Some questions about one-way ANOVA in JMP 9

One-way ANOVA can be used to estimate the contribution of a variable (x) to the variation in another one (y). I have some questions about it:

1. How to perform it in JMP 9? I am eager to get the R square.
2. Is this R square identical with the one by multivariate regression?
3. In ANOVA, is the variable necessarily nominal and more than 2 groups?

Thanks a lot!