Nikita Nawani
The goal of this article is to assess the generalizability of the results of published studies on the accuracy and reliability of fingerprint comparison to fingerprint laboratory casework and/or to provide information on the error rate of the Analysis-Comparison-Evaluation (ACE) method. We examine the 13 previously published investigations on the accuracy and reliability of fingerprint comparisons. Since criminal courts began accepting forensic fingerprint evidence approximately 120 years ago, these papers represent the whole body of experimental research published on the reliability of fingerprint comparisons. We begin with the two recent, sizable investigations by Ulery, Hicklin, Buscaglia, and Roberts in order to address the issues raised in the National Academy of Sciences report and to provide estimates of the accuracy and reliability of fingerprint comparisons. The other eleven experiments are reviewed and evaluated after the two Ulery et al. studies, taking into account any particular issues with each one. The 13 experiments are then assessed for issues that are present in all or the majority of them, particularly with regard to the applicability of the findings to laboratory case studies. The tests did not require examiners to utilize the ACE approach, nor was that method established, regulated, or evaluated in these experiments. Neither did the experiments use fingerprint test items that were known to be similar in type and particularly in difficulty to those found in casework. We conclude that new experiments modeled on these existing experiments cannot provide the fingerprint profession or the courts with information about casework accuracy and errors until there has been significant progress in defining and measuring the difficulty of fingerprint test materials and until the steps to be followed in the ACE method have been defined and measurable.