The Lack of Statistics and Distortion Can Weaken Fingerprint Accuracy
Cherry, Michael, Imwinkelried, Edward J., Schenk, Manfred, Judicature
Before the advent Of DNA testing, fingerprint examination was considered "the gold standard" of the forensic disciplines. However, in its 2009 report the National Academy of Sciences raised serious questions about the status of fingerprint evidence.1 Together with that report, our own research has identified several areas of concern.2
From the fingerprint examiner's perspective, the ideal situation is a case in which the examiner can compare a card with 10 complete, clear prints against another card with 10 equally complete, equally clear prints. Examiners sometimes encounter virtually ideal situations in the immigration context. In that context, they often have the luxury of two 10 print cards to compare. However, even then problems arise. In particular, distortion can occur. The person placing his or her fingers on a fingerprint card or an AFIS scanner may not apply uniform pressure.
Although distortion that is often invisible to the naked eye can occur in the creation of the known print, distortion in latents is an even larger problem. The perpetrator often applies pressure unevenly when he or she leaves a latent at the crime scene. If the latent appears on an object found at the scene, the perpetrator may have twisted his or her finger when gripping the object. Distorted areas in both types of prints need to be identified and excluded, since their presence can create misleading artifacts.
To make matters worse, unlike an ideal situation, in forensic casework the examiner has only a partial latent. The examiner often has a very small area, representing approximately 20 percent of each fingerprint.
Fortunately, these problems can be addressed. Currently no statistical measurements are kept on fingerprints either by die examiners or by the Automated Fingerprint Information Systems (AFIS) diat they use. …