Study tests fingerprint experts

A study published in Forensic Science International tested the reliability of fingerprint identification by having 156 American experts compare 100 pairs of images, not knowing which came from the same finger. Researchers marked which pairs were from different fingers and found the experts made false identifications in 0.2 percent of cases where the fingerprints were different.

Experts made few errors overall

When the fingerprints were actually identical, the experts incorrectly excluded the correct match in 4.2 percent of cases. This means the overall error rate for the group was relatively low, but the errors were not evenly distributed among all participants.

One expert made most errors

Out of 23 total errors in the study, thirteen were made by a single participant. This individual had limited experience, no formal certification, and minimal training, which suggests that the overall error rate may not reflect the average expert's performance.

Error rate depends on the expert

The study shows that the error rate of fingerprint identification is not a simple average. It includes many experts who rarely make mistakes and a small number who make frequent errors. This means that for legal professionals, the key concern should be who conducted the comparison and under what conditions, rather than just the overall error rate.

Review by a colleague helps prevent errors

The study also found that no error was repeated when a second expert reviewed the same pair of images. This indicates that having a colleague review the work can help catch mistakes, suggesting that peer review is an effective safeguard in forensic fingerprint analysis.