New Zealand Statistical Association 2024 Conference

James Curran

University of Auckland

What the Zeta!

This is joint work with Patrick Buzzini, Tatiana Trejos

Many people are familiar with forensic evidence such as DNA, fingerprints, fibres and so on. Many fewer people are familiar with forensic glass evidence. Glass evidence arises when glass is broken during the commission of a crime. The statistical interpretation of glass evidence preceded the statistical interpretation of DNA evidence by almost a decade. In this talk I will discuss an estimation problem that arises when considering activity level propositions for glass evidence – i.e. propositions that consider how the glass might have been deposited on a person of interest as well as the physical characteristics of the glass that link it to the crime scene. This talk will involve an obscure discrete distribution, as well as some extensions to said distribution. Some might liken it to Morris Dancing.

If you were at the IBS-AR conference last year then you might have heard this talk before.