topic Re: Terminating Intercept in Regression Analysis in New SAS User

Terminating Intercept in Regression Analysis

sasgiaz — Sun, 24 Feb 2019 15:31:58 GMT

Hello. Does somebody happen to know what should be done if we're using regression analysis but the intercept is not significant?

If It should be terminated from the model, what syntax should we use?

Thank you for your kind help and attention.

Re: Terminating Intercept in Regression Analysis

VDD — Sun, 24 Feb 2019 16:11:37 GMT

would this help by using plots=fitplot

ods graphics on;
proc reg data=sashelp.class plots=fitplot;
model height=weight/noint;
run;
quit;

https://communities.sas.com/t5/SAS-Programming/regression-analysis-without-intercept/m-p/529460#M144669

Re: Terminating Intercept in Regression Analysis

PaigeMiller — Sun, 24 Feb 2019 16:29:28 GMT

I would leave it in the model. The model will predict better and fit better.

Re: Terminating Intercept in Regression Analysis

sasgiaz — Sun, 24 Feb 2019 17:22:45 GMT

It's done! Thankyou so much!

Re: Terminating Intercept in Regression Analysis

sasgiaz — Sun, 24 Feb 2019 17:23:46 GMT

Could you tell me why is that?

Re: Terminating Intercept in Regression Analysis

PaigeMiller — Sun, 24 Feb 2019 18:05:11 GMT

Draw a plot of the regression line with the intercept.

Draw another plot of the regression line with no intercept.

See which one fits better visually.

Re: Terminating Intercept in Regression Analysis

sasgiaz — Sun, 24 Feb 2019 18:05:41 GMT

Sorry if I bother you again.. But I found new problem.
I used to use the syntax like this:
model height=weight/influence;
and it has output statistics which has the value of model's residual, but I can't find it in the syntax you gave.
Could you tell me how to make the residual come out? Thanks before.