL]

What are you rebutting? People seem convinced I'm trying to use this whole criminal profiling thing to make an argument relating to the case. I've already said that I am purely responding to the, in my view dubious, authority of the criminal profiler in the quote. I feel like Amanda... why won't anybody believe me? At this point I'd gladly agree to anything to get out of this loop...

Hi Shuttlt,
Just sign here, here and here.
And we'll just put you in this cell for your safety...