I am an epidemiologist and data scientist specializing in spatial analysis, population health, and manchine learning. My work focuses on building reproducible pipelines and identiying meaningful patterns and trends associated with health outcomes.

Bayesian spatial models (BYM2) for tract-level rate smoothing with credible intervals and population offsets.

Generates tract-level populations using ACS tables (B01001, B11016, B19001) with household structure and age distributions.

Tools for validating and diagnosing composite index inputs, including correlation screening, PCA summaries, and CFA workflows.

• Scaling synthetic population methods to national coverage
• Improving validation frameworks for composite indices
• Integrating spatial models with age-adjusted rate estimation

• LinkedIn: [https://www.linkedin.com/in/stephen-scroggins-824b1213/]
• Selected Publications: [https://www.geoscroggins.com/]