Dear All,
To estimate the average duration of FMD virus excretion in a
transmission study we want to use Weibull survival analysis because in
some cases the animals were killed before the end of the experiment, or
were still excreting at the end of the experiment.
I'm using S-Plus version 6.2 on Windows XP. With our dataset and using
"statistics/survival/parametric survival" in the GUI I get the following
short output.
*** Parametric Survival ***
Call:
survReg(formula = Surv(duur, censor, type = "right") ~ 1, data =
surlammerkarin2, na.action = na.exclude, dist =
"weibull", scale = 0, control = list(maxiter = 30, rel.tolerance
= 1e-005, failure = 1))
Coefficients:
(Intercept)
2.876181
Scale= 0.6907516
Loglik(model)= -31.9 Loglik(intercept only)= -31.9
n= 16
My question is how do I calculate the mean duration of FMD virus
excretion?
I would think the intercept is the equal to log(mean survival), so the
mean survival = exp(intercept)
But when I look at the definition of the weibull distribution I have to
calculate
mean survival = exp(scale/intercept)^scale * gamma(scale).
Which one of the two is correct? And how can I calculate the S.E. of the
mean survival?
Aldo Dekker
------------------------------------------------------------------------
---
Aldo Dekker
Senior scientist
Laboratory vesicular diseases
CIDC-Lelystad, Centraal Instituut voor DierziekteControle Lelystad
(Central Institute for Animal Disease Control Lelystad)
P.O. Box 2004
82O3 AA The Netherlands
Aldo.Dekker@wur.nl
www.CIDC-Lelystad.NL
Tel: +31 - 320 - 238 603
Fax: +31 - 320 - 238 668
Address: Houtribweg 39
------------------------------------------------------------------------
---
|