Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: tests of over-identifying restrctions after Heckman


From   "Evren Ceritoglu" <[email protected]>
To   <[email protected]>
Subject   st: tests of over-identifying restrctions after Heckman
Date   Wed, 29 Jul 2009 18:59:20 +0300

Dear Statalist,

I want to perform over-identifying restrictions (overid) test after Heckman
two-step selection model. I must introduce at least one exogeneous variable
in the probit equation, which becomes an instrument, for the Heckman model
to hold. However, this process also reqires an overid test after the
estimation to check whether the instruments are valid or not.

OLS: lny1 = X1b1 + X2b2 + u
Probit: d = X1b1 + Zb3 + e

(d: dummy variable, which is the selection criterion)
(Z: set of instruments, composed of dummy variables)

The overid command is useful after ivprobit or ivtobit, but it is not
possible to use it after the Heckman model.

I tried to perform a LR test, but there was a problem! I stored the
predictions (estimation results) from the overidentified regression
(heckoverid) and also the predictions (estimation results) from the
just-identified regression (heckjustid). However, stata says that heckoverid
does not contain scalar e(ll), which I don't understand in both cases.

I) over-identified model

OLS: lny1 = X1b1 + X2b2 + u
Probit: d = X1b1 + Zb3 + e

II) just-identified model

OLS: lny1 = X1b1 + X2b2 + Zb3 + u
Probit: d = X1b1 + Zb3 + e

III) LR test

lrtest heckoverid heckjustid, df(1) (assuming that there is only one
instrument)

Is there a feasible way to perfrom over-identifying restrictions test after
Heckman two-step selection model?


Best regards,

Evren Ceritoglu

Researcher
Research Dept.
Central Bank of Turkey
Ulus / Ankara



==========================================================-
Bu e-posta sadece yukarida isimleri belirtilen kisiler arasinda ozel haberlesme amacini tasimaktadir. Size yanlislikla ulasmissa lutfen gonderen kisiyi bilgilendiriniz ve mesaji sisteminizden siliniz. Turkiye Cumhuriyet Merkez Bankasi A.S. bu mesajin icerigi ile ilgili olarak hicbir hukuksal sorumlulugu kabul etmez.

This e-mail communication is intended for the private use of the people named above. If you received this message in error, please immediately notify the sender and delete it from your system. The Central Bank of The Republic of Turkey does not accept legal responsibility for the contents of this message.

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index