Statistical Modeling & Synthetic Populations

Stage II, Course 8 · 3 credits · 14 weeks · From systems to populations while preserving structure

Course Purpose

Students learn to build population-level models while preserving system structure. Move from single systems to aggregated ensembles. Understand how individual behavior produces population distributions, and how to work backward from population data to infer system properties.

By end: Students can build statistical models of populations, create synthetic populations that preserve structural properties, recognize Simpson's paradox and aggregation artifacts, and move between individual and population levels without losing information.

Learning Outcomes

1. Understand distributions as system outputs — how individual systems produce populations
2. Model population-level relationships — conditional dependence and correlation
3. Recognize Simpson's paradox — how aggregation can reverse correlations
4. Create synthetic populations — preserve structure while generating new instances
5. Move between levels of analysis — aggregation and disaggregation
6. Understand sampling and inference — from sample to population
7. Validate statistical models — goodness of fit and assumption checking

Core Concepts

distribution, population, sample, conditional-dependence, correlation, Simpson's-paradox, synthetic-population, aggregation, sampling, inference, parameter-estimation

Course Structure (14 weeks)

← Back to Stage II