This advanced module in Prediction Modelling & Joint Modelling of Longitudinal and Survival Data Analysis provides students with the tools to unlock the full potential of longitudinal and time-to-event (survival) data—critical in healthcare research and personalised medicine. You will gain a deep understanding of how to build and validate clinical prediction models, and how to jointly (or simultaneously) model repeated measurements over time and patient outcomes, accounting for dropouts and dynamic risk.
Through formal lectures, real-world case studies, and interactive computer practicals in R, you will learn to develop and interpret robust prediction models. You will also explore state-of-the-art joint modelling techniques to investigate complex relationships between biomarkers and health events, a skill increasingly in demand across healthcare research and industry.
Assessments are designed to mirror real-world analysis and reporting. In assessment 1, you will create and validate a clinical prediction model using R and present it in a concise, reproducible R Markdown report suitable for clinical use (50% of the final grade). Assessment 2 requires a detailed analysis of a dataset using joint modelling, presented as a journal-style research article complete with lay summary, visuals, and practical implications (50%).
You will also benefit from guest lectures by domain experts and apply your learning using real datasets, combining statistical rigour with practical problem-solving. This module is ideal for students interested in clinical research, data science, epidemiology, or any field where prediction and time-dependent data analysis are essential.
Gain confidence, fluency in R, and the statistical insight needed to handle today’s most complex health data challenges—and translate them into meaningful clinical knowledge.