A method for the statistical interpretation of friction ridge skin impression evidence: Method development and validation

Defence against the modern arts: the curse of statistics—Part II: ‘Score-based likelihood ratios’

“…We warn the reader that other models involving scores have been proposed (Armstrong et al, 2017;Swofford et al, 2018;Ausdemore et al, 2019;Hendricks et al, 2019) but are not considered to be score-based likelihood ratios. These models do not rely on the ratio of the likelihoods of the score in two sampling distributions.…”

“…These models do not rely on the ratio of the likelihoods of the score in two sampling distributions. For example, the right panel of Figure 2 shows the concept underlying a model called FRStat (Swofford et al, 2018) 10 , which relies on the ratio of two tail probabilities bounded by the score between a trace and a control object. Comparison between two different concepts for the use of summary statistics/kernel functions to provide some information on the probative value of fingerprint evidence.…”

The Need for Objective Measures in Forensic Evidence

“…By focusing on such steps, examiners were able to achieve greater consistency with each other. The exercise also prompted researchers to develop objective metrics to quantify the likelihood of the similarity observed between two prints 5,6 …”

“…As an example, Henry Swofford and colleagues created an algorithm to calculate an objective numerical measure of similarity between two prints, FRstat, based on the geometric configurations of the minutiae in the two prints being compared; see Figure 6 . Values of FRstat can be compared for different pairs of prints; for example, between a questioned latent print and potential exemplar prints.…”

“…Recreation of Figure in Swofford et al ., a conceptual illustration of the overlay and pairing of features from two fingerprint impressions (green and black) 6 . Geometric features (weighted angles and distances) are used in calculating FRstat.…”