Issues and opportunities: the application of the numerical likelihood ratio framework to forensic speaker comparison

Erica Gold; Vincent Hughes

doi:10.1016/j.scijus.2014.04.003

Issues and opportunities: the application of the numerical likelihood ratio framework to forensic speaker comparison

Erica Gold, Vincent Hughes

Language and Linguistic Science

Research output: Contribution to journal › Article › peer-review

Abstract

Across forensic speech science, the likelihood ratio (LR) is increasingly becoming accepted as the logically and legally correct framework for the expression of expert conclusions. However, there remain a number of theoretical and practical shortcomings in the procedures applied for computing LRs based on speech evidence. In this paper we review how the LR is currently applied to speaker comparison evidence and outline three specific areas which deserve further investigation: namely statistical modelling, issues relating to the relevant population and the combination of LRs from correlated parameters. We then consider future directions for confronting these issues and discuss the implications for forensic comparison evidence more generally.

Original language	English
Pages (from-to)	292-299
Number of pages	8
Journal	Science & justice : journal of the Forensic Science Society
Volume	54
Issue number	4
Early online date	9 May 2014
DOIs	https://doi.org/10.1016/j.scijus.2014.04.003
Publication status	Published - Jul 2014

Keywords

Likelihood ratio
Phonetics
Forensic speech science
Modelling
Relevant population
Correlations

Access to Document

10.1016/j.scijus.2014.04.003

Cite this

@article{05570bbe791342eabc149f720463b011,

title = "Issues and opportunities: the application of the numerical likelihood ratio framework to forensic speaker comparison",

abstract = "Across forensic speech science, the likelihood ratio (LR) is increasingly becoming accepted as the logically and legally correct framework for the expression of expert conclusions. However, there remain a number of theoretical and practical shortcomings in the procedures applied for computing LRs based on speech evidence. In this paper we review how the LR is currently applied to speaker comparison evidence and outline three specific areas which deserve further investigation: namely statistical modelling, issues relating to the relevant population and the combination of LRs from correlated parameters. We then consider future directions for confronting these issues and discuss the implications for forensic comparison evidence more generally.",

keywords = "Likelihood ratio, Phonetics, Forensic speech science, Modelling, Relevant population, Correlations",

author = "Erica Gold and Vincent Hughes",

year = "2014",

month = jul,

doi = "10.1016/j.scijus.2014.04.003",

language = "English",

volume = "54",

pages = "292--299",

journal = "Science & justice : journal of the Forensic Science Society",

issn = "1355-0306",

publisher = "Forensic Science Society",

number = "4",

}

TY - JOUR

T1 - Issues and opportunities: the application of the numerical likelihood ratio framework to forensic speaker comparison

AU - Gold, Erica

AU - Hughes, Vincent

PY - 2014/7

Y1 - 2014/7

N2 - Across forensic speech science, the likelihood ratio (LR) is increasingly becoming accepted as the logically and legally correct framework for the expression of expert conclusions. However, there remain a number of theoretical and practical shortcomings in the procedures applied for computing LRs based on speech evidence. In this paper we review how the LR is currently applied to speaker comparison evidence and outline three specific areas which deserve further investigation: namely statistical modelling, issues relating to the relevant population and the combination of LRs from correlated parameters. We then consider future directions for confronting these issues and discuss the implications for forensic comparison evidence more generally.

AB - Across forensic speech science, the likelihood ratio (LR) is increasingly becoming accepted as the logically and legally correct framework for the expression of expert conclusions. However, there remain a number of theoretical and practical shortcomings in the procedures applied for computing LRs based on speech evidence. In this paper we review how the LR is currently applied to speaker comparison evidence and outline three specific areas which deserve further investigation: namely statistical modelling, issues relating to the relevant population and the combination of LRs from correlated parameters. We then consider future directions for confronting these issues and discuss the implications for forensic comparison evidence more generally.

KW - Likelihood ratio

KW - Phonetics

KW - Forensic speech science

KW - Modelling

KW - Relevant population

KW - Correlations

U2 - 10.1016/j.scijus.2014.04.003

DO - 10.1016/j.scijus.2014.04.003

M3 - Article

SN - 1355-0306

VL - 54

SP - 292

EP - 299

JO - Science & justice : journal of the Forensic Science Society

JF - Science & justice : journal of the Forensic Science Society

IS - 4

ER -