Skip to content

A Gentle Introduction to the Likelihood Ratio: Basic Ideas, Implementation, and Limitations

Conference/Workshop:
107th International Association for Identification (IAI) Annual Educational Conference
Published: 2023
Primary Author: Alicia Carriquiry
Secondary Authors: Jeff Salyards
Research Area: Forensic Statistics

The workshop focuses on the likelihood ratio (LR) approach in forensic science. The LR, a one-number summary, quantifies how well the observations/results are explained by the prosecution’s versus the defense’s propositions. While the basic idea behind the LR is simple and intuitive, challenges arise when trying to implement the approach on different types of evidence. Presenters will discuss the statistical foundations of the LR, applications in different forensic disciplines, best practices, and limitations.

Related Resources

Significance of image brightness levels for PRNU camera identification

Significance of image brightness levels for PRNU camera identification

A forensic investigator performing source identification on a questioned image from a crime aims to identify the unknown camera that acquired the image. On the camera sensor, minute spatial variations…
Methodological problems in every black-box study of forensic firearm comparisons

Methodological problems in every black-box study of forensic firearm comparisons

Reviews conducted by the National Academy of Sciences (2009) and the President’s Council of Advisors on Science and Technology (2016) concluded that the field of forensic firearm comparisons has not…
First impressions matter: Mundane obstacles to a forensic device for probabilistic reporting in fingerprint analysis

First impressions matter: Mundane obstacles to a forensic device for probabilistic reporting in fingerprint analysis

This article investigates why statistical reasoning has had little impact on the practice of friction ridge (or ‘fingerprint’) examination, despite both interest and some modest scientific progress toward this goal.…
A Dirichlet process model for directional-linear data with application to bloodstain pattern analysis

A Dirichlet process model for directional-linear data with application to bloodstain pattern analysis

Directional data require specialized models because of the non-Euclidean nature of their domain. When a directional variable is observed jointly with linear variables, modeling their dependence adds an additional layer…