[R] anova
Lynn Govaert
lynn.govaert at gmail.com
Thu Aug 21 12:13:13 CEST 2014
I pressed enter to soon.
Again
Hi all,
I have troubles with doing an anova. So I have the following variables: a
variable group with two levels, a continuous variable trait and within each
group we have 12 organisms (clone) and for each clone we have 5 replicates.
So we want to see if for the variable trait the two groups differ,
including the information for the clones.
So the dataset looks like:
trait group clone
... 1 a1
... 1 a2
...
... 1 a12
... 1 a1
...
... 2 b1
... 2 b2
etc
trait are just some numbers.
So clone is nested in group.
Then I don't understand how to make the anova,
I was thinking we want to see if there is an effect of group
So we built the model aov(trait ~group)
but because we also want to include the effect of clone which is a random
effect we do
aov(trait ~ group + Error(clone))
but because Clone is nested in group
we do
aov(trait ~ group + Error(group/clone))
but if I do this I get the following output
Error: PopulationTNFMF
Df Sum Sq Mean Sq
PopulationTNFMF 1 0.00917 0.00917
Error: PopulationTNFMF:CloneTNFMF
Df Sum Sq Mean Sq F value Pr(>F)
Residuals 1 0.0001849 0.0001849
Error: Within
Df Sum Sq Mean Sq F value Pr(>F)
Residuals 117 0.02107 0.0001801
and I don't get any p-values, so I guess I'm doing something wrong.
Can someone help?
Thank you in advance
Lynn
2014-08-21 12:08 GMT+02:00 Lynn Govaert <lynn.govaert at gmail.com>:
> Hi all,
>
> I have troubles with doing an anova. So I have the following variables: a
> variable group with two levels, a continuous variable trait and within each
> group we have 12 organisms (clone) and for each clone we have 5 replicates.
> So we want to see if for the variable trait the two groups differ,
> including the information for the clones.
>
> So the dataset looks like:
>
> trait group
>
[[alternative HTML version deleted]]
More information about the R-help
mailing list