Which of the following is a Bradford Hill criterion for causality?

• A. Temporal relationship
• B. P-value threshold
• C. Randomization
• D. Standard error

Answer: (A)

The main idea is temporality: for something to cause an outcome, the exposure must occur before the outcome develops. This timing is essential because without it, you can’t establish a causal direction or rule out reverse causation. In the Bradford Hill framework, temporality is a necessary condition for causal inference in observational data.

The other options don’t serve as causality criteria. A p-value threshold speaks to statistical significance, not whether exposure truly causes the outcome. Randomization is a study design feature that helps reduce bias and confounding, not a criterion used to judge causality after the fact. Standard error measures precision of an estimate, not causal relation.