Fingerprints have been used in forensic identification for a hundred years. The generally accepted method of latent print identification is ACE-V. ACE-V stands for: analysis, comparison, evaluation, and verification.1 Is this process objective? Does the context in which the analyses are made affect outcomes?
Dr. Itiel Dror of the University of Southhampton set out to prove that context does affect their outcomes. According to Drors study, When Emotions Get the Better of Us: The Effect of Contextual Top-down Processing on Matching Fingerprints, emotional context and subliminal messages increased the likelihood of match judgments when fingerprints were ambiguous, thus concluding that contextual information actively biases the way gaps are filled.2 When fingerprints were clearly not a match, the same contextual information did not influence the participants to conclude that a match existed.
Dr. Drors next study, Contextual Information Renders Experts Vulnerable to Making Erroneous Identifications,3 used the background of the fingerprint found in the Madrid train bombing to create the context. On March 11, 2004, terrorists detonated a bomb on a number of trains in Madrid, killing approximately 191 people.4 After the bombing, Spanish officials found a blue plastic bag containing detonators in a van parked near a train station. Within one week of the bombing, Spanish officials sent digital images of partial latent fingerprints obtained from the bag to the FBI for analysis, and they were entered into the FBI computer.5
The FBI used the Integrated Automated Fingerprint Identification System to search and compare the unknown prints to millions of known prints stored in the database.6 The search generated a list of fifteen potential matches. FBI fingerprint examiners assessed the prints to determine whether the unknown print matched any of the known prints retrieved from the database.7
The FBI determined that the unknown print belonged to Brandon Mayfield, a 37-year old civil and immigration attorney practicing in Portland, Oregon.8 Mayfields fingerprints were on file from a 1984 burglary arrest as a teenager in Wichita, Kansas. On April 2, 2004, the FBI sent a letter to Spanish authorities informing them of the Mayfield identification. On April 13, the Forensic Science Division of the Spanish National Police responded that the purported identification was conclusively negative and that it couldnt confirm the FBIs match.9 While the FBI located fifteen points of similarity places where the particular ridge characteristics of a known and unknown print match the Spanish National Police found only eight points of similarity.10 Additionally, the Spanish government had no record of Brandon Mayfield ever traveling to Spain.11 The FBI presented the Spanish National Police with a three-page document detailing its findings, but the Spanish National Police refused to validate the FBI conclusions, stating there was no match. Nevertheless, Mayfield was detained on May 6. However, approximately three weeks later, the FBI announced that it had erred in the identification, even though the previous identification was found to be a 100 percent match with the fourth ranked print on the list.12 It turned out that the fingerprint actually belonged to Ouhnane Daoud, an Algerian.13
As a pretext, Dr. Drors newest study asked several examiners to analyze the prints from the Brandon Mayfield case. In reality, the examiners would analyze prints that each of them had personally declared a match in a previous case. The study was designed to provide empirical data on whether fingerprint experts are susceptible to extraneous contextual influences when working in normal routines and environments. Five fingerprint experts were selected from an international pool of volunteers from fingerprint bureaus, agencies, and laboratories all over the world. The experts were chosen based on two factors: 1) they had no familiarity with Mayfield’s fingerprint; and 2) identification matches they had made in the past were accessible to Dror.14
The participants agreed to be tested within the next 12 months. During this time they were presented with original pairs of fingerprints which had been previously identified as a match by them. The prints were presented to the experts during their normal course of work. Experts did not know which prints were test items and were not told that they were evaluating prints that they had already matched once before. Instead, they were told they were evaluating the fingerprints that were erroneously matched to Mayfield in the Madrid bombing case.
For the experimental print comparison, a co-worker approached each participant and told them: 1) to examine a set of prints, one from a latent print taken at a crime scene and the other a print obtained from a suspect, 2) that the prints were the same prints that were erroneously matched by the FBI as the Madrid bomber, giving an extraneous context that the prints did not match, 3) to decide if there was sufficient information available to determine whether the print was a match or a non-match, and if so, what was the conclusion, and 4) to disregard the context and background information and focus solely on the print in their evaluation. The participants were allowed to evaluate the fingerprints as they would during their normal course of work, with all the equipment they routinely used. They were provided with an unlimited amount of time to evaluate the prints.
As a result, one examiner held fast, despite the suggested context, and declared the prints a match, which was consistent with the previous identification he or she had originally made. The idea that the FBI had erroneously matched the prints was not an influence in that one case. Three examiners, however, contradicted their prior determinations that the prints had matched and now determined the prints did not match. Finally, one examiner who originally determined the prints to be a match now stated that there was not enough information to determine if the prints matched.
This study revealed that fingerprint examiners are vulnerable to irrelevant and misleading contextual influences.15 Even though the study was conducted with only five fingerprint examiners, the study was the first conducted in the real world conditions of the criminal justice system and four out of five in the study changed their prior identifications.16 Dror and his colleagues concluded that their findings of inconsistent identification decisions may reflect cognitive flaws and limitations in conducting objective and independent processing of the information, and calls for further research to examine these issues in greater depth.17
The National Institute of Justice is also funding many fingerprint studies. The first study, conducted by International Biometric Group, LLC, will examine the frequency and permanence of Level III characteristics of fingerprints. The project will also look at the tools that enable capture, processing, and statistical evaluation of these characteristics quality and strength. Indiana University is conducting another study using different quantitative approaches to fingerprint identification. In this study, relevant features that can be matched across prints will be extracted. The U.S. Department of Defense, Naval Systems Management Israeli National Police will conduct a third study. Factors that affect the life, durability, and recovery of fingerprints on hand-guns will be assessed and an attempt will be made to develop more successful methods for developing fingerprints on hand-guns, primarily using the cyanoacrylate method. A fourth study, conducted by the U.S. Department of Energy Oak Ridge Operations Office, will conduct research to gain a better understanding of fingerprint and fingerprint degradation chemistry, employ methods to enhance Raman-based latent-print visualization, and utilize the enhancement methods to modify the ChemImage system for field applications. Changes that occur in the spatial orientation of minutiae during maturity, focusing on children ages 2 18, will be examined in a fifth study, conducted by Ultra Scan Corporation. The sixth study, conducted by Research Foundation of SUNY, focuses on increasing understanding of the discriminative power of friction ridge patterns. The project will focus on assessing existing statistical models. The issue of quantity and quality of friction ridge data needed for matching will be studied and an empirical study of friction ridge patterns of identical and fraternal twins will be performed. 18
1 SWGFAST, Friction Ridge Examination Methodology for LatentT Print Examiners (Aug. 22, 2002) (available online at SWGFAST)
3 Itiel E. Dror, David Chalton & Ailsa E. Peron, Contextual Information Renders Experts Vulnerable to Making Erroneous Identifications, 156 FORENSIC SCI. INTL 74 (2006) (available online at Contextual Information Renders Experts Vulnerable to Making Erroneous Identifications)
4 Stephen Wax & Christopher Schatz, A Multitude of Errors: The Brandon Mayfield Case, Champion Magazine, Sept./Oct. 2004, at 6 (available online at A Multitude of Errors: The Brandon Mayfield Case)
5 David Heath & Hal Bernton, FBI Admits Fingerprint Error, Clearing Portland Attorney, Seattle Times, May 25, 2004, at A1 (available online at FBI Admits Fingerprint Error).
6 Statement on Brandon Mayfield Case, May 24, 2004 (available online at Statement on Brandon Mayfield Case).
7 Statement on Brandon Mayfield Case, supra note 7.
8Associated Press, FBI Apologizes to Lawyer Held in Madrid Bombings, MSNBC.COM, May 25, 2004 (available online at FBI Apologizes to Lawyer Held in Madrid Bombings).
9Wax & Schatz, supra note 5.
10 Jennifer L. Mnookin, The Achilles Heel of Fingerprints, Washington Post, May 29, 2004, at A27 (available online at The Achilles Heel of Fingerprints).
11 Hans Sherrer, Thats Not My Fingerprint, Your Honor, Justice:Denied, Summer 2004, at 11 (available online at Thats Not My Fingerprint, Your Honor ).
12 Wax & Schatz, supra note 5.
13 Mnookin, supra note 12.
14 Dror et al., supra note 4.
15 Dror et al., supra note 4, at 76.
17 Id. at 77.
18National Institute of Justice Awards.