Friday, September 02, 2005

CSI Undone

In the criminal justice system, much has been said about the impact of popular crime dramas have had in shaping laypeople's perceptions of forensic science and expectations of investigations.

The "CSI Effect" Lately, several prosecutors have commented that juries have been more willing to acquit criminal defendants simply because there is no forensic evidence that tied the defendant to the crime, even in cases where forensic evidence would normally not come into play.

With the increased attention on forensic science, defense attorneys have been make significant headway in revealing long-used forensic science techniques as "junk science." With that, common forensic techniques like bite mark comparison, microscopic hair comparisons, and bullet lead batch comparison tests have come under more scientific scrutiny and revealed to be more questionable science than originally accepted.

Today, the FBI announced that they will no longer conduct bullet lead batch comparison tests as it is simply proven to be too unreliable. The test had been routinely done by the FBI since 1966 and has served as evidence in countless criminal convictions.

Essentially, the flawed theory behind the test was that by testing the amount of common trace elements and compare the compositions of say a bullet committed in a murder to unfired bullets in the possession of the accused. The FBI analyst would say whether the composition of the two different bullets of those trace elements were similar, thus implying that they were likely from the same batch made from the same manufacturer. You can see where I'm going with this.

However, study of the technique had revealed one critical flaw. Not every bullet in a manufacturer's package came from the same batch. Furthermore, studies also showed that bullets supposedly made in the same batch did not have the same composition of the trace elements.

Although only 20% of the FBI lead batch tests were admitted as evidence, the impact of a forensic science technique proven to be unsound can not be understated. Over 100 cases since the 1980s had lead batch testing evidence introduced as sound forensic science bearing the seal of approval from the FBI. And in hindsight, it's hard to quantify how much impact that forensic test alone had on the jury's decision.

Expect to see further forensic science techniques to come under increasing scrutiny. With that, several long-held assumptions about the results of some tests to be proven to be unsupported by science. Indeed, given the lack of a uniform standard of doing comparison analysis, even the use of fingerprint analysis has become questioned. And in some courts, expert handwriting comparison evidence is inadmissible as unreliable evidence.

No comments: