How to check multicollinearity a logistic regression model?

[lionking19063]

Hi, 

Could someone tell me which method is correct to check multicollinearity a logistic regression model? Thanks.

1)  method 1

proc logistic data= sample;

model y(event='1') =x1 x2 x3;

run;

proc reg data=sample;

model  y= x1 x2 x3/vif;

run;

2) method 2

proc logistic data= sample;

model y(event='1') =x1 x2 x3;

output out=out_data p=pred;

run;

proc reg data=out_data;

model  pred= x1 x2 x3/vif;

run;

[PaigeMiller · Accepted solution]

I would vote for method 1

[sbxkoenk]

• Checking Multicollinearity in Logistic Regression model
  https://communities.sas.com/t5/New-SAS-User/Checking-Multicollinearity-in-Logistic-Regression-model/...
• SESUG 2015 -- Paper SD69
  Integrating PROC REG and PROC LOGISTIC for Collinearity
  Examination, Sample Scoring and Model Evaluation
  Alec Zhixiao Lin, Loan Depot, Foothill Ranch, CA
  https://lexjansen.com/sesug/2015/69_Final_PDF.pdf

Koen

[Ksharp]

None of them.

You should check CORRB option of PROC LOGISTIC to check the correlation between the estimated coefficient .

32471 - Testing assumptions in logit, probit, Poisson and other generalized linear models (sas.com)

Also calling @StatDave  @SteveDenham 

proc logistic data=sashelp.heart;
class sex bp_status;
model status=sex bp_status height weight ageatstart/corrb ;
run;