This course cluster is focused on basic statistical modeling and consists of 3 courses at a master level: Practical linear models using R, GLM using R, Analysis of Binary Data and log linear models using R. All courses will be avilable online during the acedemic year 2017/2018.

  • Categotical Data Analysis: The course has two parts: modeling binary data and log linear models.

  • Modelling Binary Data using R (See Binary Data).

The course covers topics from the books of David Collett Modelling Binary Data and Alan Agresti An Introduction to Categorical Data Analysis. The course consists of 3-4 classes, each of three hours and combines theory and application using R. Topics covered in the course include:

   Comparing Proportions in Two-by-Two Tables.
   The Relative Risk and the Odds Ratio.
   Chi-Squared Tests of Independence.
   Analysis of I X J tables.
   Logistic Regression: model formulation, interpretion, estimation and inference.
   Model Selection.

  • Log linear models using R

  • Applied Generalized Linear Models (GLM) using R (See GLM).

The course covers topics discussed in

  • Collet D (1994): Modeling Binary data.
  • Lindsey (1997): Applying generalized linear models.
  • Dobson (2002): An introduction to generalized linear models.
  • McCillagh and Nelder (1983): Generalized linear models (first edition).

Topics include in the course:

Linear  models with normal error.
Generalized linear models.
Exponential Family
Generalized linear model function in R
Models for Binary data.
Estimation, confidence intervals and Inference.
Model Selection and Model diagnostic.
Poisson regression and log linear models.
Beyond Poisson and binomial distributions: models with different link functions and/or       
distributions.
Over dispersion.
  • Practical linear models using R