Problem | Effect | Solution |
---|---|---|
Heteroskedasticity: variance of error term is not constant. Test using BP test BP = nR |
F-test is unreliable. Standard error underestimated. t-stat overstated. |
Robust standard errors Generalized least squares |
Serial correlation: error terms are correlated. Test using DW stat. DW 2(1 – r) |
F-stat too high. Standard error underestimated. t-stat overstated. |
Hansen method Modify the regression equation |
Multicollinearity: two or more independent variables are highly correlated. | Inflated SE’s; t-stats of coefficients artificially small High R |
Omit one or more of the independent variables |