Speaker: Emily Cantrell

Institution: Princeton

Date: 14:00 BST, 15th October, 2026

Talk Title: Machine Learning Models Trained on a Longitudinal Survey Dataset Poorly Predict Hundreds of Life Outcomes

Abstract: Understanding the predictability of life outcomes is important for public policy and science, but existing studies focus on a limited number of outcomes and therefore struggle to describe the broad patterns of predictability. We developed a high-throughput approach that uses a single dataset to study the predictability of many future outcomes using past data. We deployed this approach with thousands of machine learning pipelines for hundreds of outcomes in the Future of Families and Child Wellbeing Study. We use this high-throughput approach to assess predictive accuracy, study the “skill” of complex modeling procedures compared to simple benchmarks, and search for patterns in what types of life outcomes are most predictable.

Bio: Emily Cantrell is a PhD candidate at Princeton University. Her dissertation examines the capabilities and limitations of machine learning models that make predictions about individuals’ futures. Emily is passionate about using data science and computational methods to advance evidence-based solutions to poverty and inequality. She previously worked at the economic research institute Scioto Analysis, the child and family policy research institute Child Trends, and in state government assisting the Health and Human Services Committee in the New Mexico House of Representatives.