 |
 |
2850 Presidential Dr.
Suite 160
Fairborn, OH 45324
Phone: (937) 426-9270
Fax: (937) 426-9271
help@bioforensics.com
|
|
|
 |
The DNA Advisory Board's approach: report both RMP and DMP (Dr. Michael Raymer, Wright State University)
Within a Bayesian context, the NRC II Report's LR and
Balding and Donnelly's (1996) LR could be interpreted to yield
a coherent evaluation of the evidence. Unfortunately, Bayesian
logic has not been considered by the U.S. criminal legal system
for DNA analysis. Clearly, what is required is a formulation
of the LR that transparently conveys its import without resorting
to Bayesian statistics. In this setting, the treatment of the
database search recommended by the NRC II Report can be conservative,
but only for the unlikely scenario of a very large N is
it very conservative. Apparently the treatment of the database
search recommended by Balding and Donnelly (1996) is not conservative
when the number of loci genotyped is small and remains so until
the number of loci becomes large enough to essentially ensure
uniqueness. To put it another way, without the Bayesian
framework, the Balding and Donnelly (1996) formulation is easily
misinterpreted in a fashion unfavorable to the suspect. Stockmarr's
(1999) formulation, which is a more formal exposition of what
originally appeared in the NRC II Report (1996), communicates
value of a database search far better, and it is always conservative.
Thus, we continue to endorse the recommendation of the NRC II
Report for the evaluation of DNA evidence from a database search.
Materials
DNA Advisory Board.
Statistical and population genetics issues affecting the evaluation of the frequency of occurrence of DNA
profiles calculated from pertinent population database(s). Forensic Science Communications. 2000;2(3).
|
|
