Advanced Biostatistics for Epidemiology

Module Aims:
This module builds on the fundamentals of biostatistics from a relevant pre-requisite module, moving to more advanced considerations around when and how to apply statistical techniques with a focus on common themes of validity and interpretation. The module will consider some theoretical aspects of statistical methods but the focus will be on illustrative examples of various statistical techniques applied to different contexts (i.e. different research problems across topics from population health sciences).

Module Learning Outcomes:

By the end of the module, students should be able to:

1. Describe, implement, and interpret a range of fundamental/established techniques in statistical epidemiology in the context of population health studies.
2. Critically assess, determine, and justify the relevant statistical technique that can appropriately answer a specific research question.
3. Describe (though not necessarily implement) a range of extensions of fundamental techniques in order to account for complexities within a study (either by design or after data collection).
4. Express analysis findings to non-specialists, conveying the evidence and potential caveats.

Pre-requisites:

A module covering the fundamentals of biostatistics (i.e. Statistics for HDS or Principles of Biostatistics), data handling and analysis in GNU R (Applied Data Analysis), and how to formulate research questions and present findings (Research Skills). In particular:

• An understanding of classical hypothesis and significance testing and how to implement various statistical tests.
• An understanding of regression modelling and how to implement and interpret a regression model.
• Familiarity with the basic concepts of study design and sample size calculations.
• Familiarity with the GNU R statistical software.
• Familiarity with the types of research questions of interest in population health science.
• Ability to present findings via written reports and presentations.

Teaching Strategy:

The module will consist of lectures covering the main material, a mixture of theoretical and high-level aspects combined with examples taken from relevant contexts. The lecture material will be further developed in problem classes, specifically how to apply the methods in GNU R and developing skills to critically assess the relevant output.

Assessment:

The module will be assessed via a written report and presentation based on students completing a practical assignment. The assessment will involve students being given a (unique) dataset and associated research question to address. They will address this question individually. The students will be given a dedicated day with access to computing facilities to conduct their analysis (which will not be under exam conditions), then they will write a report and create a presentation. The presentation (20 mins) will focus on interpretation and translation of statistical aspects into context. The presentations will be given to at least two assessors and will be marked in terms of clarity of presentation, statistical content, and interpretation of key findings to a lay audience. The report will be short and focus on technical detail and explanation of statistical aspects of analysis. It is a companion to the presentation and will be assessed on clarity and statistical rigour.

Module Length: 9 days (double module)