|
Activity Number:
|
432
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
IMS
|
| Abstract - #305645 |
|
Title:
|
Calculation of Likelihood Ratios from Categorical Data with Applications to Forensic Handwriting Analysis
|
|
Author(s):
|
Christopher P. Saunders*+ and Amanda B. Hepler and Mark J. Lancaster and Linda J. Davis and Donald T. Gantz
|
|
Companies:
|
George Mason University and George Mason University and George Mason University and George Mason University and George Mason University
|
|
Address:
|
4400 University Drive, Fairfax, VA, 22030,
|
|
Keywords:
|
Forensic Statistics ; Questioned Documents ; Categorical Data
|
|
Abstract:
|
In forensic sciences there are two competing hypotheses: the suspect left the trace, or the suspect is a random man. With respect to questioned document examination, these hypotheses reduce to "the suspect wrote the questioned document" versus "the suspect did not write the questioned document." The relevant probabilities that form the likelihood ratio are the probability of observing the questioned document given the suspect wrote it, and the probability of observing the questioned document given a randomly selected individual wrote it. A recent development by the Gannon Technologies Group in the quantification of handwritten documents allows for the representation of a handwritten document as a vector of counts. A review of Gannon's quantification procedure on writing samples collected from 454 writers and the related issues of calculating forensic likelihood ratios are presented.
|
