Statalist


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

st: SE's on survival estimates from parametric models


From   "Mark Hebblewhite" <mark.hebblewhite@umontana.edu>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: SE's on survival estimates from parametric models
Date   Fri, 24 Apr 2009 10:37:20 -0600

My question is how to calculate the SE/95% CI for the predicted survival estimate from a parametric model in STATA. I know STATA is doing this with stcurve commands, but I haven't been able to figure out how to solve for a point estimate of survival and its SE for time t.  

I have analyzed a large dataset of mule deer fawn survival (n~600) and Cox/parametric, and found a parametric (Weibull) model to be the best fitting parametric model (100's of dAIC units).  One of the goals for our study is to understand the survival at some time t - i.e., for the winter of the study. 

Given a Weibull equation, for a specific time, I then solve for the hazard at time t = 153 (our study period) to get a point estimate.  

       H(t)  = exp(B0 + B1X...+ BnXn) tp for time t = 150 (length of the study) 
       For example, H(t)  = 0.5795 for a female mule deer fawn of X mass, etc...

Thanks (in advance) for your help,

Mark Hebblewhite


The model results are pasted below:

Weibull regression -- log relative-hazard form 

No. of subjects =          581                     Number of obs   =      2349
No. of failures =          253
Time at risk    =        59812
                                                   LR chi2(9)      =    139.61
Log likelihood  =   -477.86548                     Prob > chi2     =    0.0000

------------------------------------------------------------------------------
          _t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
      novdec |   .0069395    .005974     1.16   0.245    -.0047693    .0186483
      janmar |   .0139184   .0044337     3.14   0.002     .0052284    .0226083
     tminoct |   .0416139   .0584366     0.71   0.476    -.0729197    .1561475
      tminjm |  -.1767656   .0550239    -3.21   0.001    -.2846105   -.0689206
      tminnd |   .0180625   .0631216     0.29   0.775    -.1056536    .1417786
     october |  -.0026746   .0051782    -0.52   0.606    -.0128237    .0074745
        mass |  -.1298777   .0215789    -6.02   0.000    -.1721716   -.0875838
        male |   .0924923   1.024433     0.09   0.928     -1.91536    2.100345
    malemass |   .0059718   .0296013     0.20   0.840    -.0520457    .0639894
       _cons |  -6.675164    .953159    -7.00   0.000    -8.543321   -4.807007
-------------+----------------------------------------------------------------
       /ln_p |   .4487272   .0714428     6.28   0.000     .3087019    .5887525
-------------+----------------------------------------------------------------
           p |   1.566317   .1119021                      1.361656    1.801739
         1/p |   .6384402    .045612                      .5550192    .7343997
------------------------------------------------------------------------------



Sincerely, 


Mark Hebblewhite 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 
Assistant Professor
Wildlife Biology Program 
College of Forestry and Conservation 
University of Montana 
Missoula, MT, 59812 
Office Phone: +1(406)-243-6675
Fax: +1(406)-243-4557 
Email: mark.hebblewhite@cfc.umt.edu 
http://www.cfc.umt.edu/HebLab/
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 


*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index