Skip to main content

Table 3 Overview of the 7 scenarios, including six control scenarios and the seventh scenario which gives the opportunity to test the added value of controlling for the common cause

From: Instrumental variable analysis to estimate treatment effects: a simulation study showing potential benefits of conditioning on hospital

under 7 different scenarios:

ā€ƒ1. Null scenario: no effect of treatment

ā€ƒ2. RCT scenario: treatment randomly assigned

ā€ƒ3. Confounder-adjusted

ā€ƒ4. Confounder-adjusted with instrument

ā€ƒ5. Confounder-adjusted and subject to selection bias

ā€ƒ6. Confounder-adjusted and subject to selection bias with instrument

ā€ƒ7. Confounder-adjusted and subject to selection bias with instrument and common cause of instrument and outcome

With scenario number 7 being what we believe to be closest to the truth.