Is there a way of performing HAC robust standard errors in SPSS?

  • $\begingroup$ You mean a way of calculating HAC robust standard error estimates, right? Do you have a particular method you want to use? $\endgroup$ – guest Mar 22 '12 at 4:14
  • $\begingroup$ @guest - Yes! I'm using GEE Negative Binomial regression, controlling for temporal correlation on longitudinal count data. $\endgroup$ – Stephen Mar 22 '12 at 11:26
  • 1
    $\begingroup$ Use sandwich in R. Or robust option in Stata. A year and a half of waiting time is definitely worth the price of Stata :). $\endgroup$ – StasK Oct 22 '13 at 18:48

Since this question has not been answered for over a year and I stumbled upon an article with a macro for the simple HC correlation I thought it might be worth mentioning. The article is available here and has at the end a macro that allows for HC calculation. Using the R code from the sandwich-package vcovHAC (esp. the meatHAC) functions it might be feasible to create a macro for SPSS that handles HAC errors.

That said, why not just go with R and the sandwich-package? You can even work inside SPSS with R using the essentials for R-addon.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.