Causal Inference in Observational Studies

An association is not a cause, and knowing the difference is what makes a researcher credible. This level teaches you to interrogate any finding through four rival explanations before you dare claim a cause. It is the intellectual backbone of the entire course.

Who this level is for

Anyone who will interpret observational data, researchers, clinicians, analysts and evidence users at any level. Recommended after the observational-design levels (2-4).

What you'll learn

• Why measures like risk ratios and odds ratios never prove causation on their own
• The four rival explanations for any association: chance, bias, reverse causation and confounding
• An intuitive grasp of chance, error and p-values (the maths of power lives in the Sample Size course)
• How specific design choices fight bias and confounding
• How to judge causality honestly and choose stronger designs

You finish with: Confounders, likely biases & a design justification for your question.