I have a variable "group" whose values 1,2 and 3 divides the sample into three groups. The second variable is "cheating" that the values 1 and 2 determines whether someone cheated or not. Now I want to know how that groups cheated (at %), and if there are any significant differences between groups. Perhaps I should use the anova, but ... my resulst are weird. What should i do? Thx for any help!