Multiple Linear Regression using SAS


In stead of just using engine displacement (displ) to predict highway mpg (hwy) in the simple linear regression example, the multiple regression here we will use the number of cylinders (cyl), city mpg (cty) and engine diplacement (displ) together to predict the highway mpg (hwy). The stb option on the model statement tells SAS to display the standardized regression coefficients. Note we treat cyl as interval variable here.

proc reg data = mpg;
  model hwy = displ cyl cty/ stb;

SAS multiple linear regression 1


The results indicate that the overall model is statistically significant (F = 821.93, p = <.0001).  Furthermore, all of the predictor variables are not statistically significant except for cty.

And SAS also provide plots for model assumption check.

SAS multiple linear regression 2


From the plots above, we think the model is valid.



See Also:

Do the same test in R

Leave a Reply

Your email address will not be published. Required fields are marked *