CLPEX Chat Board

This is cool for those interested in a probabilistic approach and qualified opinions. These Swiss experts are using a verbal scale and reported "moderate support" for the hypothesis that he died by radioactive poisoning (vs. that he did not). However, I have a problem that the media is not telling us the scale. The report clearly establishes on pages 67-69 the evidence in favor of death by poisoning versus not. In the end, they conclude "moderate support" that he died by radioactive poisoning. But what does "moderate" mean? Without the entire scale reported by the media, it could be easily confused. The researchers used a 6 point scale: slight support, moderate support, and strong support (for one proposition or the other). So moderate support is actually higher than one might first assume. In other words, there is good evidence to suggest this, but not conclusive.

[Maybe I missed it, but I also didn't see that the researchers reported the expected levels of Po210 in an average human being from that region. Or are we to assume it should have been zero? It seems necessary to establish that the levels in Arafat were significant].

Anyway, for those with an interest in probabilities and use of qualified opinions, here is a great report that clearly combines numerical data, uncertainties, and expert judgments into a probabilistic verbal scale.

Here is the link to the Washington Post Article.
http://www.washingtonpost.com/world/mid ... story.html

You can find the full report at:
http://www.aljazeera.com/investigations ... 63780.html

g.

This is interesting, on a philosophical level

However

He was either poisoned or he was not.

The report seems like a complicated answer to a simple question

You've got to love Bill's uber practical sentiments on the issue, and to some extent I think his statement hits at the heart of the matter, namely causality. I'll keep it short so Steve will actually read it.

When we look at other statistical measures of forensic utility, they deal with brute facts about the world. The frequency of alleles in populations is illustrated by the Hardy-Weinberg Principle and over simply stated is the result of standard mechanisms of Mendelian inheritance.

The mechanisms of Polonium, namely Ionizing Radiation are a function of it's physical nature (unstable atoms). How those atoms interact with human biological systems is also well known through the study of Acute Radiation Syndrome and cancer research.

The point I'm getting at is that these are facts about the natural world that cause other facts in the world. Fingerprint Individualizations are not found in nature nor are they the result of some brute fact in the world, they're a function of cognition, which is not reducible to some lower level phenomena. For instance, we don't say ridge features in spatial/sequential agreement 'cause' individualizations, nor do we say that ridge features in spatial/sequential agreement 'cause' cognitive states that cause us to perceive individualizations. We say ridge features in spatial/sequential agreement 'count as' individualizations. The arrangement of ridge features do not determine individualizations in the same way that the arrangement of physical structures of a building determine its load bearing capacity. One is causally determined, the other is not.

What 'counts as' an individualization is a different kind of question than what causes an observer independent fact like allele frequency and alpha radiation. So, given that premise, let me ask a simple question. What brute fact about fingerprints makes them subject to qualified statements of probability in same way they are applicable to a DNA case or the Arafat Poisoning?