There have been several instances recently of people convicted of crimes they did not commit and unjustly sentenced to prison for years -- based on faulty forensic science. Critics say that the practice of matching hair samples of victims with the accused is particularly unreliable.

• Hair experts have never been able to agree on consistent criteria for associating a suspect with an unknown hair -- nor have they produced persuasive data showing the probative value of such association, critics charge.

• In the early 1970s, the federal government sponsored a proficiency-testing program for 240 crime laboratories -- and discovered that the labs rated poorly on this exercise.

• Among the first 74 prisoners exonerated by DNA testing over the past decade, 26 were implicated, in part, by hair analysis.

• In the wake of revelations that some crime-lab analysts around the country have given fraudulent testimony against innocent parties in criminal trials, the profession is under increased scrutiny.

Some judicial observers recommend that forensic labs be run as independent agencies, rather than as adjuncts of police departments. Put on an independent basis, they could serve as fact finders for both the prosecution and the defense.

There is a model for improvement. The 1988 Clinical Laboratory Improvement Act provided accountability for laboratories that perform medical tests. Labs which make mistakes in health tests can lose their accreditation.

In forensic laboratories, by contrast, few are held accountable for bad practice or botched results.

Source: Barry Scheck and Peter Neufeld, "Junk Science, Junk Evidence," New York Times, May 11, 2001.

For NY Times text http://www.nytimes.com/2001/05/11/opinion
/11SCHE.html