[R-sig-ME] Calculating upper and lower confidence limits on a population estimate derived from multiple point estimates

Farrar, David Farrar.David at epa.gov
Tue Jul 29 15:26:13 CEST 2014


Do you have a model for autocorrelation along stream networks? If you do you should be in position for parametric bootstrap.  If you don't, something to consider may be the SSN package.  I haven't used that but I know some people who are want to try. 

cheers,
David


-----Original Message-----
From: r-sig-mixed-models-bounces at r-project.org [mailto:r-sig-mixed-models-bounces at r-project.org] On Behalf Of Jeff Newmiller
Sent: Monday, July 28, 2014 9:47 PM
To: Reuben Smit; r-sig-mixed-models at r-project.org
Subject: Re: [R-sig-ME] Calculating upper and lower confidence limits on a population estimate derived from multiple point estimates

In general, no. Depends on the level of correlation between all values being added. This is a pretty basic statistical theory question.

---------------------------------------------------------------------------
Jeff Newmiller                        The     .....       .....  Go Live...
DCN:<jdnewmil at dcn.davis.ca.us>        Basics: ##.#.       ##.#.  Live Go...
                                      Live:   OO#.. Dead: OO#..  Playing
Research Engineer (Solar/Batteries            O.O#.       #.O#.  with
/Software/Embedded Controllers)               .OO#.       .OO#.  rocks...1k
---------------------------------------------------------------------------
Sent from my phone. Please excuse my brevity.

On July 28, 2014 10:05:27 AM PDT, Reuben Smit <smit.reuben at gmail.com> wrote:
>I am generating a river reach population estimate for a freshwater 
>mussel by summing point estimates made across a gridded point network 
>(within the
>reach) using a generalized linear mixed model framework. I have 
>generated 95% confidence/prediction intervals at each of the ~150,000 
>point locations in R. I have summed all of the point estimates to 
>derive the reach population estimate, but am unsure how to derive a 
>single confidence interval for the population estimate using the 
>150,000 individual-point confidence intervals. My question: Is it 
>statistically valid to simply sum all the lower estimates and upper 
>estimates to obtain the absolute upper and lower most population 
>confidence limits?
>
>	[[alternative HTML version deleted]]
>
>_______________________________________________
>R-sig-mixed-models at r-project.org mailing list 
>https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

_______________________________________________
R-sig-mixed-models at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models



More information about the R-sig-mixed-models mailing list