ANOVA with unequal sample sizes

#1
hello,

I have a question about ANOVA. This is the output I received when testing the effect of 'opnop' on mmse1.
opnop is divided in 4 groups, but one group had a sample size of 4, while the other groups are bigger. What can the problem be when you decide to continue your exercise with such a difference??

OPNOP mean group size SD variance
1 19,00 4 6,583 43,33
2 20,00 22 9,217 84,95
3 19,50 20 7,178 51,53
4 16,00 13 8,727 76,17
Totaal 18,88 59 8,253 68,11

I really hope someone can help with this because I can't find the answer ...

Thanks!
 
#2
With such low number of cases, you would have a high chance of type II error. Furthermore, ANOVA is sensitive to sample size. Instead, in this case, you can run nonparametric Kruskal-Wallis test followed by Mann-Whitney U test for each pair.