If it is not standardized, I have the following script that I use to
calculate the r^2 for each variable of a model. I do wonder, however,
if there is a way to calculate a model-wide r^2 like index.
-Jarrett
##
#returns the r^2 for endogenous variables and error path coefficients
#
# Last updated 8/18/08
##
rsquare.sem<-function (sem.obj){
output.info<-list()
for (index in 1:length(sem.obj$S[,1])){
var.name<-sem.obj$var.names[index]
obs.var<-sem.obj$S[index,index]
est.var<-sem.obj$P[index,index]
r.sq<-1-(est.var/obs.var)
std.error.coef<-sqrt(1-r.sq)
#if it's 0, or close to (due to rounding error), it's exogenous
if(r.sq > 1e-10){
output.info<-c(output.info, var.name, obs.var, est.var, r.sq,
std.error.coef)
}
}
output.matrix<-matrix(output.info, ncol=5, byrow=TRUE)
colnames(output.matrix)<-c("Variable", "Observed.Variance",
"Estimated.Variance", "R^2", "Standardized.Error.Coefficient")
output.matrix
}
----------------------------------------
Jarrett Byrnes
Postdoctoral Associate, Santa Barbara Coastal LTER
Marine Science Institute
University of California Santa Barbara
Santa Barbara, CA 93106-6150
http://www.lifesci.ucsb.edu/eemb/labs/cardinale/people/byrnes/index.html
On Feb 9, 2010, at 4:17 AM, John Fox wrote:
> Dear Kathryn,
>
> I assume that MWDError is the error variance associated with MWD. If
> MWD is
> a standardized variable (i.e., with a variance of 1) then 59% of its
> variance is unaccounted for.
>
> I hope this helps,
> John
>
> --------------------------------
> John Fox
> Senator William McMaster
> Professor of Social Statistics
> Department of Sociology
> McMaster University
> Hamilton, Ontario, Canada
> web: socserv.mcmaster.ca/jfox
>
>> -----Original Message-----
>> From: r-help-boun...@r-project.org [mailto:r-help-boun...@r-project.org
>> ]
> On
>> Behalf Of Kathryn Barto
>> Sent: February-09-10 6:23 AM
>> To: r-help@r-project.org
>> Subject: [R] interpreting error estimate in SEM
>>
>> hi,
>>
>> I'm using the sem package, and I want to make sure I'm interpreting
>> the
>> output correctly. Here is an excerpt of my output. When I report
>> the MWD
>> error, should I say that 59% of the variation in MWD is not
>> explained by
> the
>> model, or that 59% of the variation is explained, or something
>> entirely
>> different.
>>
>> Parameter Estimates
>> Estimate Std Error z value Pr(>|z|)
>> NplantRoots 0.462029 0.120803 3.82466 1.3095e-04
>> plantRoots
> <--
>> - N
>> ...
>> CoError 1.000011 0.151850 6.58553 4.5325e-11 Co <--> Co
>> carbonateError 1.000009 0.156217 6.40142 1.5394e-10 carbonate
> <-->
>> carbonate
>> MWDError 0.589788 0.092128 6.40185 1.5351e-10 MWD
>> <-->
> MWD
>>
>> Thanks
>> Kathryn
>>
>> ______________________________________________
>> 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.
>
> ______________________________________________
> 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]]
______________________________________________
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.
