Dear all,
I am estimating an IV-oprobit using cmp. How can I test the exdogeneity and relevance of my instrumental variables? My situation is complicated as 1) I have two endogenous variables, both of which are dummies 2) I also used clustered standard errors 3) I use cmp instead of 2sls. The main equation for Y is ordered probit, and two equations for endogenous variables are probit.
cmp (Y=X1+X2+X3) (X1=Z1+Z2+X3) (X2=Z1+Z2+X3), ind ($cmp_oprobit $cmp_probit $cmp_probit) vce(cluster school)
Y_ordered categorical variable;
X1, X2_endogenous variables, dummies;
Z1_IV for X1;
Z2_IV for X2;
X3_exdogenous variables
In this situation, how could I test the exdogeneity (not related to error terms) and relevance (related to endogenous variable) of IVs?
Thank you very much!
I am estimating an IV-oprobit using cmp. How can I test the exdogeneity and relevance of my instrumental variables? My situation is complicated as 1) I have two endogenous variables, both of which are dummies 2) I also used clustered standard errors 3) I use cmp instead of 2sls. The main equation for Y is ordered probit, and two equations for endogenous variables are probit.
cmp (Y=X1+X2+X3) (X1=Z1+Z2+X3) (X2=Z1+Z2+X3), ind ($cmp_oprobit $cmp_probit $cmp_probit) vce(cluster school)
Y_ordered categorical variable;
X1, X2_endogenous variables, dummies;
Z1_IV for X1;
Z2_IV for X2;
X3_exdogenous variables
In this situation, how could I test the exdogeneity (not related to error terms) and relevance (related to endogenous variable) of IVs?
Thank you very much!
