Dear friend. I think it can be done using a generalized liniear mixed model -
implemented for instance in genstat:
read Y
.5 .6 .1 .2
.7 .8 .2 .3
.8 .9 .3 .4 :
units[nval=12]
factor [lev=3] a
factor [lev=2] b
factor [lev=2] sub
generate a,b,sub
print a,b,sub,Y
glmm[distrib=bin;random=sub;fixed=a*b] Y;nbin=100
vplot
vdisplay[print=wald]
This produces totally plausible output to the toy example.
A reference could be to Breslow and Clayton, ASA 88: 9 ff 1993
hope this could help-
Troels
Bill Simpson skrev:
> I don't know much about GLM in general or glm in R. Can anyone tell me
> how to do the following (in R or some other stat system) or refer me to a
> textbook discussion?
>
> Two factor ANOVA repeated measures design.
> Each subject gives a percent correct.
>
> I am assuming the correct way to proceed is to fit a generalized linear
> model with binomial responses and logistic link. But I have not found a
> detailed discussion of this in either McCullagh & Nelder or Dobson.
>
> Here is a toy example. Each datum is proportion correct out of 100
> trials for one subject:
> b1 b2
> a1 .5,.6 .1,.2
> a2 .7,.8 .2,.3
> a3 .8,.9 .3,.4
> (2 subjects)
>
> How does the analysis proceed? Thanks very much for any help.
>
> Bill Simpson
>
>
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
> r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
> Send "info", "help", or "[un]subscribe"
> (in the "body", not the subject !) To: r-help-request at
stat.math.ethz.ch
>
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !) To: r-help-request at
stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._