Deutsch | English  | print

E5007 Multivariate Analysis

Course title: E5007 Multivariate Analysis

Instructor: Alexandra Adveenko

Offered: fall semester 2015

Method (hours per week): lecture (2)

Course level: Master

Course language: English

Requirements and Exam: regular attendance, written exam

ECTS-Credits: 5

Course Description:
The majority of data sets collected by researchers in all disciplines are multivariate. The correct application of statistical methods allows extracting as much information as possible from the data at hand. Therefore, main objective of is to give students a practical introduction to multivariate analysis and econometrics. The lecture will include methods both for describing and exploring data and for making formal inferences about them. Additionally, this course will also offers an introduction to advanced programming in Stata. The lecture will start with an introduction to efficiently written do-files (including data processing). Different data types will then be presented, i.e. the German Socio-Economic Panel (SOEP). In hands-on sessions students will be taught how to prepare the data for analysis. Variables will be generated and their distributions explored; data will be merged; and regression results will be critically discussed. Moreover, in this course students will learn how to implement new commands for Stata and to conduct Monte Carlo simulations. These are important for verification of implementations and are used as a very important tool to analyse the small sample properties of estimators and to complement the theoretical properties of estimators making them an integral part of econometric analyses. We will also touch upon Stata's matrix programming language Mata. Moreover, we will apply the programming techniques to implement selected cross-section models.

Literature: Cameron/ Trivedi (2009). Mircoeconometrics using Stata. Stata Press.

Contact person: Alexandra Avdeenko, Zimmer 102, Tel.: +49-621-181-3500, E-Mail: avdeenko(at)