Advanced Quantitative Methods: Modelling Non-Continuous Outcomes and Panel/Longitudinal data using STATA (2 Day Schools)

Code: FCS680 (0 Credits) FCS679 (5 Credits) , Offered By: Social Sciences DTP

Social Sciences Researchers Only

This is a 2 day course focused upon modelling discrete outcomes, e.g. probabilities, and censored outcomes, e.g. consumption. These models depart from the multiple regression framework which is based upon a linear specification. There will also be a focus upon panel data and the advantages/disadvantages this brings in terms of estimation and interpretation. The course will introduce students to STATA which is one of the main statistical packages used by social scientists. The course will be ideal for those participants who want to be able to analyse non discrete outcomes and/or who are using panel data in the most robust and accurate.


Recommended For

Year 1 Social Science PhD students only

First Year PhD students who have completed the Quantitative Methods Core Module in Semester 1.  A reasonable working knowledge of multiple regression, i.e. Ordinary Least Squares or Maximum Likelihood, is highly recommended and will be assumed throughout. No prior knowledge of STATA is necessary. No prior knowledge or experience of modelling non-continuous outcomes or panel data will be assumed.

Aims Objectives

It is envisaged that the following topics will be covered:
  • An introduction to modelling binary discrete dependent variables, e.g. probability of graduating, and the problems associated with the linear probability model (OLS).
  • Modelling categorical ordered outcomes, e.g. psychological health, bullying, and unordered outcomes, e.g. mode of transport to work, consumer brand choices.
  • Count models and over-dispersion, e.g. number of doctor visits, patents.

  • Censored outcomes and corner solutions, e.g. consumer demand.

  • Analysis of panel/longitudinal data, i.e. covering the same cross sectional units over multiple time periods. This will be applied to continuous and binary outcomes

  • The focus is upon understanding the basic models and in particular the interpretation of the parameter estimates and what they mean in the real world.

Teaching Methods

The course comprises a mixture of lectures on the underlying and is taught throughout via examples. Participants will also have the opportunity to practice the skills learnt on real data.

Assessments and Credit

Details will be provided during the day schools

Other Information

Staff Contact

Karl Taylor  


Sorry, there are no sessions scheduled for this seminar at the moment. You can express add an expression of interest to your basket and you will be notified when any new sessions become available.

Your Basket
Your Basket is Empty