[R] Anova

[S Ellison]

> Now my question: Am I allowed to use these functions given 
> that my data is unbalanced? 

Unusually, Yes, assuming all the other requisite assumptions are reasonably well satisfied. One-way ANOVA interpretation is not much affected by imbalance because (among other things) with only one factor there is no 'model order' dependence or hypothesis separation to worry about. You don't need to use alternate sum-of-squares calculations (eg type II and III as in car) because types I, II and III sums of squares are all the same in a one-way case.

In a two-factor or higher analysis it's a (very) different story. 

However, one-way anova is still adversely affected by failures in homogeneity of variance, non-normality etc. 

S Ellison

*******************************************************************
This email and any attachments are confidential. Any use...{{dropped:8}}