Wouldn't conditioning on Z=4-year mortality violate one of the structural causal model assumptions needed to adjust for biases? The four-year satisfaction is the outcome variable and conditioning on the outcome would introduce biases in the adjustment. (at time point 19:28)
Wouldn't conditioning on Z=4-year mortality violate one of the structural causal model assumptions needed to adjust for biases? The four-year satisfaction is the outcome variable and conditioning on the outcome would introduce biases in the adjustment. (at time point 19:28)