Analysis of linear and nonlinear panel data with empirical applications and interpretation of final results.
Models and techniques for the analysis of cross-sectional data encountered in microeconomics and finance. Binary data, count data, duration and multinomial data models are covered. The course will also cover the generalized method of moments estimation techniques. The focus is on applied analysis and replication of published papers.
Wooldridge J.M. (2010) Econometric Analysis of cross-section and panel data, The MIT Press.
Cameron, A.C. and P. K. Trivedi (2005) "MICROECONOMETRICS: Methods and Applications", Cambridge University Press, New York.
Research articles and teaching material provided by the professor
Learning Objectives
The aim of the course is to expose the student to practicalities ancountered when doing applied research. Some programming skills will be developed and reproduction of published results attempted.
Students will be able to select the most appropriate modelling technique for an array of data types, conduct robustness checks and statistical tests. Students will be able to interpret and discuss the empirical results both in terms of their statistical implications and the implications they have for economic, financial and social theories.
Prerequisites
Econometrics (both Macro and Micro modules)
Teaching Methods
Hands on approach. Lectures followed by practical examples (reproduction of published papers results)
Further information
Additional material available on the Moodle platform
Type of Assessment
A paper on a topic chosen by the student and the instructor
(2 papers per student).
Course program
Longitudinal data analysis: definition and benefits for estimation and inference. Fixed and random effects estimators. Linear mixed regression models with both fixed and random effects. Nonlinear regression models for categorical repeated measurements: binary, categorical and counts data. Techniques for analysing longitudinal data with non-ignorable missing observations. Empirical applications will be provided using statistical software.
In the second part:
1. Maximum Likelihood: Principles, Properties, Mechanics, Classical Test Principles.
2. Binary Data Models: Link Functions, Interpretation of Coefficients, Latent Variable Models, Likelihood Analysis.
3. Count Data Models: Poisson Regressions, Likelihood Analysis, Over Dispersion: Negative Binomial Types I and II,
4. Duration Data Models: Survival Function, Hazard Rate, Likelihood Analysis.
5. Generalized Method of Moments: Moment Conditions and Identification, Instrumental Variables, MM Estimation, GMM: estimation, consistency, asymptotic distribution, Efficient GMM.
6. Multinomial Data Models: Multinomial Logit, Nested Logit, Multinomial Probit, Example: Delisting of Public Companies.