Dear Statalist:
I am running Stcox and would really appreciate any help that someone what me able to provide. How does one interpret missing standard errors, misin z scores etc. When I run it, several of the variables, (e.g., "lang") come back with a hazard ratio but nothing else. I checked the Stata FAQ response. However, it is unclear to me if it best practice to leave those variables out. Are the results credible with some of the variables coming out like this. Also, many of the P values are coming back as 1.000. How does one interpret this result? Can I use/trust the overall results and the results of other variables?
Cox regression --
Breslow method for ties Number of obs = 22667
Gamma shared frailty Number of groups = 103
Group variable: company
No. of subjects = 2163 Obs per group: min = 194
No. of failures = 71 avg = 220.068
Time at risk = 7739538 max = 395
Wald chi2(46) = 6209.90
Log likelihood = -313.29552 Prob > chi2 = 0.0000
------------------------------------------------------------------------------
_t | Haz. Ratio Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
pstock | .9992232 .0121881 -0.06 0.949 .9756183 1.023399
mstock | 1.0124 .0179597 0.69 0.487 .9778044 1.048219
expstock | 1.094391 .1897217 0.52 0.603 .7791295 1.537218
grupo2 | 4.231667 2.718451 2.25 0.025 1.201417 14.9049
priv | 2.162367 1.563238 1.07 0.286 .524289 8.918421
alliance | 2.812804 1.462422 1.99 0.047 1.015281 7.79279
own2 | 1.310643 1.584542 0.22 0.823 .1225741 14.01427
lsales | 1.371852 .3741872 1.16 0.246 .8037721 2.341432
nf | .9062903 .105697 -0.84 0.399 .7210994 1.139041
pre91 | 4.332353 1.997612 3.18 0.001 1.754857 10.69562
age | .9949293 .0094322 -0.54 0.592 .9766132 1.013589
sic1 | 1.859124 2.472855 0.47 0.641 .1371227 25.20619
sic2 | 4.299693 4.926218 1.27 0.203 .4552019 40.61353
sic3 | 2.843917 3.657909 0.81 0.416 .2286034 35.37945
gdpcaph | 1.000517 .0005426 0.95 0.340 .9994545 1.001581
gdpdiff | 1 3.27e-07 0.25 0.802 .9999994 1.000001
lang | 1.17e+09 . . . . .
lndist | 1.38712 3.576611 0.13 0.899 .008858 217.215
agree | .9933545 .6794694 -0.01 0.992 .2599379 3.796111
border | 4.169336 7.285759 0.82 0.414 .1357139 128.0884
lninfdi | 1.272696 .5410447 0.57 0.571 .5531771 2.928093
lnexport | 1.226262 .5871333 0.43 0.670 .4797653 3.134281
coun1 | 3.22e-08 .7510258 -0.00 1.000 0 .
coun2 | 3.94e+07 1.46e+08 4.72 0.000 27456.55 5.66e+10
coun3 | 3408175 . . . . .
coun4 | 2.17e+08 8.30e+08 5.02 0.000 120135.3 3.92e+11
coun6 | 6.54e+07 2.00e+08 5.87 0.000 160486 2.66e+10
coun7 | 5.03e-07 21.54161 -0.00 1.000 0 .
tarcou1 | 4.11e-18 1.44e-09 -0.00 1.000 0 .
tarcou2 | .8961112 2.568654 -0.04 0.969 .0032542 246.7658
tarcou3 | .302835 .835457 -0.43 0.665 .0013581 67.52881
tarcou4 | .3512424 .8451288 -0.43 0.664 .0031443 39.2368
tarcou5 | 5.08e-18 1.86e-09 -0.00 1.000 0 .
tarcou6 | 7.41e-18 2.79e-09 -0.00 1.000 0 .
tarcou7 | .5215511 1.251615 -0.27 0.786 .004727 57.54455
tarcou8 | .2315671 .6427914 -0.53 0.598 .0010043 53.39115
tarcou9 | 6.90e-18 2.70e-09 -0.00 1.000 0 .
tarcou11 | 6.73e-18 2.16e-09 -0.00 1.000 0 .
tarcou12 | 3.20e-18 1.17e-09 -0.00 1.000 0 .
tarcou13 | 2.44e-18 8.51e-10 -0.00 1.000 0 .
tarcou15 | 2.15e-18 8.31e-10 -0.00 1.000 0 .
tarcou16 | 7.80e-18 2.98e-09 -0.00 1.000 0 .
tarcou17 | 1.043378 2.932431 0.02 0.988 .0042279 257.4887
tarcou18 | 4.31e-09 1.06e-08 -7.84 0.000 3.50e-11 5.32e-07
tarcou19 | 3.93e-09 9.88e-09 -7.69 0.000 2.84e-11 5.44e-07
tarcou20 | 3.63e-18 1.37e-09 -0.00 1.000 0 .
tarcou21 | .4769469 1.298323 -0.27 0.786 .0022981 98.98547
tarcou22 | .234499 .5961574 -0.57 0.568 .0016076 34.20708
-------------+----------------------------------------------------------------
theta | .4990421 .4101498
------------------------------------------------------------------------------
Likelihood-ratio test of theta=0: chibar2(01) = 3.58 Prob>=chibar2 = 0.029
Note: Standard errors of hazard ratios are conditional on theta.
Thanks,
Doug Thomas
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
