Statalist


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

Re: st: stdes, SD for time at risk?


From   Steven Samuels <sjhsamuels@earthlink.net>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: stdes, SD for time at risk?
Date   Mon, 2 Feb 2009 19:06:56 -0500

---

You asked for the SD of "time at risk" as described by -stdes-. Are you saying now that you meant something different?

I'm not sure what you mean by "visit". If events can happen between "visits", you have a an interval-censoring or grouped-data setup. If that is the case, some -st- commands might not apply. Also, "time at risk" might not be time to the last visit.

I suggest that you create a toy subset of observations in your data and -stset- it. List the data, and calculate by hand from the listed variables, "time elapsed" as you want it and descriptive statistics for time elapsed. When you have done that, come back to the list and show us the data and computations, including the -stset- commands. Then perhaps we can help you with further commands to compute the descriptive statistics.

-Steve



On Feb 2, 2009, at 6:35 PM, MAY BAYDOUN wrote:

Thank you. Yes, I do have a multiple record data. So, that means each visit has a different time depending on how much time elapsed from the origin which is 50 years of age. Is there a way to figure this out with the variables that are created automatically by Stata (i.e. _t0, _st, _t and _d)? I need to find the time elapsed at last visit at which each person was at risk and summarize that time for the whole sample. Thanks for any help you can provide,

--- On Mon, 2/2/09, Steven Samuels <sjhsamuels@earthlink.net> wrote:

From: Steven Samuels <sjhsamuels@earthlink.net>
Subject: Re: st: stdes, SD for time at risk?
To: statalist@hsphsun2.harvard.edu
Date: Monday, February 2, 2009, 5:53 PM
If you have single record data, run -summary- on your
failure time variable.  If you have multiple record data,
keep only one record for each person first and then run
-summary-.
n Feb 2, 2009, at 11:52 AM, MAY BAYDOUN wrote:

Hello Statalisters,

I was wondering if you have used the stdes command and
were able to get a standard deviation for time at risk in
addition to mean, median and range. I am not sure if that is
at all possible. Is there another command one can use to get
the SD? Thanks a lot!
*
*   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