[R] appropriate error term for AOV repeated block design

hpdutra hpdutra at yahoo.com
Mon Feb 9 19:25:14 CET 2009


I am using AOV and I am not sure if my model structure is correct for the
repeated measures.

I want to verify if there is a interaction of a random factor(block) with
fixed factors(veget,fruit, time,)

> model<-aov(trackarcsin~veget*fruit*time*block
> +Error(block/(veget*fruit)),data=mice)

Because it is repeated measures do I have to use an error term for that?

here is my experimental design
My experiment consists of manipulating fruits (F) and vegetation (V) to two
levels
each(intact (I) or removed (R)  
VIFR 
VRFI
VIFI
VRFR
=============================
 summary(model<-aov(trackarcsin~veget*fruit*time*block
+Error(block/(veget*fruit)),data=mice) )

Error: block
      Df Sum Sq Mean Sq
block  2 2804.2  1402.1

Error: block:veget
            Df Sum Sq Mean Sq
veget        1 537.66  537.66
veget:block  2  19.69    9.85

Error: block:fruit
            Df  Sum Sq Mean Sq
fruit        1  79.417  79.417
fruit:block  2 204.688 102.344

Error: block:veget:fruit
                  Df Sum Sq Mean Sq
veget:fruit        1 31.053  31.053
veget:fruit:block  2 88.472  44.236

Error: Within
                       Df  Sum Sq Mean Sq
time                   14 3160.61  225.76
veget:time             14  554.85   39.63
fruit:time             14  336.02   24.00
time:block             28 1306.89   46.67
veget:fruit:time       14  623.75   44.55
veget:time:block       28  584.31   20.87
fruit:time:block       28 1401.77   50.06
veget:fruit:time:block 28  810.79   28.96



-- 
View this message in context: http://www.nabble.com/appropriate-error-term-for-AOV-repeated-block-design-tp21918890p21918890.html
Sent from the R help mailing list archive at Nabble.com.




More information about the R-help mailing list