Causation Math Example 4
Follow the full solution, then compare it with the other examples linked below.
Example 4
hardA pharmaceutical company's observational study finds Drug X correlates with recovery. Explain three criteria needed to claim Drug X causes recovery, and why a randomized trial is required.
Solution
- 1 Criterion 1: Correlation — Drug X and recovery must be statistically associated (already shown)
- 2 Criterion 2: Temporal order — Drug X must be administered before recovery occurs
- 3 Criterion 3: Elimination of confounders — people taking Drug X might be healthier overall, have better healthcare access, etc.
- 4 Why randomized trial: randomly assigns Drug X vs placebo, balancing all confounders (known and unknown) between groups, allowing causal conclusions
Answer
Need correlation + temporal order + no confounders. Only a randomized trial eliminates confounders definitively.
The 'gold standard' for causal inference is the Randomized Controlled Trial (RCT). Randomization uniquely eliminates all confounders in expectation — this is impossible in observational studies no matter how many confounders you control for statistically.
About Causation
Causation exists when one variable directly produces or influences a change in another variable — distinct from mere correlation or association.
Learn more about Causation →More Causation Examples
Example 1 medium
Ice cream sales and drowning deaths are positively correlated. Explain why correlation does not impl
Example 2 hardA study finds students who eat breakfast score higher on tests. Design an argument for why this migh
Example 3 easyCountries with more TV sets per capita have higher life expectancy. Does this mean buying TVs causes