Description: Abstract: This overview of fixed and random effects models from an econometric perspective will describe how panel data can be used to mitigate concerns about unobserved factors affecting a regression. The presenters will describe the assumptions that need to be met for each model to be used and describe tests that can be used to choose between fixed and random effects models. Finally, how statisticians think about fixed, random, and mixed effects models and how this can differ from an econometric perspective will be addressed.
Target audience: Researchers who would like an introduction to econometric methods for observational studies in health services research. Seminar material will assume knowledge of basic probability and statistics and familiarity with linear regression.