Every day, Fingerprint Experts in every latent office across the globe examine fingermarks that they determine to fall short of an identification. Whether it be due to distortion, the lack of clarity or quantity of information observed, or down to the training, knowledge and experience of the examiner, all these issues can result in this outcome.
How often do such prints occur? Are inconclusive determinations of 5 corresponding features have the same weight of 7-8 features in agreement? They’re often reported the same way if your agency uses the approach #2 level of reporting. What is their potential value for association? Would they actually impact case investigations or prosecutions in a useful way? Well a study published in Forensic Science International in February, 2020 sought to investigate these particular issues for answers.
Titled “Occurrence and associative value of non-identifiable fingermarks”, the authors collected latent fingermarks from casework of nine state and local jurisdictions within the USA. These fingermarks had fallen below the expert-determined threshold “of value for identification,” but had some clear Level 2 detail (i.e. minutiae) within an area of contiguous ridge flow. After reviewing all collected latent fingermarks, 828 non-identifiable fingermarks (NIFMs) were chosen that met program requirements for the study. These qualifying NIFMs showed three or more clear and reliable minutiae occurring within an area showing continuity of ridge flow (see Image 1).
An expected score-based likelihood ratio (ESLR) for each mark was measured, without reference to a putative source, based on modelling within-variability and between-variability of AFIS scores. These scores were calculated from an AFIS system consisting of a database with 963,710 fingerprints, stripped from any personal information. These fingerprints are from retired records, purged over 20 years ago by the Swiss Federal Police following an upgrade of their AFIS system. These records have been made available only to the University of Lausanne (UNIL) for research purposes.
What the researchers found was that, out of the 828 NIFM, 78 of them had ESLR values that exceeded a world population estimate of 77 billion fingers (7.7 billion individuals) (Log10 > 10.88).
Of the remaining 750 NIFMs, 540 were from property crimes (largely burglaries and thefts), whereas 210 were from violent crimes (largely homicides, robberies and assaults, but including 33 from drug and firearm related crimes). From this pool of 750 NIFMs, they showed values of Log10 ESLR ranging from 1.05–10.88, with a mean value of 5.56 (s.d. 2.29), corresponding to an ESLR of approximately 380,000. What this average ESLR value means to explain the weight of evidence would be if we found matching characteristics that would occur randomly in one in 38,000 individuals (one in 380,000 fingers).
I guess the main question from this study is just how useful would NIFM evidence be in actual practice? Could this information be enough to justify a search warrant against a particular person? Could it be used to order surveillance or initiate a wiretap on a person’s phone calls? It could lead down a slippery slope if this kind of reporting is abused and not represented properly, within context of the actual associative weight of the evidence.
It is clear from this study that there are large numbers of cases where non-identifiable fingermarks occur that have high potential associative value, as indicated by their ESLR. These NIFMs are readily available, but not used, yet have associative value that could provide useful information. Using ESLRs to calculate associative strength is still quite novel and is not yet sufficiently defined and vetted for widespread use and acceptance. Of interest would be how the Courts and jury members interpret this information when it’s presented in our reports and given in oral testimony. How much probative weight do they give to this associative evidence? Both I and the authors from the study feel very certain though that such methods will be here very soon and new methods to measure the associative value of fingerprints is currently under active development.
If you wish to find out more on this study, you can find further details about the article and where to find it below.
Stoney, D. A., De Donno, M., Champod, C., Wertheim, P. A., & Stoney, P. L. (2020). Occurrence and Associative Value of Non-Identifiable Fingermarks. Forensic Science International, 110219. https://doi.org/10.1016/j.forsciint.2020.110219