Authorised by Academic Registrar, April 1996
Objectives On successful completion of this subject students can expect to have an appreciation of likelihood methods and their role in econometrics; be able to construct likelihood functions; know how to find maximum likelihood estimates either analytically or with the aid of a computer; given the information matrix, know how to construct likelihood ratio, Lagrange multiplier and Wald tests; have an understanding of linear simultaneous equation models and their methods of estimation; have an appreciation of the identification problem in simultaneous equation models.
Synopsis This subject presents a unified theory of likelihood methods used to estimate and test the validity of economic relationships. Maximum likelihood estimation and likelihood ratio, Wald and Lagrange multiplier tests form the backbone of this methodology. Applications include the classical linear regression model, the generalised linear regression model, tests for linear restrictions and autocorrelation in the regression model. Other topics include an introduction to simultaneous equation models, the identification problem, instrumental variable estimators, limited information and full information simultaneous equation estimators.
Assessment Written (6 fortnightly assignments): 40% + Examination (2 hours): 60%