I am an Assistant Professor at the University of Pennsylvania in the Department of Criminology and the Department of Statistics and Data Science. My training is primarily in statistics and data science, and my work is driven by interdisciplinary projects in public policy, public health, and forensic science.
I am interested in the intersection of statistics and the law, particularly how and whether we can (and should!) answer legal questions by using statistics. I have written about nonparametric estimation in causal inference, as it relates to the legal question of whether one can attribute an outcome (e.g. cancer) to an exposure (e.g. exposure to a harmful chemical).
I have also written about the statistical foundations of forensic science. I have studied the proper way to state a conclusion in pattern-matching disciplines (e.g. fingerprints, hair microscopy, toolmarks identification) and what information a forensic examiner should not have access to when analyzing evidence. I developed an algorithm for toolmark identification using machine learning classification with 3D data. I am currently studying the accuracy of facial recognition technology as it is used by law enforcement.
I often serve as an expert witness and consultant in trials. I have discussed the statistical foundations of forensic science (in cases regarding hair microscopy, toolmarks, and statistics more broadly) and the basis for statistical arguments in child abuse (shaken baby syndrome). I enjoy explaining complex statistical or machine learning ideas to a lay audience, especially in high-stakes situations, such as the law or public health.
I am affiliated with these groups:
- CSAFE
- ASA Social Statistics Section
- Crime and Justice Policy Lab
- Penn Center for Causal Inference
- Quattrone Center for the Fair Administration of Justice
I am the mother of twins. My husband, Justin Humphreys, is a philosopher who studies ancient mathematics and the imagination.
Last updated: August 2024.