CONTENT

Become an expert in the analysis and implementation of linear, nonlinear, and dynamic panel-data estimators using Stata. This course focuses on the interpretation of panel-data estimates and the assumptions underlying the models that give rise to them. The course is geared for researchers and practitioners in all fields. The breadth of the lectures will be helpful if you want to learn about panel-data analysis or if you are familiar with the subjects.

The concepts presented are reinforced with practical exercises at the end of each section. We also provide additional exercises at the end of each lecture and access to a discussion board on which you can post questions for other students and the course leaders to answer.

 

PREREQUISITES

Stata 16 installed and working

Course content of NetCourse 101 or equivalent knowledge

Familiarity with basic time-series, cross-sectional summary statistics and linear regression

Internet web browser, installed and working (course is platform independent)

 

PROGRAM

 

LESSON I

An introduction to panel data and its features

Getting started with panel data

Summary statistics and dynamics

Overview of basic concepts

Data generation

The regression model

Variance–covariance estimators

Margins and marginal effects

Basic panel-data estimation concepts

Moment-based estimation

Panel data, regression, and efficiency

Closing remarks

 

LESSON II

Random-effects model

The model

Fixed-effects model

Within estimator

Comparing within and random-effects estimates

First-differenced estimator

Deciding between random and fixed effects

Hausman test

Mundlak test

Population-averaged models

LESSON III

Probit model

Probit models for panel data: Random effects

Probit models for panel data: Population averaged

Probit models for panel data: Remarks

Logit model

Logit models for panel data: Random effects

Logit models for panel data: Fixed effects

Logit models for panel data: Population averaged

Poisson model

Poisson models for panel data

 

LESSON IV

Endogeneity

Cross-sectional estimation under endogeneity

Panel-data estimation under endogeneity

Dynamic models

Building your own dynamic models

A more complex dynamic structure

Concluding remarks

 

Note: There is a one-week break between the posting of Lessons 2 and 3; however, course leaders are available for discussion.