3

When using Newey-West robust standard errors to deal with heteroskedasticity and autocorrelation:

http://support.sas.com/kb/40/098.html

is it correct to state that the coefficients are not different between a robust and a non-robust estimation? Only the standard errors change, is that correct?

adrCoder
  • 473
  • 2
  • 9
  • 3
    Yes, as the name suggests, only the standard errors change. – Christoph Hanck Jan 28 '20 at 09:29
  • Thanks Christoph for confirming this. Some reference would also be nice, thanks. – adrCoder Jan 28 '20 at 09:39
  • 2
    See, e.g., the textbooks by Stock and Watson or Hayashi. Also, here are a couple of threads on CV which should be relevant: https://stats.stackexchange.com/questions/353922/eicker-huber-white-robust-variance-estimator https://stats.stackexchange.com/questions/139614/robust-ols-standard-errors-newey-west https://stats.stackexchange.com/questions/142289/newey-west-standard-errors-in-regression-model-without-constant https://stats.stackexchange.com/questions/286894/difference-between-hccme-and-wls-estimator/287122#287122 – Christoph Hanck Jan 28 '20 at 09:44
  • Many thanks Christoph, very helpful. – adrCoder Jan 28 '20 at 10:32

0 Answers0