Recently published research from the Center for Statistics and Applications in Forensic Evidence (CSAFE) shows how a novel application of statistical methods provides fresh insights into how different jurors may respond to presentations of forensic evidence.
The study, published in Law, Probability & Risk, used mixture models to determine whether subpopulations exist among samples of mock jurors and whether these subpopulations have clear differences in how they perceive forensic evidence.
Naomi Kaplan-Damary, a post-doctoral researcher in statistics at the University of California, Irvine, and the co-authors of the study analyzed data from three previously published studies in which jury-eligible adults evaluated pairs of statements regarding forensic evidence and judged which statements they felt were stronger. Each study used different types of forensic evidence and different presentations of the data, including both quantitative and qualitative statements. The studies measured five participant characteristics: gender, age, numeracy, education level and forensic knowledge.
In all three studies, the authors utilized the mixture model approach, which does not require pre-defined hypotheses regarding important characteristics, to analyze participant responses and identify subpopulations.
Data from the three studies suggest that subpopulations exist and perceive statements presented by the examiner differently. By using the mixture model approach, the researchers found subpopulations undetected by the hypothesis-driven approach in which researchers determine in advance the characteristics or group differences that they deem relevant. From the results, they also found that participants with higher numeracy tended to respond more strongly to quantitative statistical statements, such as those based on likelihood ratios; while those with lower numeracy preferred more categorical statements.
Kaplan-Damary states that the existence of group differences in how evidence is perceived suggests that forensic scientists may need to present their findings in multiple ways.
“Perhaps there is no single approach that works best for all potential jurors, but by offering several alternative statements, forensic experts may be able to assure that their testimony is more broadly understood,” she said.
Also contributing to the study were William C. Thompson, professor emeritus of criminology, law and society; Rebecca Hofstein Grady, a former doctoral student; and Hal Stern, CSAFE co-director and Chancellor’s Professor of statistics. They are all from the University of California, Irvine.
View and download the journal article at https://forensicstats.org/blog/portfolio/using-mixture-models-to-examine-group-difference-among-jurors-an-illustration-involving-the-perceived-strength-of-forensic-science-evidence/.
View insights from this study: https://forensicstats.org/blog/2021/02/06/insights-using-mixture-models-to-examine-group-differences-among-jurors/.
