Testing for the Conditional Heteroskedasticity and Serial Correlation appears to involve correlation of the residuals. Does anybody know what I have to regress these residuals against in order to derive that residual correlation?
Thanks
Testing for the Conditional Heteroskedasticity and Serial Correlation appears to involve correlation of the residuals. Does anybody know what I have to regress these residuals against in order to derive that residual correlation?
Thanks
Testing for Heteroskedasticity, the BP test requires a regression of the Squared Residual (Yi-Ŷ1)2 on the independent variable(X) Decision Rule Heteroskedasticity exists: If X explains the variation in the squared residual. Heteroskedasticity doesn’t exist:X doesn’t explain the variation. BP test for Heteroskedasticity: χ2= nR2 Someone please correct me if I am wrong.