The course starts with the basic applications of biostatistics methodology for the analysis of data from medical and public health research. The following topics are given attention: data types, measures of location and spread, population and samples, distributions, confidence intervals, hypothesis testing, comparison of two or more proportions (parametric and non-parametric methods), relationships between two variables (correlation, single linear regression, logistic regression, Poisson regression, time-to-event regression).
This module handles the statistical methods needed to study the association between (various) determinants and the occurrence of a certain event.
Considerable attention is given to multiple linear regression. Likelihood theory and maximum likelihood estimation is introduced, using examples and keeping mathematical derivations to a minimum. Subsequently, the most important regression techniques that are applied in biomedical and related research, are sketched and exemplified. This refers to: logistic regression (including model validation and regression diagnostics), Poisson regression, analysis of event history data, with specific attention for the Cox proportional hazards model. Finally, students are introduced to the analysis of longitudinal data. Attention is given to missing data.