topic Re: Linear Regression in Statistical Procedures

Linear Regression

jlima22 — Sat, 23 Jan 2021 03:19:13 GMT

Part of my assignment reads as follows:

Your goal is to create a regression model that can predict the income based on age in years.

Create a scatter plot for age vs. income.
1. I believe I have done this.
Fit a linear regression model in SAS
How good is the overall fit of the income determination model? (Use both R2 as well as the F statistics to justify your answer.)
Compute the predicted income for a 27-year-old individual.

The second part of the assignment is giving me a hard time. I'm using the Linear Regression task but I'm confused as to how to set it up. Would age or income be the dependent variable? Would the remaining variable then be used as the classification variable/intercept?

Any help would be greatly appreciated!

Re: Linear Regression

mkeintz — Sat, 23 Jan 2021 04:05:46 GMT

predict the income based on age

One of these variables is the predictor, one the "predictee". Which is which should be evident from the quoted phrase above.

Google "dependent variable" and you will likely have started your road to understanding regression.

Re: Linear Regression

PaigeMiller — Sat, 23 Jan 2021 12:49:36 GMT

In addition to what @mkeintz said, there are no CLASS variables here, there are only continuous variables, so if you include CLASS variables in your model, you get gibberish. So please look up what a CLASS variable is and is not.

Re: Linear Regression

STAT_Kathleen — Tue, 26 Jan 2021 21:55:32 GMT

The INCOME variable is the dependent variable and AGE is the independent variable. So if you want to use PROC REG to estimate a linear regression model. For example,
/* syntax for linear regression model, provide parameter estimates, creates output data set with predicted observations */
PROC REG DATA=YOURDATA;
MODEL INCOME=AGE;
OUTPUT OUT=OUT P=PREDINC;
QUIT;

/* provides output data set with the predicted observations */
PROC PRINT DATA=OUT;
RUN;