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]

From |
Cameron McIntosh <cnm100@hotmail.com> |

To |
STATA LIST <statalist@hsphsun2.harvard.edu> |

Subject |
RE: st: Margins command after Cox regression |

Date |
Fri, 30 Dec 2011 10:54:34 -0500 |

You might be able to get them through a different type of hazard model: Kyyrä, T. (2009). Marginal Effects for Competing Risks Models with Piecewise Constant Hazards. Oxford Bulletin of Economics and Statistics, 71(4), 539–565. Chang, S.-H. (2004). Estimating Marginal Effects in Accelerated Failure Time Models for Serial Sojourn Times Among Repeated Events. Lifetime Data Analysis, 10(2), 175-190. Lo, S.M.S., & Wilke, R.A. (April 2011). A Regression Model for the Copula Graphic Estimator. University of Nottingham Discussion Paper No. 11/04. Nottingham, UK: University of Nottingham, School of Economics.http://www.nottingham.ac.uk/economics/documents/discussion-papers/11-04.pdf Cam ---------------------------------------- > From: fwolfe@arthritis-research.org > Date: Fri, 30 Dec 2011 08:36:58 -0600 > Subject: Re: st: Margins command after Cox regression > To: statalist@hsphsun2.harvard.edu > > Thanks, Maarten. > > Fred > > On Fri, Dec 30, 2011 at 8:13 AM, Maarten Buis wrote: > > On Fri, Dec 30, 2011 at 1:46 PM, Fred Wolfe wrote: > >> I am trying to get margins expressed in probability units rather than > >> hazard ratios after Cox regression (Stata 12). > > > > I don't think you can. First, it is hard to define the probability in > > the context of a survival analysis. Is it the probability of the event > > occuring at time = 0 or at the end of the study or somewhere in > > between? Second, even after you have defined what probability you > > want, it is still the aim of Cox to make such statements hard. That is > > not because Cox(*) is such a nasty person, but because he avoided > > making assumptions on the baseline hazard and by avoiding such > > assumptions it prevents you from making wrong assumptions. However, > > the price you'll have to pay is that it is only a useful model if you > > want to make statements on relative quantities, like the hazard ratio, > > and it becomes very hard to make statements on absolute quanitties > > like the hazard or the probability. That is a logical consequence of > > not estimating the baseline hazard, so there is not much that can be > > done about that. > > > > Hope this helps, > > Maarten > > > > (*) David not Nick > > > > -------------------------- > > Maarten L. Buis > > Institut fuer Soziologie > > Universitaet Tuebingen > > Wilhelmstrasse 36 > > 72074 Tuebingen > > Germany > > > > > > http://www.maartenbuis.nl > > -------------------------- > > * > > * 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/ > > * > * 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/ * * 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/

**Follow-Ups**:**Re: st: Margins command after Cox regression***From:*Fred Wolfe <fwolfe@arthritis-research.org>

**References**:**st: Margins command after Cox regression***From:*Fred Wolfe <fwolfe@arthritis-research.org>

**Re: st: Margins command after Cox regression***From:*Maarten Buis <maartenlbuis@gmail.com>

**Re: st: Margins command after Cox regression***From:*Fred Wolfe <fwolfe@arthritis-research.org>

- Prev by Date:
**Re: st: Margins command after Cox regression** - Next by Date:
**Re: st: Margins command after Cox regression** - Previous by thread:
**Re: st: Margins command after Cox regression** - Next by thread:
**Re: st: Margins command after Cox regression** - Index(es):