causation

Last reviewed 01/2018

Causation - that is, causal linkage between two factors - is suggested by evidence which is:

  • consistent and strong
  • biologically plausible
  • showing a dose / response relationship
  • temporally correct
  • coherent
  • specific

The strength of evidence is probably most strong from randomised clinical trials, and progressively less strong from cohort studies, case control studies and case series.

Note that causation is definitely distinct from association.