Testing the Accuracy and Reliability of Palmar Friction Ridge Comparisons: A Black Box Study

As submitted by the proposer:

The latent print comparison field has long been at the forefront of forensic science reform, conducting research to meet the challenges of critics, and producing error rate studies that establish the foundational validity of the comparison process. However, these studies have largely focused on fingerprints only, and palm impressions have been hitherto neglected.

RTI International and the University of Lausanne have partnered to conduct a black box error rate study using exclusively palm print impressions. The researchers seek to begin establishment of the foundational validity of palmar friction ridge comparisons by empirically determining estimated error rates. The researchers will also investigate the relative strengths and weaknesses of both frequentist and Bayesian approaches to the presentation of error rate information in the courtroom, and provide recommendations that will assist examiners with their testimony about all friction ridge comparisons -- not only palms.

This project will take place over two calendar years, and will involve partnerships with several forensic laboratories and the participation of approximately 150 working latent print examiners. More than 500 known ground truth palmar impressions of diverse size, quality, substrate, and development techniques will be collected from our partner laboratories to make up the test sample pool, and each latent print examiner study participant
will complete 75 trials, resulting in an estimated 11,000+ presentations of palmar images.

Trials will include both same-source and challenging different-source trials to allow both false-positive and false-negative estimates to be made. Error rates will be calculated for all trials in aggregate, but also stratified by size and quality, which will allow examiners to select the error rate that most closely corresponds to the size or quality of the impression(s) in the case when presenting testimony.

Results will be disseminated to the forensic and legal communities via presentations at professional meetings, an article in a peer-reviewed scientific journal, and the presentation of a webinar. Additionally, the researchers will develop a web-based application specifically designed for computing confidence intervals and credible intervals. This application will demonstrate the benefits and limitations of both frequentist and Bayesian approaches in the context of error rates in friction ridge comparisons; and will be freely available for public use. Finally, a primer on error rates in friction ridge comparisons will be developed and distributed to the court system to educate jurors who will hear friction ridge testimony.

Note: This project contains a research and/or development component, as defined in applicable law, and complies with Part 200 Uniform Requirements - 2 CFR 200.210(a)(14).

ca/ncf