Maybe the following will be of some help. Some problems are that the specificity depends on exactly what testing is done, and even on the circumstances of the crime (e.g., if the victim and the defendant are related), not to mention errors in handling (contamination or switching of samples, for example).
http://www.councilforresponsiblegenetics.org/pageDocuments/H4T5EOYUZI.pdf
http://projects.nfstc.org/workshops/resources/articles/How%20the%20Probability%20of%20a%20False%20Positive%20Affects%20the.pdf
http://dna-view.com/profile.htm
https://www.newscientist.com/article/mg20727743-300-how-dna-evidence-creates-victims-of-chance/
A more general consideration of the legal ramifications of Bayesian computations at trial:
https://www.eecs.qmul.ac.uk/~norman/papers/likelihood_ratio.pdf
>>Kathy
------------------------------
Katherine Godfrey
------------------------------
Original Message:
Sent: 09-17-2015 02:17
From: Ulderico Santarelli
Subject: Forensic use of DNA test
here in Italy a major crime is now at the attention of the TV audience. While somebody says that DNA test is so accurate that it wipes away any doubt of innocence, somebody else boldly criticizes the seemingly blind faith in DNA real capacity of supporting a "certain" guilty evidence. I went through some literature and found nothing about these 4 basic points
1. how much is the scale of the test valid?
2. how much is it reliable?
3. what about the sensibility of the test?
4. and, more important, specificity of the test?
I find that when you are going to send somebody to a life sentence, test specificity is a basic information. In words, how many other citizens would show the same "correspondence level" between two DNA samples? It should be none, in the sense that the claimed correspondence should find just the guilty guy to correspond at that level.
Because I ignore completely the biology theory that underpins the DNA test, I ask anybody having expertise on the subject to provide me with an answer about DNA test specificity. I assume that the scale is valid and also reliable and that sensibility is OK.
Thanks.
Ulderico Santarelli.