Episode 291 - Mundane Obstacles to Developing a Stats Model
from Double Loop Podcast Hosted by Glenn Langenburg and Eric Ray
Eric and Glenn start the episode with a listener email and then play a numbers themed game of Regional Quirkisms. For the topic of this episode, the guys discuss a 2025 paper from Simon Cole and Justin Sola titled "First impressions matter: Mundane obstacles to a forensic device for probabilistic reporting in fingerprint analysis". The article discusses practical and realistic obstacles and hurdles to developing a validated, accepted, and commercially-available statistical model for fingerprint evidence. The article takes the novel view that most sources cite either "practitioner disinterest" (or worse: practitioner rejection) or lack of push and influence from the Courts, Daubert challenges, or Authoritative bodies. However, once you remove those two obstacles and assume examiner 'buy-in' and Courts pushing for empirical data over subjective examiner expertise, what 'mundane' obstacles exist that most people don't realize or think about? The article discusses a number of surprising, non-trivial obstacles that have a huge influence on the state of statistical model development in the field today. Article: Cole, S.A.; Sola, J.L. First impressions matter: Mundane obstacles to a forensic device for probabilistic reporting in fingerprint analysis. Social Studies of Science (2025), 55(5): 683-710. https://doi.org/10.1177/03063127251333074
Embed this episode
No transcript yet.