Significance Analysis but Unclean Residuals
Posted: Thu Jul 18, 2013 9:26 am
I would like to remove statistically insignificant variables from my model’s equation but I could only do so if the residuals are clean ie. no autocorrelation and heteroskedasticity. Am I able to conduct the significance analysis if I will to use Newey West estimator to “clean up” my residuals? Thanks in advance for your response.