|
|
This course is the first course
in applied statistical methods for social scientists. Students
will
learn a variety of basic cross-section regression models (as
time permits!) including linear regression model, discrete choice
models, duration (or hazard) models, event count models,
structural
equation models, and others. Unlike traditional courses on
applied
regression modeling, I will emphasize the connections between
these
methods and causal inference, which is the primary goal of social
science research. Prerequisites, POL
502 and POL 571.
|
Statistical Framework of Causal
Inference: Counterfactuals, Potential Outcomes,
Principal Stratification
|
Neyman's Design-Based Inference
: Statistical Analysis of Randomized Experiments,
Difference-in-Means Estimator
|
Simple Regression
: Linear Regression with a Single Variable, Finite Sample and
Asymptotic Inference, Regression Discontinuity Design
|
Multiple Regression
: Regression with Multiple Variables, Finite Sample and Asymptotic
Inference, Omitted Variables Bias, Measurement Error
|
Structural Equation Modeling
: Linear Structural Equation Modeling, Causal Mediation Analysis,
Instrumental Variables
|
Likelihood Inference
: Maximum Likelihood Estimation, Bootstrap, Likelihood Ratio Test, Normal
Regression, Logit/Probit Models
|
Discrete Choice Models
: Ordered/Multinomial Logit/Probit Models, Sample Selection Model
|