an overidentification test after heckprobit with weights

[osnat israeli]

hello,
I have run a heckprobit model with two instruments in the selection regression. I want to check the exclusion restrictions, but because I'm using weights, I can't use the lrtest command after the estimation. Could anyone suggest an alternative overidentification test procedure?
Thanks
Osnat

[Jeff Wooldridge]

A simple test is to include one of to two IVs in the outcome model and use a t test. The standard errors reported will account for the weighting.

[osnat israeli]

Thank you very much! I truly appreciate your willingness to help.