Boxplots compare very clearly the data in different groups:

**code:**

> ggplot(bwdf, aes(x=race, y=age))+ geom_boxplot(notch=T, varwidth=T)

**Output graph:**

Again overlapping notches of 2 groups indicate no significant difference between 2 groups.

**Effect size:**

Eta-squared is used as an effect size for ANOVA. This gives fraction of change in outcome variable accounted for by grouping variable.