Telling the stories in shattered glass

20 July 2017

PhD student Anjali Gupta
PhD student Anjali Gupta

Anjali Gupta has been fascinated with true-crime shows and Sherlock Holmes-style stories since her childhood in India. But as a schoolgirl, she never thought she’d end up in New Zealand doing a doctorate on involving the forensic properties of shattered glass at crime scenes.

Glass is one of the most common types of trace evidence, and, outside of DNA, is the field of evidence interpretation that makes the most use of statistics, she explains. As you’d expect, glass evidence usually arises from broken windows. But there may be headlamp and windscreen glass at the scene of a hit-and-run, shards from someone’s spectacles at the wake of an assault, or bottle or vase glass at the site of a domestic dispute. Tiny fragments, typically smaller than a grain of salt, often end up in people’s hair and clothing, making them critical evidence.

Anjali says that forensic scientists can use the physical properties, refractive index and elemental characteristics of glass to identify the origin of shards. The elemental characteristics can be measured using Laser Induced Breakdown Spectroscopy (LIBS), an analytical chemistry technique. “LIBS is cheap compared to other methods. It’s portable so you can use it at the crime scene, and samples don’t need preparation,” she says.

But there are a few drawbacks: LIBS generates data as a spectrum (rather than dots) to represent the isotopes of the elements present and their intensities, so it’s a challenge to analyse and interpret the results. Forensic scientists have also been concerned about the reliability of LIBS, and Anjali’s research focus is improving its precision.

Anjali started her academic career with a BSc(Hons) in Statistics at the University of Delhi followed by an MSc in Applied Statistics at the University of Oxford; her thesis focused on modelling and forecasting yield rates of US Treasury securities. She then worked in the energy, finance and credit sectors, enjoying the variety: “The underlying tools and concepts of statistics that you use are the same,” she says. “I enjoy learning different things, new concepts and applying statistics to varied domains.”

She was looking for a change of field when a browse on the internet took her to Professor James Curran’s work on the statistical interpretation of forensic glass evidence. Anjali was fascinated, though her knowledge about the forensic world was limited to the crime investigation shows and thrillers she has loved since her childhood. She got in touch with Professor Curran, and he supplied some papers to read. “As I read, it inspired me even more towards working with him.” When Professor Curran posted a PhD project in the scientific examination of glass, Anjali applied, and started her research in February 2015.

She says she is grateful that Professor Curran took her on. “Even though I didn’t have any prior experience in forensic statistics, he shared the relevant reading material so that I could learn and pursue research under his mentorship.”

Anjali’s now hooked, and plans to continue in forensic statistics after her thesis is complete.