[R] Anova unbalanced

paladini paladini at beuth-hochschule.de
Wed Apr 17 11:47:09 CEST 2013

Hello everybody,
I have got a data set with about 400 companies. Each company has a 
score for its enviroment comportment between 0 and 100. These companies 
belong to  about 15 different countries. I have e.g. 70 companies from 
UK and 5 from Luxembourg,- so the data set is pretty unbalanced and I 
want to do an ANOVA. Somthing like aov(enviromentscore~country). But the 
aov function is just for a balanced design.
So I wonder if I can use fit=lm(enviromentscore~country), anova (fit) 
instead? Would this be okay or can it also only be used with balanced 

Thanking you in anticipation, best regards


More information about the R-help mailing list