testing suspected endogenous variables

[btwhow]

what is the best way to test suspected endogenous variables in eviews - so far I have run tests via gmm and 2sls
thanks
m

[startz]

Run TSLS then choose View/IV Diagnostics/Regressor Endogeneity test

[btwhow]

Thanks - its sounds like i am on the right track

the result for the three variables that i tested show as:

"Endogenous variables to treat as exogenous: var1 var2 var3"

does that mean that the three variables that i tested are not endogenous?

m

[startz]

No, that means that's those are the variables being tested.

[btwhow]

Thanks,

i had interpreted that part of the output from the test that I should treat the suspected endogenous variables as exogenous - meaning that they were exogenous and not endogenous.

how do I interpret the results then?

does the answer lie in the remainder of the output from the endogenous test of these variables - ie. the difference in value, degrees of freedom and/or probability, the j-statistic summary, the restricted test equation, or the unrestricted test equation?

m

[startz]

You want to look at the difference in J-stat

[btwhow]

Thanks,
m