Logistic Regression omits some of my characteristic variables

[Maria Philiposian]

Hi forum!

I have done a logistic regression with many characteristic variables. When I examine the result table I find that some of the characteristic variables have been omitted - or when I export the table to LaTex, then these specific variables have an estimated value =0 and a SD=0.

E.g. the wealth variable also has a 5th category "poorest", but in the regression table, it has been excluded (picture attached), and in LaTex it occurs, but with SD value 0 and estimated value 0.
I am unsure why this happens - is it due to multicollinearity? How can it be explained?

Thank you in advance, and have a great day.

Best,
Maria

[Leonardo Guizzetti]

The omitted category here is your reference category. It is not estimated by design. Its coefficient will always be 0 (on whatever raw scale is used, here it’s the log odds scale) and has no associated variance.

[Maria Philiposian]

Hi Leonardo

Thank you very much for the answer! I think it makes more sense now!
Have a great day.

Best,
Maria