Microcredential: Data Analysis for Precision Health
July 28 @ 2:00 pm – July 29 @ 5:00 pm
As we enter the data revolution, the scale and accessibility of health and medical data has reached unprecedented levels creating a growing need for expertise in extracting insights from this data.
This course will provide participants with essential statistical skills to analyse and interpret health and medical data. Key topics include linear models, mixed effect models, logistic regression and survival analysis.
Real health and medical data will be utilised to explore common challenges, practical workarounds, and translate data into actionable insights.
By the end of this course, you will be able to:
- formulate and interpret appropriate linear models to describe the relationships between multiple factors
- train and evaluate logistic regression models for binary data
- understand and apply linear mixed effect models for data with repeated measures
- visualise survival data with Kaplan-Meier curves and perform inference with Cox proportional hazards models.
Aims
This course will give participants the necessary understanding and skills to perform statistical analyses on health and medical data.
Participants will gain experience in working with real data and develop critical thinking skills to address common challenges.
Participants will learn to communicate their data and findings through graphical and statistical summaries.
Content
The course covers four main topics:
- Linear models: Fit, refine, interpret and visualise regression models. We will discuss making predictions, evaluating model fit and feature selection.
- Logistic regression: Interpret odds ratios to evaluate risk factors, assess model performance when working with binary health outcomes.
- Mixed effect models: Analyse repeated measures and hierarchical data to understand individual and group-level patterns in health and medical contexts.
- Survival analysis: Model time-to-event data using Kaplan-Meier curves and Cox proportional hazards models, and assess model accuracy with metrics like the C-index.