Gauss-Markov: Assumption no.5.

"Correlation between independent variables and the error term is 0".

How do i check this (in SPSS)?
I know this is basically the same as the demand for isolation (from causality) and by introducing control variables.. But how do i check if there is correlation between independent variables and the error term?

Can i do this by looking at the coefficients (Standardized Betas in SPSS) in the multiple regression analysis? If the control variable is low correlation, and not significant, while the "main" independent variable has a "ok" correlation (and significant) one can say that regression assumption five (Gauss-Markov) is met?


Ninja say what!?!
The most straightforward way I see is for you to set up the regression, have the model output the error terms, and run those error terms along with the independent variables in a correlation analysis. If the correlation coefficients are 0, then the assumption is met.

I don't know SPSS though. Maybe someone else can chime in.