[R] another aov question: unbalanced multiple responses

[RenE J.V. Bertin]

Hi,

This question is related to the bwplot issue I reported yesterday. I have a 3 factors (2x3x2) dataset that I collapsed into a 2 factors dataset (3x2 = sizexModality). For size==small, I have 2 observations per subject (Snr), for the other sizes only 1.

I reckoned that aov (and underneath, lm) might handle this as it should, since the subjects are idendified, when I do

> aov( y~Modality*size + Error(Snr/(Modality*size)) )

I indeed get a result without any warning. Remembering a remark on unbalanced designs in aov, I ran the same (?..) test in NCSS. It too gives results without any warnings or errors, but they're not exactly the same. The DF, SumSQ, MeanSq and F values are the same for Snr:size and for Snr:size:Modality, but they are not (except DF) for Snr:Modality and Snr. (On fully balanced data, I did get the same values in both progs.)

I have not found anything purporting to this specific situation in the documentation of either programme, so I am lost.

My simple question is: can aov do what I am trying to do here (cf. the call above), or if not, what is the correct way to do this?

Thanks again in advance,
RenE Bertin
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._