Some explanation needed about "ABSORB" statement

[Phil_NZ]

Hi all,

@PaigeMiller  gave me some useful suggestions regarding the ABSORB statement. However, I think it is better to raise a question in separate thread. in this discussion, @PaigeMiller documented that:

ABSORB allows the model to be fit. You cannot get predicted values if you use the ABSORB statement. You cannot get least squares means of the ABSORBed variables. ABSORB does not help you understand multicollinearity, and it does not allow you to reduce multicollinearity.

I much appreciate if you can help me to clarify it. Because not only in SAS but in STATA, we also use absorb statements quite frequently.

[PaigeMiller]

What part needs clarification?

Here is the documentation, where SAS explains it:

https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.4/statug/statug_glm_syntax02.htm

https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.4/statug/statug_glm_details24.htm