Next time you see one of those television crime-scene investigators crack a case with high-tech analysis, better take it with a grain of salt. The National Academy of Sciences, the nation’s most prestigious scientific organization, has surveyed the field of forensic science and found it grossly deficient.
It’s not just that many forensic laboratories are poorly funded and staffed with “experts” who are poorly trained. The more fundamental problem, according to the study, is that there is little evidence of the accuracy and reliability of most forensic methods — especially those that rely on expert interpretation.
The most thoroughly validated technique is nuclear DNA analysis, which has a minuscule likelihood of error when done right. But other well-known methods that can supposedly identify a guilty person or link a weapon or other evidence to a particular crime have no rigorous scientific proof that they work consistently.
That goes for analyses of hair, bite marks, fibers, documents, tools, firearms, shoe impressions, tire tracks, handwriting and blood spatters, among others. The analyses can help focus an investigation but can seldom provide infallible evidence of guilt.
Even fingerprint analysis depends on a subjective judgment by experts as to how closely two prints match, a conclusion that can be biased by the examiner’s knowledge of the suspect or the case. Examiners have sometimes disagreed with their own past conclusions when viewing the same prints in a different context.
The academy’s panel makes sensible suggestions for improvement, such as certification of forensic professionals, accreditation of laboratories, uniform standards for analyzing evidence and independence of the laboratories from police and prosecutors who might bias judgments. In the long run, research is needed to determine the accuracy of forensic methods. For now, judges, lawyers and juries are on notice that high-tech forensic perfection is a television fantasy, not a courtroom reality.
