The difficulty of assessment of causation can involve separating direct effects of an exposure from intermediate or indirect effects of that exposure. If the exposure is randomized, direct or indirect effects are not separately identifiable. If the exposure is not randomized but is controlled, direct and indirect effects can be assessed without bias

Robins JM; Greenland S

The difficulty of assessment of causation can involve separating direct effects of an exposure from intermediate or indirect effects of that exposure. If the exposure is randomized, direct or indirect effects are not separately identifiable. If the exposure is not randomized but is controlled, direct and indirect effects can be assessed without bias

Robins JM, Greenland S1992Identifiability and exchangeability for direct and indirect effects. Epidemiology

Summary

No abstract is available for this paper. It was published in Identifiability and exchangeability for direct and indirect effects. Epidemiology in 1992 by Robins JM and colleagues.

Part of CIRS Nexus · Curated by Lauren Lowenstein, The Biotoxin Lady