Regression modelling for biostatistics II

This unit presents the theory and application of generalised linear models (GLMs) and survival analysis, paying proper attention to the underlying assumptions of these models. 

The unit covers the implementation of GLMs to analyse count data using Poisson and negative binomial regression; how logistic regression models can be applied to binary, multinomial, and ordinal data; and the use of GLMs with continuous data. 

The unit presents methods to analyse time to event survival data including the Kaplan Meier curve, the Cox proportional hazards model, and parametric accelerated failure time models.  A major focus is on selection of appropriate methods, assessing the model fit and diagnostics of GLMs and survival models, and the practical interpretation and communication of model results.