Statistical population modelling is a powerful tool for producing gridded population estimates that can support census activities. In this project we partnered with the United Nations Population Fund (UNFPA) to share our expertise and provide support to national statistics offices in training and production of high-resolution gridded population estimates from existing data sources (e.g. household surveys, building footprints, administrative records, census projections).

We developed a series of tutorials in Bayesian statistics for population modelling that are openly accessible to national statistical offices and other stakeholders globally. The tutorial series covers key concepts in Bayesian statistics that are required for population modelling and include example code and other resources designed to expedite the learning curve as much as possible.

These tutorials serve as stand-alone resources for self-guided learning that provide the core curriculum for in-person capacity strengthening workshops.

The key concepts covered in the tutorial include:

• Introduction to software for Bayesian statistical modelling: R and Stan,
• Simple linear regression in a Bayesian context,
• Random effects to account for settlement type (e.g., urban/rural) and other types of stratification in survey data,
• Quantifying and mapping uncertainties in population estimates using hierarchical variances, and
• Diagnostics to evaluate model performance (e.g., cross-validation)