Share
Explore BrainMass

Significance of regression coefficient in Probit analysis

Test of significance for binary regression models:

I have estimated both probit and logit regression models, and I want to check to see if a particular result is significant. Specifically speaking I want to test to determine whether or not V9 and V31 are significant. Please show me how to conduct such a test. Here is a sample of my output:

Probit regression Number of obs = 13421
LR chi2(8) = 1255.28
Prob > chi2 = 0.0000
Log likelihood = -6646.2534 Pseudo R2 = 0.0863

------------------------------------------------------------------------------
empst31 | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
v2 | -.1887461 .0357127 -5.29 0.000 -.2587417 -.1187505
v3 | -.1388738 .0306249 -4.53 0.000 -.1988976 -.07885
v5 | -.0056533 .0014383 -3.93 0.000 -.0084723 -.0028342
v9 | -.0139441 .0129414 -1.08 0.281 -.0393088 .0114206
v31 | .0984975 .0334974 2.94 0.003 .0328438 .1641511
_cons | .0709432 .0757518 0.94 0.349 -.0775275 .219414

Solution Summary

Significance of regression coefficient in Probit analysis.

$2.19