Schedule

Multilevel Modeling Workshop
University of Kentucky

Brandon Bartels
George Washington University

Day 1: Monday, May 23, 2011

· Multilevel data structures and key concepts

· Linear variance components model

· Four general modeling approaches: (1) complete pooling (OLS), (2) no pooling (fixed effects, or within-cluster), (3) partial pooling (random intercept and random coefficient models), and (4) between-cluster models.

Day 2: Tuesday, May 24, 2011

· Fixed versus random effects

· “Controversial assumption” in random intercept model

· Cluster confounding

· Issues in panel and time-series cross-sectional (TSCS) data

· Random coefficient model

Day 3: Wednesday, May 25, 2011

· Random coefficient model (continued)

· Nonlinear models for noncontinuous dependent variables