[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
The overid routine of Baum, Schaffer, Stillman and Wiggins generates
"Sargan" tests of overidentifying restrictions after an instrumental
variables estimator has been applied to an overidentified equation.
Earlier versions of this routine could be used after -ivreg- or -
ivreg2-, but given that -ivreg2- routinely calculates the Sargan(-
Hansen) statistic, their utility was somewhat limited.
overid has now been extended to compute tests of overidentifying
restrictions for three additional IV estimators:
* ivprobit using the twostep (non-ML) option
* ivtobit using the twostep option
* reg3, with and without linear constraints (following up on one of
my Statalist postings)
The new routine uses Mata, thus requires Stata 9.2. It is available
from SSC or via -adoupdate if you have previously installed -overid-.
Kit Baum, Boston College Economics
An Introduction to Modern Econometrics Using Stata:
* For searches and help try: