ECON686: Panel Data Analysis, Term II  2018-19

This course introduce models and methods for analysing panel and longitudinal data that arise from repeated measurements taken from a cross-section of subjects. These models and data have found substantive applications in many disciplines within the social and biological sciences where the describers “panel” and “longitudinal” respectively arose. This course is primarily oriented to the applications in social science including business, economics, education and sociology. Common models include (i) one-way error components panel data models (with fixed effects, random effects, or correlated random effects); and (ii) two-way error components panel data model (with fixed effects, random effects, or correlated random effects). Panel data model with heteroskedasticity and serial correlation and tests of hypotheses with panel data will also be covered. Major inference methods such as (quasi) likelihood and GMM will be briefly introduced. A brief introduction will be given to the dynamic panel data models and spatial panel data models. Throughout, the course emphasizes on the models and their applications with computer software STATA.

Prerequisites include a good background in mathematical statistics and econometrics at the level of Greene (2012): Econometric Analysis, 7th Ed (Prentice Hall, New Jersey).

Upon completion of the course, students should be able to analyse the panel data frequently arose in business, economics, health, education, sociology, etc., using an appropriate panel data model with the assistance of the computer software STATA or others.


Course Outline

 Lecture Material


 Data Sets

 Stata Docs


Lecture time:

Tues. 8:15:00-11:45,   First Class: 19 March 2019

Lecture Venue:



Tuesday 4:00-5:00pm, at my Office 50-69, SOE;

Or by appointment, Office Phone: 68280852, email: