R help - Nov 2010 - Probit Analysis: Confidence Interval for the LD50 using Fieller's and Heterogeneity (UNCLASSIFIED)

Classification:  UNCLASSIFIED 
Caveats: NONE

A similar question has been posted in the past but never answered.  My
question is this: for probit analysis, how do you program a 95%
confidence interval for the LD50 (or LC50, ec50, etc.), including a
heterogeneity factor as written about in "Probit Analysis" by
Finney(1971)?  The heterogeneity factor comes into play through the
chi-squared test for homogeneity and is equal to h=chi^2/(k-2), where k
is the number of doses and k-2 are the degrees of freedom.

I have done a lot of research on this and really appreciate any help
anyone can provide.  Thank you!


Classification:  UNCLASSIFIED 
Caveats: NONE

On Nov 22, 2010, at 2:24 PM, Kenney, Colleen T CTR USA AMC wrote:
> Classification:  UNCLASSIFIED
> Caveats: NONE
>
> A similar question has been posted in the past but never answered.  My
> question is this: for probit analysis, how do you program a 95%
> confidence interval for the LD50 (or LC50, ec50, etc.), including a
> heterogeneity factor as written about in "Probit Analysis" by
> Finney(1971)?  The heterogeneity factor comes into play through the
> chi-squared test for homogeneity and is equal to h=chi^2/(k-2),  
> where k
> is the number of doses and k-2 are the degrees of freedom.
>
> I have done a lot of research on this and really appreciate any help
> anyone can provide.  Thank you!
The reason it may not have had a rely ... assuming the questions was  
homomorphic to this one... is that there is no R content in this  
question. You may want to post on:

http://stats.stackexchange.com/

But I suggest you be somewhat more descriptive than just citing a 40  
year-old text.

-- 

David Winsemius, MD
West Hartford, CT

Hi:

The MASS package has a function dose.p() to produce a CI for ED50, ED90 or
EDp in general (0 < p < 100).  It takes a model object (presumably from a
suitable logistic regression) as input. You could always take the code
already available and adapt it to your situation or you could investigate
one or more of the packages devoted to dose-response models.

A useful thing to know in R is the sos package, starting with its primary
function, findFn():

library(sos)
findFn('ED50')

scares up the three major packages related to dose-response models with 40
matches to ED50.

HTH,
Dennis

On Mon, Nov 22, 2010 at 11:24 AM, Kenney, Colleen T CTR USA AMC <
colleen.t.kenney@us.army.mil> wrote:
> Classification:  UNCLASSIFIED
> Caveats: NONE
>
> A similar question has been posted in the past but never answered.  My
> question is this: for probit analysis, how do you program a 95%
> confidence interval for the LD50 (or LC50, ec50, etc.), including a
> heterogeneity factor as written about in "Probit Analysis" by
> Finney(1971)?  The heterogeneity factor comes into play through the
> chi-squared test for homogeneity and is equal to h=chi^2/(k-2), where k
> is the number of doses and k-2 are the degrees of freedom.
>
> I have done a lot of research on this and really appreciate any help
> anyone can provide.  Thank you!
>
>
> Classification:  UNCLASSIFIED
> Caveats: NONE
>
> ______________________________________________
> R-help@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.
>
	[[alternative HTML version deleted]]