Previous methods to evaluate evidence from handwriting examinations were usually associated with a redefinition of how these examinations are to be made. Here we propose the likelihood ratio method for handwriting evidence evaluation which is fully compatible with the current handwriting examination protocols. The method is focused on the similarity between handwriting samples, quantified using Jaccard index from results of a usual forensic handwriting comparison. The numerator of the likelihood ratio is the probability of a given class of similarity, assuming that a given person wrote the questioned sample. The denominator is the probability of the same class of similarity, assuming that a randomly selected person wrote questioned sample. The similarity distribution to quantify the numerator is derived from comparisons across reference handwritings. To calculate the denominator we propose to develop similarity distributions relevant for particular forensic scenarios. In the proof-of-a-concept study, we developed the distribution for the simulation scenario.
You can access the full paper here: Likelihood ratio to evaluate handwriting evidence using similarity index
This project was my first foray into the world of combining forensic science with data science. As with any first steps, the journey was a bit messy—think of it as my “origin story.” The scripts might look like a chaotic battlefield at times, and honestly, I’m not even sure if I’ve included all the R scripts here. It was so long ago, and the chaos was real! But from that disorder emerged a successful and insightful outcome. It was a learning process, and like any good hero, I grew stronger from the experience.
Due to the sensitive nature of the data involved, and in compliance with data protection regulations, I am unable to share the datasets used in this project. Rest assured, all analyses were conducted with the utmost respect for privacy and confidentiality.