One way ANOVA and R2

#1
I have used XLSTAT to run an Anova, and in the output I got a figure for the R2 (which I wasn't actually expecting as haven't had this information in the past when using SPSS). As the R2 was very low, it has made me question the robustness of the Anova results. Is there a recommended minimum R2 when running one way Anova? Many thanks in advance for any advice.
 
#3
R^2 should not be used as the sole measure of "success" for a model. In psychology fields, it's common to see models with low R^2. So, it depends on what you are studying. Perhaps there is literature to reference in your case. I have seen R^2 used as a measure of effect size.
https://en.wikipedia.org/wiki/Effect_size
Thank you so much for your response. As the independent variable is categorical, I'm not even sure whether I should be considering the R2 in my example. I want to know whether the mean varies among 5 groups (i.e. 5 categories). Thanks again
 

Buckeye

Active Member
#4
Just to be clear. R^2 can still be interpreted with a categorical predictor. It is defined as the proportion of variability in the dependent variable that can be explained by the model with the categorical predictor.