topic Re: Accounting for clustering within my dataset in a logistic regression analysis in Statistical Procedures

Accounting for clustering within my dataset in a logistic regression analysis

derbygun — Fri, 29 Apr 2022 02:21:55 GMT

Hi All,

I am currently working on an analysis with a binary outcome. I have a dataset in which some of the variables I want to include in the model were measured at the zipcode level. Meaning patients in the same zipcodes would have the same values. How do I account for this in my logistic regression model?

Thanks

Re: Accounting for clustering within my dataset in a logistic regression analysis

StatDave — Fri, 29 Apr 2022 02:31:00 GMT

Fit the logistic model in PROC GEE. See the Generalized Estimating Equations example in the Getting Started section of the GEE documentation.

Re: Accounting for clustering within my dataset in a logistic regression analysis

derbygun — Fri, 29 Apr 2022 03:32:03 GMT

Okay, thank you. I wanted to ask what the repeated measure would be in this case. I am confused if it would be the zip code or the person_id. in my dataset, patients in the same zipcodes were assigned the same values for the variables I am interested in