Applied Quantitative Methods – Intermediate Econometrics I (November/December 2009)

Faculty
Keld Laursen, Professor, Department of Innovation and Organizational Economics, CBS; Toke Reichstein, Associate Professor, Department of Innovation and Organizational Economics, CBS; Francesco Rullani, Assistant Professor, Department of Innovation and Organizational Economics, CBS; Larissa Rabbiosi, Assistant Professor, Center for Strategic Management and Globalization, CBS

Course Coordinator
Toke Reichstein, Associate Professor, Department of Innovation and Organizational Economics, CBS

Prerequisite
The course requires that the students have taken the course on “Introduction to basic econometric techniques” or a similar course. Students that have a master educational background with a strong econometrics program may also feel comfortable with entering at this stage.
Aim of the course

This course focuses on categorical variables and limited dependent models. We provide an introduction to the analysis of counts and ordered categorical variables with the binary being the special case. The course therefore starts by introducing simple cross tabulation analysis and then moves on to debate simple logistic, multinomial logistic and ordered logistic regression. We illustrate the use of the logistic regression in developing useful samples for analysis such as matching procedures. We provide basic understanding of the specification and estimation of count regression models. The aim is to train the students to carrying out robust quantitative studies on their own, but also to provide them the tools that will enable them to give constructive critique to econometric work in general.

The course's development of personal competences

After the course, students are expected to be able to:

Understand the pitfall of running simple OLS regressions when studying relationships in which the dependent variable is categorical
Do tabulations of categorical variables and carry out simple tests to investigate skewness across categories
Be able to identify how and when to do logistic, ordered logistic, multinomial logistic, and count regressions
Estimate marginal effects and understand why they are useful
Identify the pitfalls of including interaction terms in limited dependent variables regressions and what to do about it
Create a control sample to match a treatment sample and understand why this is of importance
Carry out difference and difference models
Identify count variables, carry out count variables regressions, and correct for any bias that may be attributed to overdispersion through zero inflation.

Lecture plan

Time/period	Faculty	Title
30 November		Discrete and Categorical data - why treat them differently
		Tabulations and tests of skewness across tabulations
		Dummies as dependent variables
		Logistic Regressions
		Interaction terms in limited dependent variable regressions
1 December		Ordered Logistic Regressions
		Marginal Effects
		Treatment samples and Control Samples
		Matching procedures – creating a control sample
		Difference in Differences models
2 December		Multinomial Logistic Regressions
		Tobit Model
		Count models and Zero-inflated count models
		Training the learnt methods through practice

Teaching methods

Lectures, discussion and exercise.

The course employs STATA throughout the sessions furnishing students with examples and assignments for illustration. The methods taught will hence be displayed and demonstrated in STATA. Participants can bring own laptops with STATA installed and use these throughout the course. However, there will be desktop computers with STATA installed available. We encourage students to bring a memory stick or external hard drive so that they can work directly on these and bring all files used and programs written with them home after the course.

Examination

TBA

Course literature

Students attending the course will receive a more detailed recommended reading list assisting them in preparing for the course and pointing to readings that may prove helpful as a follow up after attending a session.

Recommended literature

Wooldridge, J. M. (2001), “Econometric Analysis of Cross Section and Panel Data”, The MIT Press.
Hamilton, L. C. (2006), “Statistics with Stata”, Brooks/Cole, Belmont, CA.

Enrolment

Applications should be sent to:

The Doctoral School in Economics and Management
Copenhagen Business School
Attn: Shi Hua Chen Kold
Department of Innovation and Organizational Economics
Kilevej 15, 3rd. floor
2000 Frederiksberg
Denmark
E-mail: shc.ino@cbs.dk

Other

Fee:

525 Euro

Not to be paid by PhD students enroled at a Danish University

Sidst opdateret af Kristine Olsen 20.10.2009