We can use an ANOVA test for multiple types of variables and "values" of them.
An example should make this clearer: if a researcher wants to study whether the time of birth, or gender, of a student, has any effect on their GPA, they can set up a two-factor ANOVA.
One factor would be the time of birth: this could have 12 levels, the months of the year.
The other factor would be the student's gender: this would have 2 levels, male and female.
This would make 2*12=24 different treatment conditions, ie: combinations of levels of the two factors.
We could then use software like SPSS to test for any main effects, ie: differences in GPA in student's born in different months, or in different genders, and then (for a more detailed test) for any interactions between the month of birth and gender.
