tag:blogger.com,1999:blog-5354567765897135804.post909858442735608484..comments2023-03-16T23:35:17.345-04:00Comments on Forensic Science, Statistics & the Law: Latent Fingerprint Identification in Flux? DH Kayehttp://www.blogger.com/profile/09329862957840849989noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-5354567765897135804.post-33432815096309865902016-01-28T10:04:26.522-05:002016-01-28T10:04:26.522-05:00With regards to the Likelihood Ratio, what are you...With regards to the Likelihood Ratio, what are your thoughts on <a href="https://forensiccoe.org/module/registration/Default.aspx?moduleid=BBFD983F-843E-4BF8-816B-AF47781ACB75" rel="nofollow">Hari Iyer's (NIST)</a> claims that there are some concerns with the LR due to the fact that the priors cannot be known and assigning them all 1 is not an adequate solution?<br /><br />Perhaps, as conclusions move further from binary (ID/Exclusion)decisions, they might fare better using non Bayesian diagnosticity as outlined in <a href="http://www.jonathandnelson.com/papers/usefulQuestions.pdf" rel="nofollow">Jonathan Nelson's 2005 paper</a><br />Boydhttps://www.blogger.com/profile/05434111509389990203noreply@blogger.com