Hello,
I'm working with S-PLUS 8.04 in Windows XP.
I recently became proficient with two-factor ANOVA's and am now trying
to sort out a nested design.
For the two-factor ANOVA, I was able to organize my data using.....
> fnames <- list(sugar=c("lac", "mel", "suc", "OMG", "gal", "tre", "tal"),
+ concentration=c("22mM", "200mM", "500mM"))
> trials.design <- fac.design(c(7,3), fnames, rep=4)
> trials.df <- data.frame(trials.design, viability)
......to give me a table with 7 sugars and 3 treatments, and 4
replicates per combination.
For the nested ANOVA, I am comparing 3 different concentrations, and I
also want to see the differences between the 16 replicates that make
up each of concentration.
I have...
> fnames <- list(concentration=c("22", "200", "500"), wells=LETTERS[1:16])
> cytotox.design <- fac.design(c(3,16), fnames, rep=3)
> results.df <- data.frame(cytotox.design, reading)
... but I don't think this is correct. I'm not clear on how the nested
ANOVA is called in S-PLUS but the data.frame just doesn't seem right.
When I run the anova function, I don't get the F-statistic or p-value
(only Sum of Squares and Mean Square).
Does anyone know how to set it up correctly?
My ANOVA call:
> aov.cytotox <- aov(reading~treatment/wells, data=results.df)
I'm sitting in front of several textbooks, but I am stumped nonetheless.
Thanks once again,
Liz
|