Statalist


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

st: WG: discrete Time Proportional Hazard Model


From   "Martin Weiss" <martin.weiss1@gmx.de>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: WG: discrete Time Proportional Hazard Model
Date   Fri, 17 Apr 2009 09:23:25 +0200

<> 

I am forwarding this on behalf of Tracy. My advice to her is the same as to
xiaohua:

You can post if you are subscribed to Statalist and post _in text-only
format_ from the address where you receive the messages. Make sure you tick
those boxes and it should work :-)


HTH
Martin

Von: Tracy Bai [mailto:tracy.bai@flinders.edu.au] 
Gesendet: Freitag, 17. April 2009 09:19
An: Martin Weiss
Betreff: discrete Time Proportional Hazard Model

Hi Martin,

I post the following email to Statalist nearly two hours ago, it seems that
it doesn't go through to the Statalist as well.

Could you please help me to put on it.

I am sorry if any inconvenient to you.

Thanks,

Tracy


Dear Statlist

I am going to fit a discrete Time Proportional Hazard Model to estimate
hazard rate for transition from school to the first job.

There is a survey question as below:
After you left school how long did it take you to find a full time job?
1.     Straight away/ already had one/ less than one month/One month
2.     Two months
3.     Three months
4.     Four months
5.     Five months
6.     Six months
7.     Seven months to less than one year
8.     One year or more

I am wondering how to deal with the intervals 7 and 8,
7. Seven months to less than one year
8.     One year or more
 I am going to introduce eight dummy variables d1-d8, for each interval, and
include them in the P.H. model. 
 If any individual falls into the interval seven without censored, then
response variable Y should have 0 0 0 0 0 0 1 
 If any individual falls into the interval eight without censored, then
response variable Y should have 0 0 0 0 0 0 0 1
It doesn't mean that those individuals have exactly time seven months or
eight months, just because I code them 7 for Seven months to less than one
year, and 8 for One year or more.
I just want to make sure that I am right to re-organise the data in this way
so that I can properly fit a discrete Time Proportional Hazard Mode.
Thanks,
Tracy 


*
*   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