Causality

Question 0

Explain how to assess causation

Answer 0

Internal validity:
Eliminate non-causal associations, Bradford Hill
External validity:
Generalisation of results with source population, other populations, your population
Compare results with other studies’ evidence, established biological mechanisms

Question 1

Describe how to distinguish causal from non-causal associations

Answer 1

Bradford Hill Criteria:
Strength of association (strong correlation = more likely)
Specificity of association
Temporal sequence (exposure --> disease)
Consistency of association
Dose response (biological gradient)
Reversibility (what causes what?)
(Weaker arguments - coherence of theory, biological plausibility, analogy)