Module MA2600-KP07

Duration

1 Semester

Turnus of offer

each summer semester

Credit points

7

Course of studies, specific fields and terms:

• Bachelor CLS 2023, compulsory, mathematics
• Bachelor CLS 2016, compulsory, mathematics

Classes and lectures:

• Biostatistics 2 (practical course, 2 SWS)
• Biostatistics 2 (exercise, 1 SWS)
• Biostatistics 2 (lecture, 2 SWS)

Workload:

• 40 hours private studies
• 15 hours exam preparation
• 85 hours programming
• 70 hours in-classroom work

Contents of teaching:

• Assumptions in the classical linear model
• Last squares method and geometric representation
• Stochastic properties, testing the general linear hypothesis, construction of confidence intervals and confidence ellipsoids
• Regression diagnostics and model choice
• Logistic regression: basics, model specification, threshold model, maximum likelihood estimation, tests and confidence intervals
• Survival Analysis: Kaplan-Meier curves, Log-Rank test, assumptions and parameter estimation in Cox regression
• Data structures in R, functions and functionals in R
• Statistical analysis in R: descriptive statistics (frequency tables, metrics), graphical representation, statistical tests (t-, X2-, U-, Log-Rank-), executable protocolls (literate programming) with knitr, bootstrapping, cross-validation, linear regression, logistic regression, Cox regression

Qualification-goals/Competencies:

• The students are able to enumerate and explain the assumptions of the classical linear model.
• They are able to describe typical applications of the classical linear model.
• They are able to list the differences between the linear model and the logistic regression model.
• They are able to describe possible error sources in modelling the linear model.
• They are able to calculate the estimators (point and interval estimators, residual, prediction) in the linear model by hand.
• They are able to evaluate the graphics for regression diagnostics in the linear model.
• They are able to interpret the results of studies, where a linear, a logistic or a Cox regression model was applied.
• They are able to draw and interpret Kaplan-Meier curves.
• They are able to perform data transformations.
• They are able to program their own R functions.
• They are able to present data by suitable and pleasing graphics.
• They are able to conduct linear, logistic and Cox regression analysis by means of R packages and to evaluate the results on the computer.
• They are able to execute statistical tests (t-, X2-, U-, Log-Rank-) in R, to formulate the hypotheses and to make a test decision.
• They are able to illustrate the principle of bootstrapping and cross-validation and to implement it in R.
• They are able to create a report that meets the requirements of academic work by means of the R package knitr.

Grading through:

• written exam

Is requisite for:

• Genetic Epidemiology 2 (MA4661-KP08, MA4661)
• Interdisciplinary Seminar (MA3300-KP04)
• Generalized Linear Models (MA4962-KP05)
• Multivariate Statistics (MA4944-KP05)

Requires:

• Introduction to Programming (CS1000-KP08, CS1000SJ14-MML/MI, CS1000SJ14-MIW)
• Biostatistics 1 (MA1600-KP04, MA1600, MA1600-MML)

Responsible for this module:

• Prof. Dr. rer. biol. hum. Inke König

Teacher:

• Institute of Medical Biometry and Statistics
• Dr. rer. hum. biol. Björn-Hergen Laabs
• MitarbeiterInnen des Instituts

Literature:

• Fahrmeir, Ludwig; Kneib, Thomas; Lang, Stefan (2009) : Regression: Modelle, Methoden und Anwendungen Springer: Heidelberg
• Dobson, Annette J & Barnett, Adrian (2008) : An Introduction to Generalized Linear Models, 3rd ed. Chapman & Hall/CRC: Boca Raton
• Sachs, Lothar; Hedderich, Jürgen : Angewandte Statistik: Methodensammlung mit R 15. Auflage, Springer: Heidelberg
• Ligges, Uwe : Programmieren mit R 3. Auflage, Springer: Heidelberg

Language:

• offered only in German

Notes:

Last Updated:

07.02.2023