Dear all, I would like to run in R the anova showed in the following pamphlet. http://www.for.gov.bc.ca/hre/biopamph/pamp14.pdf For A = 0 and B =0 I have de control group. Best regards. A B replication response 0 0 1 24 0 0 2 27 0 0 3 36 0 0 4 28 0 0 5 32 1 1 1 43 1 1 2 39 1 1 3 32 1 1 4 36 1 1 5 50 1 2 1 34 1 2 2 35 1 2 3 45 1 2 4 37 1 2 5 52 1 3 2 34 1 3 3 35 1 3 4 58 1 3 5 35 1 4 1 26 1 4 2 49 1 4 3 44 1 4 4 39 1 4 5 37 1 5 1 37 1 5 2 32 1 5 3 33 1 5 4 37 1 5 5 36 2 1 1 43 2 1 2 33 2 1 3 33 2 1 4 38 2 1 5 41 2 2 1 42 2 2 2 45 2 2 3 34 2 2 4 29 2 2 5 33 2 3 1 33 2 3 2 42 2 3 3 30 2 3 4 28 2 3 5 40 2 4 1 39 2 4 2 33 2 4 3 20 2 4 4 29 2 4 5 24 2 5 1 37 2 5 2 46 2 5 3 33 2 5 4 27 2 5 5 34
Charles C. Berry
2007-Aug-15 17:29 UTC
[R] ANOVA: Factorial designs with a separate control
Lamac, The ANOVA shown in 'pamph14' may not be suitable for your data. If the replicates are separate experiments or blocks, you will need to become familiar with the nlme package and the lme function in it. ( Your labelling of replicates suggests that this is the case, viz. no A==1 & B == 3 & replicate == 1 combination was found.) There is an excellent book that serves as reference for that package (Pinheiro, J.C., and Bates, D.M. (2000) "Mixed-Effects Models in S and S-PLUS", Springer.) If this turns out to be too 'deep' for you, you will do best to find a statistician who is well versed in linear mixed effects models to help you. -- However, if the replicates are independent realizations, then> # copy Lamac's data to the clipboard first > > dat <- read.table("clipboard",header=T) > fit <- lm ( response ~ I( A==0 ) + as.factor(A)*as.factor(B), dat ) > anova( fit )Gives you the sequential sums of squares ( your design is unbalanced hence the 'sequential' qualifier ). If you want what Wendy Bergerud (author of 'pamph14') called the 'TREAT' sum of squares, you can add all but the 'Residuals' SS. Likewise with the df. You can also get the TREAT ANOVA table directly by revising the formula in lm() above. The resulting formula is very simple, and you should have some fun - as well as strengthen your skill with formulas in R - trying to find it. You may want to review Chapter 11 Statistical models in R in An Introduction to R before trying this. HTH, Chuck On Tue, 14 Aug 2007, lamack lamack wrote:> Dear all, I would like to run in R the anova showed in the following > pamphlet. > > http://www.for.gov.bc.ca/hre/biopamph/pamp14.pdf > > For A = 0 and B =0 I have de control group. > > Best regards. > > A B replication response > 0 0 1 24 > 0 0 2 27 > 0 0 3 36 > 0 0 4 28 > 0 0 5 32 > 1 1 1 43 > 1 1 2 39 > 1 1 3 32 > 1 1 4 36 > 1 1 5 50 > 1 2 1 34 > 1 2 2 35 > 1 2 3 45 > 1 2 4 37 > 1 2 5 52 > 1 3 2 34 > 1 3 3 35 > 1 3 4 58 > 1 3 5 35 > 1 4 1 26 > 1 4 2 49 > 1 4 3 44 > 1 4 4 39 > 1 4 5 37 > 1 5 1 37 > 1 5 2 32 > 1 5 3 33 > 1 5 4 37 > 1 5 5 36 > 2 1 1 43 > 2 1 2 33 > 2 1 3 33 > 2 1 4 38 > 2 1 5 41 > 2 2 1 42 > 2 2 2 45 > 2 2 3 34 > 2 2 4 29 > 2 2 5 33 > 2 3 1 33 > 2 3 2 42 > 2 3 3 30 > 2 3 4 28 > 2 3 5 40 > 2 4 1 39 > 2 4 2 33 > 2 4 3 20 > 2 4 4 29 > 2 4 5 24 > 2 5 1 37 > 2 5 2 46 > 2 5 3 33 > 2 5 4 27 > 2 5 5 34 > > ______________________________________________ > R-help at stat.math.ethz.ch 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. >Charles C. Berry (858) 534-2098 Dept of Family/Preventive Medicine E mailto:cberry at tajo.ucsd.edu UC San Diego http://famprevmed.ucsd.edu/faculty/cberry/ La Jolla, San Diego 92093-0901