Page 1 of 1

Singular Matrix Problem

Posted: Thu May 17, 2018 12:47 pm
by KIHO KWON
Hi.
When I add control variables, Singular Matrix Problem occurred.
Each control variables are categorical variable. So I transformed all of them to dummy variable.
My sample size is 307.
The number of control variable is 12 which are demographic variables(age, income, etc) and use experiences(watching time, platform type, favorite genre, etc).
So dummy variables are a lot.
For example, Age - dummy variable : age1(10~19old), age2(20~29old) like this.

Unfortunately, Even I read information that the software said, I don't know what dummy variable is troublesome.
How can I solve this problem?

Re: Singular Matrix Problem

Posted: Thu May 17, 2018 8:37 pm
by jmbecker
If you have three categories, you add only 2 dummy variables. If you add dummy variables for all categories you will get this error.
The category that does not have a dummy variable is the reference category (baseline).

Re: Singular Matrix Problem

Posted: Wed Jul 14, 2021 5:39 pm
by Harman
Hi Prof. Becker,

How to interpret this aspect, i.e. baseline/reference vs dummy variables for rest of the categories?
I will appreciate if you can share some resource which I can read/watch on this.