Hi everyone!

I working on a problem using linear IV models. The main specification has one endogenous variable e one interaction of the endogenous variable and exogenous variables. I am using one excluded instrument and interactions of exogenous variables as IVs. For inference, I used ivreg2 and twostepweakiv, but the results are conflicting: ivreg2 gives an interval including 0 and twostepweakiv lc2 interval doesn´t. How should I choose in this case?
Thank you!