Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

Re: st: RE: Command "margin" after Cox Regression


From   Yuval Arbel <yuval.arbel@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: RE: Command "margin" after Cox Regression
Date   Wed, 26 Oct 2011 18:54:12 +0200

I ran the experiment I proposed before where full is obtained for
max_red=0, full1 is obtained for max_red=10 etc. What I got is that
survival rates indeed deteriorate but a very slow pace, which
correspond to a hazard rate of 1.1227. Note that the differences are
statistically significant

. collapse (mean) full full1 full2 full3 full4 full5 full6 if
fail==1,by(time_index)

. summ

    Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
  time_index |       103          63    29.87753         12        114
        full |       103    .9951255    .0492695   .4999498          1
       full1 |       103    .9950817     .049276   .4998402          1
       full2 |       103    .9949423    .0492972   .4994917   .9999999
       full3 |       103     .994499    .0493665   .4983847   .9999997
-------------+--------------------------------------------------------
       full4 |       103    .9930912    .0496077    .494879    .999999
       full5 |       103    .9886374    .0505757   .4838885    .999997
       full6 |       103    .9747187    .0553393   .4505192   .9999903

. ttest full==full1

Paired t test
------------------------------------------------------------------------------
Variable |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
    full |     103    .9951255    .0048547    .0492695    .9854963    1.004755
   full1 |     103    .9950817    .0048553     .049276    .9854512    1.004712
---------+--------------------------------------------------------------------
    diff |     103    .0000438    4.04e-06     .000041    .0000358    .0000518
------------------------------------------------------------------------------
     mean(diff) = mean(full - full1)                              t =  10.8553
 Ho: mean(diff) = 0                              degrees of freedom =      102

 Ha: mean(diff) < 0           Ha: mean(diff) != 0           Ha: mean(diff) > 0
 Pr(T < t) = 1.0000         Pr(|T| > |t|) = 0.0000          Pr(T > t) = 0.0000

. ttest full1==full2

Paired t test
------------------------------------------------------------------------------
Variable |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
   full1 |     103    .9950817    .0048553     .049276    .9854512    1.004712
   full2 |     103    .9949423    .0048574    .0492972    .9853077    1.004577
---------+--------------------------------------------------------------------
    diff |     103    .0001394    .0000128    .0001303    .0001139    .0001649
------------------------------------------------------------------------------
     mean(diff) = mean(full1 - full2)                             t =  10.8551
 Ho: mean(diff) = 0                              degrees of freedom =      102

 Ha: mean(diff) < 0           Ha: mean(diff) != 0           Ha: mean(diff) > 0
 Pr(T < t) = 1.0000         Pr(|T| > |t|) = 0.0000          Pr(T > t) = 0.0000

. ttest full2==full3

Paired t test
------------------------------------------------------------------------------
Variable |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
   full2 |     103    .9949423    .0048574    .0492972    .9853077    1.004577
   full3 |     103     .994499    .0048642    .0493665    .9848509    1.004147
---------+--------------------------------------------------------------------
    diff |     103    .0004433    .0000408    .0004142    .0003623    .0005242
------------------------------------------------------------------------------
     mean(diff) = mean(full2 - full3)                             t =  10.8616
 Ho: mean(diff) = 0                              degrees of freedom =      102

 Ha: mean(diff) < 0           Ha: mean(diff) != 0           Ha: mean(diff) > 0
 Pr(T < t) = 1.0000         Pr(|T| > |t|) = 0.0000          Pr(T > t) = 0.0000

. ttest full3==full4

Paired t test
------------------------------------------------------------------------------
Variable |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
   full3 |     103     .994499    .0048642    .0493665    .9848509    1.004147
   full4 |     103    .9930912     .004888    .0496077    .9833959    1.002786
---------+--------------------------------------------------------------------
    diff |     103    .0014078    .0001294    .0013131    .0011512    .0016645
------------------------------------------------------------------------------
     mean(diff) = mean(full3 - full4)                             t =  10.8814
 Ho: mean(diff) = 0                              degrees of freedom =      102

 Ha: mean(diff) < 0           Ha: mean(diff) != 0           Ha: mean(diff) > 0
 Pr(T < t) = 1.0000         Pr(|T| > |t|) = 0.0000          Pr(T > t) = 0.0000

. ttest full4==full5

Paired t test
------------------------------------------------------------------------------
Variable |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
   full4 |     103    .9930912     .004888    .0496077    .9833959    1.002786
   full5 |     103    .9886374    .0049834    .0505757    .9787529    .9985219
---------+--------------------------------------------------------------------
    diff |     103    .0044538    .0004069      .00413    .0036466     .005261
------------------------------------------------------------------------------
     mean(diff) = mean(full4 - full5)                             t =  10.9445
 Ho: mean(diff) = 0                              degrees of freedom =      102

 Ha: mean(diff) < 0           Ha: mean(diff) != 0           Ha: mean(diff) > 0
 Pr(T < t) = 1.0000         Pr(|T| > |t|) = 0.0000          Pr(T > t) = 0.0000

. ttest full5==full6

Paired t test
------------------------------------------------------------------------------
Variable |     Obs        Mean    Std. Err.   Std. Dev.   [95% Conf. Interval]
---------+--------------------------------------------------------------------
   full5 |     103    .9886374    .0049834    .0505757    .9787529    .9985219
   full6 |     103    .9747187    .0054527    .0553393    .9639032    .9855342
---------+--------------------------------------------------------------------
    diff |     103    .0139187    .0012487    .0126726     .011442    .0163954
------------------------------------------------------------------------------
     mean(diff) = mean(full5 - full6)                             t =  11.1469
 Ho: mean(diff) = 0                              degrees of freedom =      102

 Ha: mean(diff) < 0           Ha: mean(diff) != 0           Ha: mean(diff) > 0
 Pr(T < t) = 1.0000         Pr(|T| > |t|) = 0.0000          Pr(T > t) = 0.0000

.
end of do-file

.


On Wed, Oct 26, 2011 at 5:36 PM, Weichle, Thomas <Thomas.Weichle@va.gov> wrote:
> Hi Yuval,
> The predict() option for -margins- specifies the response and the
> response will be the default prediction that would be produced by
> -predict- after the underlying estimation command.  For -stcox-, the
> default prediction is the hazard ratio.
>
> Based on your first margins command (margins, at(max_red=10) atmeans
> predict(nohr)), it appears as though you are predicting the hazard since
> you specified 'nohr' in the predict() option.
>
> For your second margins command (margins, at(max_red=20) atmeans
> predict(basesurv)), it appears as though you are receiving the error
> message because you specified 'basesurv' in the predict() option.
> Baseline survival is a function of time and not the beta coefficients.
>
> Tom Weichle
> Math Statistician
> Center for Management of Complex Chronic Care (CMC3)
> Hines VA Hospital, Bldg 1, C202
> 708-202-8387 ext. 24261
> Thomas.Weichle@va.gov
>
> *
> *   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/
>



-- 
Dr. Yuval Arbel
School of Business
Carmel Academic Center
4 Shaar Palmer Street, Haifa, Israel
e-mail: yuval.arbel@gmail.com

*
*   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   |   Site index