[R] mixed anova models with aov

Nicolas Schuck nico.schuck at googlemail.com
Sun Feb 19 22:44:22 CET 2012


Dear R Users, 

I am trying to calculate a mixed effect Anova with 2 nested between subject factors and two nested within subject factors. Specifically, I have data from two Agegroups (between subject variable old vs young) that are characterized by their genotype (between subjects, NESTED within Agegroup). All subjects make tests in two conditions (within subject, 2 levels), where the same test is repeated (within subject factor Block, NESTED within condition). 

I run the following model: 

aov(RT ~AGEGROUP*GENE*CONDITION*BLOCK + Error(ID/(CONDITION+BLOCK)), data = SRTT$LearnBlocks_KIBRA.df)   

Is this model correct? 

With the summary command, I get a whole lot of effects and I am partly confused what the correct error term for some of them is. I have the following four Error strata: 

ID
ID:CONDITION
ID:BLOCK
Within

I assume the condition and block effects can be obtained from the respective ID:CONDITON and ID:BLOCK error terms, respectively, correct? I am uncertain with respect to the between subject factors. Gene appears under all Error terms. Where is the correct effect listed? Where are the correct effects listed for the interactions involving Agegroup and Gene? 

Many thanks!
Nico


 


More information about the R-help mailing list