Generalised Linear Models

Module code: MA3201

This module builds upon the principles of linear modelling to introduce a more comprehensive framework for data analysis: generalised linear models (GLMs). By integrating link functions and flexible error structures, GLMs extend the applicability of linear models to a wide range of data types, including count data and binary outcomes. Through this unified approach, you will learn to perform log-linear model and Poisson regression for count data, and logistic regression for binary outcomes.

The module emphasises two primary objectives: identifying key explanatory variables and understanding their relationships with the response variable. You will also gain skills to compute confidence intervals, conduct hypothesis testing, and assess model validity using deviance as a diagnostic tool. Hands-on training with R software ensures you will be able to apply these techniques effectively to real-world datasets.

By the end of this module, you will be able to independently fit and evaluate GLMs, critically assess model assumptions, and communicate findings. With its balance of theory, applications, and computational tools, this module equips you with valuable skills for advanced statistical modelling and data-driven decision-making in research and industry.

Topics covered

  • Analysing contingency tables
  • Log-linear models for analysing contingency tables
  • The general theory of generalised linear models
  • Poisson regression models
  • Linear logistic regression models
  • Statistical inferences and model selections
  • Fitting models using R programming
Back to top
MENU