More CSI stuff. "Study shoots holes in bullet analyses by FBI," is the headline this morning to this story in the Chicago Tribune. The lead:
A group of experts Tuesday called for an overhaul of the FBI's process of analyzing the chemical makeup of bullets to try to link them to specific crimes, citing, in part, broad overstatements examiners have made in criminal cases that had no scientific basis.The NY Times today has this story. A quote:The National Research Council study said the 4-decade-old technique of analyzing the chemical content of a bullet's lead was still permissible for use in court, even before recommended changes are implemented, though such evidence at best should be considered circumstantial.
After a yearlong study, a scientific panel of the National Research Council found that the Federal Bureau of Investigation's scientific method for comparing bullets was generally sound but that its examiners had sometimes overstated its reliability in court testimony and had played down the likelihood of false matches.A 4-page summary of the report may be accessed here. The entire nearly 200-page report may be accessed here.The F.B.I. asked the research council to conduct the independent study after a metallurgist who had worked at the bureau publicly challenged the process used for comparing and "matching" bullet fragments at crime scenes to bullets found in the possession of suspects. The study cost the bureau $300,000.
It found that although expert witnesses for the bureau had sometimes asserted in court that chemical analysis could trace a bullet fragment to a factory batch or even a single box of bullets, the data did not appear to support such a specific conclusion. In fact, millions of bullets could have similar or nearly indistinguishable characteristics, panel members said.
[Update] Here is an update to the entry Monday (scroll down) on digital evidence, an AP story titled "Digital evidence raises doubts" at CNN.com.
Posted by Marcia Oddi at February 11, 2004 08:05 AM