For all of the problems with the North Carolina State Bureau of Investigation crime lab, we have seen only a few comments about the accuracy of the lab’s work. We have talked about evidence withheld from defense counsel and reports altered to favor the prosecution; we have talked about a lack of oversight of lab personnel and the need for additional and more consistent training.
But we have not talked much about the reliability of the tests or whether the tests are producing questionable evidence. Does some of the fault lie in the science, not in the people?
The answer is fairly typical of attorneys, either defense or prosecution: Yes and no. Forensic science is actually both science and art. Either way, it is far from foolproof.
According to fingerprints.com, fingerprints have been used to identify individuals for thousands of years, but the first documented fingerprints date back to 1858. Techniques have changed as technology has become more sophisticated, and fingerprints are almost universally accepted as evidence in criminal justice systems, especially ours.
The truth is that fingerprint analysis is fairly subjective. Analysts compare the shape, ridge count, thickness, scars and creases — the “points” — of the suspect’s print to a print from the crime scene or from a database. The analyst reviews the measurements of the possible matches and comes up with a probability that the suspect’s fingerprint matches the one from the crime scene or the database.
The analyst’s perception is key here, and researchers have found that perceptions differ enough that the results can vary markedly between analysts. In another state, the crime lab’s review of a sampling of fingerprint examinations turned up irregularities in almost half of them. The state ended up hiring consultants to review all of the fingerprint examinations on file and to handle the backlog — more than 10,000 records in all.
We’ll continue this in our next post.
Source: Washington Post, “How accurate is forensic analysis?” April 16, 2012