This unit covers statistical analysis techniques that are commonly required for analysing data that arise from clinical or epidemiological studies. Students will gain hands on experience applying model-building strategies and fitting advanced statistical models. In particular, students will learn a statistical software package called Stata, how to handle non-linear continuous variables, and how to analyse correlated data. Correlated data arise from clustered or longitudinal study designs, such as, cross-over studies, matched case-control studies, cluster randomised trials and studies involving repeated measurements. Statistical models that will be covered include fixed effects models, marginal models using Generalised Estimating Equations (GEE), and mixed effects models (also known as hierarchical or multilevel models). This unit of study focuses on data analyses using Stata and the interpretation of results.
2hr lec/tut/week x 12 weeks, also offered fully online.
2 x data analysis report (2x50%)
No mandatory text books. Course notes are provided.