Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: IVPROBIT and marginal effects
From
Shikha Sinha <[email protected]>
To
[email protected]
Subject
Re: st: IVPROBIT and marginal effects
Date
Wed, 8 Aug 2012 18:49:42 -0700
For example, below I report the outputs from -ivreg2, -ivprobit,
margins, and margins with predict(pr)
-ivprobit coeff is exactly similar to margins,dydx(dist_min), however,
margins with predicted prob is different.
My question is which coeff (-0.0273112 vs -.0774099) is comparable
to -ivreg2 coeff of -0.02773.
Thanks,
Shikha
ivreg2 ifd $transport (dist_min=distindex)
IV (2SLS) estimation
--------------------
Estimates efficient for homoskedasticity only
Statistics consistent for homoskedasticity only
Number of obs = 163659
F( 3,163655) = 1474.76
Prob > F = 0.0000
Total (centered) SS = 37846.80961 Centered R2 = -0.0220
Total (uncentered) SS = 59422 Uncentered R2 = 0.3491
Residual SS = 38677.86458 Root MSE = .4861
------------------------------------------------------------------------------
ifd | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
dist_min | -.02773 .0006473 -42.84 0.000 -.0289987 -.0264612
allwtd | .0521262 .0036099 14.44 0.000 .0450509 .0592015
carmoto | .3531489 .0100414 35.17 0.000 .3334681 .3728297
_cons | .4503572 .0053071 84.86 0.000 .4399556 .4607589
------------------------------------------------------------------------------
Underidentification test (Anderson canon. corr. LM statistic): 1.5e+04
Chi-sq(1) P-val = 0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic): 1.7e+04
Stock-Yogo weak ID test critical values: 10% maximal IV size 16.38
15% maximal IV size 8.96
20% maximal IV size 6.66
25% maximal IV size 5.53
Source: Stock-Yogo (2005). Reproduced by permission.
------------------------------------------------------------------------------
Sargan statistic (overidentification test of all instruments): 0.000
(equation exactly identified)
------------------------------------------------------------------------------
Instrumented: dist_min
Included instruments: allwtd carmoto
Excluded instruments: distindex
------------------------------------------------------------------------------
. ivprobit ifd $transport (dist_min=distindex), nolog asis
Probit model with endogenous regressors Number of obs = 163659
Wald chi2(3) = 5631.07
Log likelihood = -622595.14 Prob > chi2 = 0.0000
------------------------------------------------------------------------------
| Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
dist_min | -.0774099 .0015262 -50.72 0.000 -.0804012 -.0744186
allwtd | .1422951 .0098945 14.38 0.000 .1229024 .1616879
carmoto | .8970391 .0280716 31.96 0.000 .8420199 .9520583
_cons | -.0949859 .014482 -6.56 0.000 -.1233702 -.0666016
-------------+----------------------------------------------------------------
/athrho | .329658 .0105356 31.29 0.000 .3090087 .3503074
/lnsigma | 1.748964 .0017479 1000.61 0.000 1.745538 1.752389
-------------+----------------------------------------------------------------
rho | .3182135 .0094687 .299535 .3366481
sigma | 5.748641 .010048 5.728981 5.768369
------------------------------------------------------------------------------
Instrumented: dist_min
Instruments: allwtd carmoto distindex
------------------------------------------------------------------------------
Wald test of exogeneity (/athrho = 0): chi2(1) = 979.06 Prob > chi2 = 0.0000
. margins, dydx(dist_min) pred(pr)
Average marginal effects Number of obs = 163659
Model VCE : OIM
Expression : Probability of positive outcome, predict(pr)
dy/dx w.r.t. : dist_min
------------------------------------------------------------------------------
| Delta-method
| dy/dx Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
dist_min | -.0273112 .0005023 -54.37 0.000 -.0282957 -.0263266
------------------------------------------------------------------------------
. margins, dydx(dist_min)
Average marginal effects Number of obs = 163659
Model VCE : OIM
Expression : Fitted values, predict()
dy/dx w.r.t. : dist_min
------------------------------------------------------------------------------
| Delta-method
| dy/dx Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
dist_min | -.0774099 .0015262 -50.72 0.000 -.0804012 -.0744186
------------------------------------------------------------------------------
On Wed, Aug 8, 2012 at 6:09 PM, Shikha Sinha <[email protected]> wrote:
> Hi all,
>
> Are -ivprobit coefficients marginal effects?
>
> or will have to use -margins, dydx(_all) after -ivprobit to estimate
> the marginal effect?
>
> Thanks,
> Shikha
> *
> * 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/
*
* 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/