[Forside] [Hovedområder] [Perioder] [Udannelser] [Alle kurser på en side]
• derive models for binary dependent variable
• apply multinomial response variable models
• derive models for ordered response variables
• Analyze models for censored data
• Analyze models for truncated sample
• analyze selection model
• derive model for duration and count data
• Analyze data using a log-linear model
In many empirical situations, the purpose is to analyze variables, which are qualitative. For example, the choice to buy a product or to accept a job is qualitative. In this course, the purpose is to explain such variables. For example, the choice to buy a product depends on the price of the product, the prices of other products and individual characteristics like income and education.
In addition, the course considers situations where the dependent variable is a result of a qualitative outcome. For example, the sample may be truncated (do not observe certain individuals) or censored (do not observe the outcome of certain individuals). Finally, the course also considers models for duration in a particular state, for example, the duration of unemployment.
Course subject areas
• binary logit and probit
• multinomial logit model
• ordered logit and probit model
• truncated samples
• censored data
• selection model
• event count and duration model
• log-linear model
Regression analysis
Valeri Voev
The lectures present the main points of the analysis of discrete dependent variables. There will be computer exercises to practice the application of the methods.
Heij, De Boer, Franses, Kloek, Van Dijk, 2004: Econometric Methods with Applications in Business and Economics. Oxford University Press Inc., New York. Chapter 6 (app. 90 pages)
Wayne S. Desarbo and David K. Hildebrand: "A Marketer's Guide to Log-Linear Models for Qualitative Data Analysis." The Journal of Marketing, Vol. 44, No. 3 (Summer, 1980), pp. 40-51
Greene 2008: Econometric Analysis. Ch. 25.1-2 (9 pages)
Form of assessment: A four hour written exam together with the course regression analysis.
Examination aids allowed: All - except any means of electronic communication including PCs.