Chi-squared test for imbalance class

Hi all !

I am dealing with chi-squared test performing with Crosstabe node.
I get through this properly, but i have a doubt: classes are imbalanced a lot. How can I deal with this ?

Hi @mdibiase2,

is it possible for you to share some sample data illustrating the issue please?


Sure !
I got the target variable presenting 3 classes and 2113 total observations.
Most of them, namely about 1800 belong to class 1, while the remaining are splitted among class 2 and 3. So a strong imbalance over classes does exist.

I want to check out if a categorial explanatory variable of 2 classes show different distribution among the 3 classes of the target variable, so Chi-squared is required.

Should I handle the class imbalance in this case ?