[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Re: Help for survial analysis,21st Aug

From   Maarten buis <[email protected]>
To   [email protected]
Subject   Re: st: Re: Help for survial analysis,21st Aug
Date   Tue, 21 Aug 2007 10:28:29 +0100 (BST)

--- khan ali <[email protected]> wrote:
> They contend that if prices contain bubbles, then
> runs of positive returns will exhibit negative
> duration dependence, i.e., an inverse relation exists
> between the probability of a run ending and the length
> of the run. They adopted log-logistic model for
> testing the null hypothesis of no rational bubbles.
> The null hypothesis of no bubbles implies that the
> probability of a run ending is independent of  the
> prior returns or that the returns are random.In terms
> of the model, the null hypothesis of  no duration
> dependence is that slope(beta=0). 

The hazard will decline as time passes in a log-logistic, that is not
an empirical finding, it is just an assumption you have made that is
implicit in the log-logistic. If you want to test for negative duration
dependence you can either go for a Weibull or a Gompertz. In case of a
Weibull you are testing whether p < 1, and in case of the Gompertz you
are testing whether gamma < 0. See the example below and page 10 and 11
of .

This test is however problematic as any variation not captured by
covariates tend to induce negative duration dependence, see page 18-21
of . So if you find
negative duration dependence it could just as well be the result of
variotion between individuals not captured by covariates.

*--------- begin example -------------
sysuse cancer, clear
stset studytime, failure(died)
xi: streg i.drug age, distribution(loglogistic)
stcurve, hazard
xi: streg i.drug age, distribution(gompertz)
stcurve, hazard
xi: streg i.drug age, distribution(weibull)
stcurve, hazard
*---------- end example --------------
(For more on how to use examples I sent to the Statalist, see )

Hope this helps,

Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

Want ideas for reducing your carbon footprint? Visit Yahoo! For Good
*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index