[R] Nonlinear regression question

[Greg Snow]

What null hypothesis are you trying to test?  There is a standard null for linear models that makes sense in a large number of cases, but what the null is for non-linear regression is not obvious (and the coefficient = 0 may not even be possibly, let alone interesting).  If you can state what your null hypothesis is then there are ways to get p-values, but easier is to just compute confidence intervals for the parameter of interest and see if the null value is in the interval.

-- 
Gregory (Greg) L. Snow Ph.D.
Statistical Data Center
Intermountain Healthcare
greg.snow at imail.org
801.408.8111


> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
> project.org] On Behalf Of Tatiana Donnay
> Sent: Thursday, September 15, 2011 6:41 AM
> To: r-help at stat.math.ethz.ch
> Subject: [R] Nonlinear regression question
> 
> Hello,
> 
> If I understand good, I can't have p-value for the nls model.
> 
> I have 2 vectors. And I'am doing
> 
> model <- nls(crf ~a*(1-exp(-x/b)) + c, data= d,
>               start = list(a=1, b=3, c=0))
> 
> and I want to know If my result is significat, if I can't have p-value,
> how can I know it?
> 
> Thank you
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-
> guide.html
> and provide commented, minimal, self-contained, reproducible code.