Course Director: Dr. James O'Malley
- Course description: The first component of the course extends regression methods for analyzing data with one or more forms of statistically dependence. Clustered, multi-level, longitudinal and other forms of structured data will be discussed along with the appropriate hierarchical (mixed-effect) models for accounting for such dependence. Estimation will be performed from a frequentist perspective. The consideration of random effects and their conditional distribution given that data will set the scene for the second major topic. Bayesian statistical methods are an important component of the course due to their affinity for solving challenging problems and their ubiquity across modern statistical applications. Bayesian methodology is carefully developed and compared to the classical (frequentist) approach using a variety of applications. Bayesian computation via Markov-chain Monte-Carlo (MCMC) is developed and illustrated. The course concludes with methods for social network analysis including representation, visualization and summarization of networks. Statistical models of networks, models of peer effects and social influence processes, and analyses involving multiple networks will be developed. The R statistical package will be used throughout with the BUGS/JAGS language also being incorporated to assist the estimation of Bayesian models.
- Dartmouth Coursework: QBS 120 and QBS 121.
- Linear algebra, multivariate calculus, statistics, probability and basic computer programming with an emphasis on mathematical/statistical programming.
- Programming: Intermediate proficiency in R.