Criteria of causation

From the book The Art of Statistics, David Spiegelhalter relates 7 criteria to be considered when determining causation.

The criteria were created by Austin Bradford Hill, then further developed by Jeremy Howick & co.

The criteria

1. The size of the effect is so large that it cannot be explained by confounding
2. There is appropriate temporal and/or spatial proximity
   • Effect follows cause
   • Effect occurs within a reasonable amount of time after the cause
3. Dose responsiveness and reversibility
   • The effect increases as the dosage increases
   • The effect decreases as the dosage decreases
4. Plausible mechanism of action
   • Effect conceptually relates to cause
   • Evidence of a causal chain
5. The effect fits with what is known already
6. The effect is found when the study is replicated
7. The effect is found in similar, but not identical, studies

3 categories for the criteria

1. Direct evidence
   • 1-3
2. Mechanical evidence
   • 4
3. Parallel evidence
   • 5-7

Application

Using the criteria, we can assess causal effects even without randomized control trials (RCTs). The criteria provide a useful framework for understanding how we can look at causality and make conclusions even without perfect evidence.