Massachusetts Institute of Technology
Department of Urban Studies and Planning

11.220 Quantitative Reasoning and Statistical Methods for Planning

SPSS

11.220-HOME   |   TEST-OUT   |   BRUSH-UP   |   Brushup-DATA

SPSS > Regression

1. Now we'll build a model to predict income based on years of education. This model will be a simple bivariate linear regression. Go to Analyze->Regression and click "Linear...."

2. Set "incws" as "Dependent:" and "educ_yr" as "Independent(s)."

3. Check the result. An R2 of .141 means that only about 14% of the variance in income is attributable to years of education.

The F test is another measure of the utility of the model. It tests the Null Hypothesis that B1 = 0, like the t test below it.

In this model, every year of education adds about $3,349.15 to annual income. The t value associated with this amount is significant to a value of .015, also meaning that the B1 coefficient does not equal 0. So, years of education does contribute to predicting annual income, but the model still seems to be lacking some important predictors.

Created by Myounggu Kang on January 25, 2004. Edited by Rhonda Ryznar on January 20, 2005.