Theses

We are pleased that you would like to write your bachelor/master thesis at our institute. On this website you may find information about topics, registration and general guidelines. We wish you much success!

BACHELOR THESES

Bachelor theses in Statistics consist of a description of new statistical methods and their application on specific data.

  • Registration

    We would like to ensure that all students of the Economics and Management bachelor receive a topic for their theses within standard period of study. Therefore, allocation is made by the Office of the Dean of Studies via a centralized application procedure. Afterwards, the Office of the Dean of Studies will inform you about the institute which you were assigned to.

  • Assignment of Topics

    We have prepared a selection of topics for students who were assigned to our institute. Topics for summer term 2021 as well as winter term 2021/22 will be assigned within a online meeting, which take place on Tuesday, 4th May 2021 at 16.00. You can access the online meeting here. You can find a list with all current topics on our website. 

    Please communicate your preferences to the office by email till Friday, April 30th, 2021.

    Pre-registrations to start the bachelor thesis earlier can still be sent to the office by email till Friday, April 23th, 2021.

TOPICS

We have arranged a selection of 30 topics from different areas.

In the following, you may find a list of currently available topics which may help for your choice. If a topic is no longer available, we will remove it from our website.

The descriptions of content as well as liturature references can be taken from the general overview which is linked above.

  • Limits of Classic Linear Regression

    Autocorrelation

    Regression with time series data could cause classical assumptions about the OLS estimator to be violated, rendering it ineffective. Autocorrelation is an example of this. When autocorrelation exists, the errors of a linear regression are time dependent. In this thesis the AR(1) error model should be presented. In addition, a test on autocorrelation should be presented and shown how to estimate linear regression models efficiently despite autocorrelation.

    Introductory Literature:

    • J.M. Wooldridge. 2013. Introductory econometrics: A modern approach. Nelson Education (Chap. 12)



    Endogeneity: Generalized Method of Moments (GMM)

    Endogeneity bias can lead to inconsistent estimates and incorrect inferences, which may provide misleading conclusions and inappropriate theoretical interpretations. GMM is a statistical method that combines economic data with the information in population moment conditions and is able to estimate all coefficients simultaneously. The idea behind GMM must be explained and then applied to solve the system.

    Introductory Literature:

    • W.H. Greene. 2012. Econometric analysis. Pearson Education (Chap. 13)
    • J.M. Wooldridge. 2010. Econometric analysis of cross section and panel data. MIT Press (Chap. 8)
    • F. Hayashi. 2000. "Econometrics". Princeton University Press (Chap. 8)
    • J.M. Wooldridge. 2001. "Applications of generalized method of moments estimation". Journal of Economic perspectives 15 (4): 87-100





  • Specific Regression Models




    Modelle für zensierte Daten: Das Lognormal-Hurdle-Modell

    Zensierte Daten sind Daten, die "abgeschnitten" sind. Wenn wir Daten über Arbeitsentgelte erheben, so sind die Daten nicht-negativ, d.h. sie sind ab dem Wert Null abgeschnitten. Das Lognormal-Hurdle-Modell modelliert Auswirkungen von unabhängigen Variablen auf die beobachtete Variable. Es interpretiert den Prozess der Erzeugung von den beobachteten Variablen als zweiteiliges Modell, welches die Hurdle und die lognormalverteilte Größe beinhaltet. Die Hurdle zeigt an, ob die beobachtete Variable zensiert wird, während die lognormalverteilte Größe anzeigt, wie groß die nicht-zensierte Variable ist.

    Einstiegsliteratur: 

    • J.M. Wooldridge. 2010. Econometric analysis of cross section and panel data. MIT Press (Kap. 17)
    • W.H. Greene. 2012. Econometric analysis. Pearson Education (Kap. 19)







  • Multivariate Methods

    Principal Component Analysis

    A principal component analysis (PCA) is concerned with explaining the variance-covariance structure through a few linear combinations of the original variables. Its general objectives are, first, data reduction and second, interpretation. An analysis of principal components often reveals relationships that were not previously suspected and thereby allows interpretations that would not ordinarily result. Therefore, it is extensively used for finding patterns in data of high dimension in the field of finance, data mining, bioinformatics and environmental studies.

    Introductory Literature: 

    • R.A. Johnson, D.W. Wichern u. a. 2002. Applied multivariate statistical analysis. Prentice Hall, NJ (Chap. 8)
    • A.J. Izenman. 2013. "Multivariate regression". In Modern Multivariate Statistical Techniques, 159-194. Springer (Chap. 7)
    • W.J. Krzanowski. 1995. Recent advances in descriptive multivariate analysis. Clarendon Press (Chap. 5)
    • M. Ringnér. 2008. "What is principal component analysis?" Nature biotechnology 26 (3): 303

    Factor Analysis

    In factor analysis, we take multiple observed variables that have similar response patterns. Like the original variables, the factors vary from individual to individual; but unlike the variables, the factors cannot be measured or observed. Each factor captures a certain amount of the overall variance in the observed variables, and the factors are always listed in order of how much variation they explain. The goal of factor analysis is to reduce the redundancy (needlessness) among the variables by using a smaller number of factors. Motivation for factor models, model definition and assumptions as well as the estimation procedure should be covered.

    Introductory Literature: 

    • A.C. Rencher und W.F. Christensen. 2012. Methods of Multivariate Analysis. John Wiley & Sons, Inc. (Chap. 13)
    • J.F. Hair u. a. 2014. Multivariate Data Analysis. Pearson Education Limited (Chap. 3)

    Cluster Analysis

    In cluster analysis we search for patterns in a data set by grouping the (multivariate) observations into clusters. The goal is to find an optimal grouping for which the observations or objects within each cluster are similar, but the clusters are dissimilar to each other. To group the observations into clusters, many techniques begin with similarities between all pairs of observations. In many cases the similarities are based on some measure of distance. Other cluster methods use a preliminary choice for cluster centers or a comparison of within- and between-cluster variability. The techniques of cluster analysis have been extensively applied to data in many fields, such as medicine, psychiatry, sociology, criminology, anthropology, archaeology, geology, geography, remote sensing, market research, economics, and engineering.

    Introductory Literature: 

    • A.C. Rencher und W.F. Christensen. 2012. Methods of Multivariate Analysis. John Wiley & Sons, Inc. (Chap. 15)
    • J.F. Hair u. a. 2014. Multivariate Data Analysis. Pearson Education Limited (Chap. 8)



  • Models of Stochastics





    Extremwertverteilungen

    In vielen Anwendungsgebieten spielt die Modellierung extremer Ereignisse eine besondere Rolle. Mithilfe der Extremwerttheorie kann z.B. das Risiko auf Finanzmärkten oder die Wahrscheinlichkeit für Überflutung eines Deichs abgebildet werden. Eine übliche Herangehensweise ist das Aufteilen des Datensatzes in Blöcke, deren Maxima bestimmten Extremwertverteilungen folgen. Dies sind die Gumbel-, Fréchet- und Weibullverteilung, die in der allgemeinen Extremwertverteilung zusammengefasst werden.

    Einstiegsliteratur: 

    • S. Coles u. a. 2001. An introduction to statistical modeling of extreme values. Springer (Kap. 3)
    • R.-D. Reiss und M. Thomas. 2007. Statistical analysis of extreme values. Springer (Kap. 4)


MASTER THESES

Master Theses in Statistics consists of a description of new statistical methods and their application on specific data. This is similar to bachelor theses. Furthermore, new statistical methods could be described in detail and more critically. Another option is to do an empirical study on a statistical problem.

  • Registration

    For master theses allocation is made on student's requests. If you would like to write your master thesis at our institute, you may contact Prof. Dr. Sibbertsen by email.

  • Topic Assignment

    Topics for master theses are very diverse. They range from methodical work (method presentation, method comparison, method development) to own empirical work (data collection and analysis) with references to all other economic elective courses.

    Topic assignment takes place in coordination with you. We will gladly consider your suggested topics.


NOTES FOR THESES

Below you may find informationen about requirements for bachelor theses as well as a template for LaTeX.

CONTACT FOR GENERAL QUESTIONS ABOUT YOUR THESIS

Laura Bub
Office
Address
Königsworther Platz 1
30167 Hannover
Building
Room
011
Address
Königsworther Platz 1
30167 Hannover
Building
Room
011