Christian Caron, PhD

Bayesian multilevel regression and poststratification (MRP) is a statistical technique that combines hierarchical modeling with demographic and geographic population data to generate reliable estimates for subgroups, such as individual states, even when direct survey samples are sparse. By pooling information across groups and then weighting results to reflect the true population, Bayesian MRP produces more stable and nuanced estimates than simple disaggregation. The following series of interactive maps apply Bayesian MRP to visualize variation across states, allowing users to explore estimated levels of support along with the associated uncertainty. By clicking on each state, viewers can examine point estimates alongside 95% credible intervals. Estimates were derived from the 2024 Cooperative Election Study.