2SLS first-stage Probit - Testing instrumental strength

[Dax Tops]

Dear statalist,

I am using a first-stage probit and use the fitted probabilities from this regression as an instrument in the second-stage. I have an original regression with an endogenous binary independent variable and continuous dependent variable (Y). Furthermore, I add an instrument (Z) as well.

1. first-stage probit

 
 probit X1 Xi-Xn Z i.year i.ffi, vce(robust), where X1 is the endogenous dummy, Xi-Xn are the exogenous variables and Z is the instrument

predict shat, pr

2. Use predicted values from previous step as an instrument and perform second stage

 
 ivregress 2sls Y X1 Xi-Xn (X1 = shat z), vce(robust) first

Since I use both shat (predicted probabilities) and Z as instruments, how can I test the strength of these instruments (F-stat > 10)? Or how can I verify whether I used relevant instruments?

[Phil Bromiley]

Generally, you'll find it better to use the built in routines rather than doing it yourself. I don't see why you buy yourself anything with the probit over just including the same variables as instruments in the second stage. User written ivreg2 has diagnostics that may be helpful that do not appear in ivregress.