[R] Adjusted means and generalized chain block designs

[Giles Crane]

To whom it may concern:

I compared several R package results,
and manual checked two generalized chain block
design experiments. The correct adjusted
treatment means can be computed
by using the effects library as follows:

    library(effects)

    aov1 = aov(y~blocks+rows+trt)
    means.aov = as.vector( allEffects(aov1)$trt$fit ) 

As a further check, I programmed an R function
to compute the adjusted treatment means
via John Mandel's manual method in his paper,
a somewhat lengthy and tricky task.

Finally, I will present  motivating examples
on behalf of my colleagues Cynthia Collins and 
Karin Mille, an index of Generalized Block Designs, and
GCB library in draft form, includingD-efficiency,
at the NIST UseR 2010 Conference
in July.

These designs are interesting combinatorially,
and quite applicable to physical measurements
which are made with high precision.

Best wishes,
Giles

-- 
View this message in context: http://r.789695.n4.nabble.com/Adjusted-means-and-generalized-chain-block-designs-tp1556545p2267205.html
Sent from the R help mailing list archive at Nabble.com.