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]
st: RE: logit or probit : why one row is missing S.E., z-score, and confidence interval??
From
Nick Cox <[email protected]>
To
"'[email protected]'" <[email protected]>
Subject
st: RE: logit or probit : why one row is missing S.E., z-score, and confidence interval??
Date
Wed, 26 Jan 2011 12:10:43 +0000
You've got one clue to difficulties:
"Note: 9 failures and 0 successes completely determined."
Thus see how the data look in terms of cross-combinations of -depvar-, -varI-, -varR-.
I don't see that your regression results indicate lack of multicollinearity: you'd need to consider that more directly.
Nick
[email protected]
Hong, Sounman
All my variables (dependent and independent) are binary. I ran a regression and I found there is no multicollinearity. Everything looked good.
But if I run a logit or probit then, as you can see below, some stats (standard error, z-score) do not show for one variable.
I wonder why??? Can someone help me, please?
. reg depvar varI varR varW varAA varRvarW varIvarR, r
Linear regression Number of obs = 102
F( 6, 95) = 4.96
Prob > F = 0.0002
R-squared = 0.1885
Root MSE = .39009
------------------------------------------------------------------------------
| Robust
depvar | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
varI | .4132157 .1222138 3.38 0.001 .1705906 .6558407
varR | .5424289 .1517043 3.58 0.001 .2412578 .8436
varW | .1506902 .1567265 0.96 0.339 -.1604513 .4618317
varAA | .2461644 .0670512 3.67 0.000 .1130509 .3792779
varRvarW | -.3753536 .1827751 -2.05 0.043 -.7382082 -.0124991
varIvarR | -.5573226 .1643252 -3.39 0.001 -.8835492 -.2310959
_cons | -.2919213 .1258739 -2.32 0.023 -.5418126 -.0420301
------------------------------------------------------------------------------
. probit depvar varI varR varW varAA varRvarW varIvarR, r
Probit regression Number of obs = 102
Wald chi2(5) = .
Prob > chi2 = .
Log pseudolikelihood = -41.68023 Pseudo R2 = 0.2344
------------------------------------------------------------------------------
| Robust
depvar | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
varI | 6.200518 . . . . .
varR | 6.766806 .4989077 13.56 0.000 5.788965 7.744647
varW | .4070662 .6244081 0.65 0.514 -.8167511 1.630884
varAA | 1.436624 .5493662 2.62 0.009 .3598866 2.513362
varRvarW | -1.470836 .7256594 -2.03 0.043 -2.893103 -.0485701
varIvarR | -6.964342 .3732943 -18.66 0.000 -7.695985 -6.232699
_cons | -7.856409 .6625587 -11.86 0.000 -9.155 -6.557818
------------------------------------------------------------------------------
Note: 9 failures and 0 successes completely determined.
. logit depvar varI varR varW varAA varRvarW varIvarR, or
Logistic regression Number of obs = 102
LR chi2(6) = 25.21
Prob > chi2 = 0.0003
Log likelihood = -41.837978 Pseudo R2 = 0.2315
------------------------------------------------------------------------------
depvar | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
varI | 9.48e+07 8.16e+07 21.35 0.000 1.76e+07 5.12e+08
varR | 2.69e+08 . . . . .
varW | 2.136371 2.232017 0.73 0.467 .2756554 16.5572
varAA | 14.17273 15.6 2.41 0.016 1.638821 122.5676
varRvarW | .0790781 .1006545 -1.99 0.046 .0065254 .9583132
varIvarR | 2.99e-09 2.16e-09 -27.10 0.000 7.22e-10 1.24e-08
------------------------------------------------------------------------------
Note: 9 failures and 0 successes completely determined.
*
* 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/