I am working with categorical data. How do I normalize a variable for effects outside the category?

Question

I have a continuous variable that I want to normalize across 4 categorical variables. This is not normalizing in the sense that making it have more like a normal distribution. Rather, it is to "clean" from effects not due to an observation's membership in that category.

Answer 1

I'M A MOM THE EASIEST WAY I CAN DESCRIBE TO
NORM: DIVIDE BY THE NUMBER OF VARIABLES : )

Answer 2

People who are not familiar with certain fields of scientific investigation probably don't realize that you are asking your question in statistical code language. However, I will answer in un-coded common English so that others might benefit from my advice.

I understand that you are very upset by your boyfriend cheating on you (a variable outside of its category). That's perfectly normal.

However, cutting off his genatalia (normalizing a variable) is not an appropriate response.

At the same time, it is very thoughtful of you to seek advice in how to cauterize the surgical site ("clean" from effects).

And, of course, you're correct. He'd never be able to set foot in the boy's showerroom again (observation's membership in that category).

Let me suggest that you might achieve similar results by correlating his coefficient. Then, run a chi square on the rest of his category. If his n = more than one, it's because of a standard deviation. In which case, he'll be guilty of covariation, and I'm afraid that you'll have to truncate his Yule's Q, making him a null hypothesis.

I hope that this satisfies your lambda.