MultiGLM Macro

This macro assists you in building multivariable linear regression models and allow you to test associations of the explanatory variables with the outcome after accounting for other variables, confounders or effect modifiers.

You may download macro (sas) and the Implementation Code (sas) but we would strongly recommend you to work through the eLearning module before implementing the macro. If you are new to model building, also read a brief Introduction to model building or consult other Resources before you begin.

What it can do for you

The MultiGLM macro builds models using a forward, backward or stepwise approach (according to your specifications) and presents results in an Excel sheet to facilitate comparisons between models. For each model fitted, the output includes:

  • Overall ANOVA F-statistics and p-values.
  • Partial F-statistics and p-values.
  • R2 and adjusted R2.
  • Parameter estimates and standard errors of the study/exposure variable.

In addition, the macro can construct and test interaction terms and create a ready-to-publish table of final results.

Work through the eLearning module presented on the following pages to learn how to implement this macro.