Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • xtsktest - overwriting the input model

    Hi there,

    I am working with panel random effect models and I need to perform Galvao test to check the normality of residuals (even if I am aware that normality shouldn't be an issue as explained by Professor Wooldridge: https://www.statalist.org/forums/for...-effects-model).

    I found that for Galvao there is an useful command: xtsktest (https://www.econstor.eu/bitstream/10...las-wp-178.pdf). However, xtsktest command can only be deployed right after estimating the model by using "xtreg re".

    Here is my question: is it possible to overwrite the coefficients returned by xtreg command and then run xtsktest on my modified model?
    I need to do this because, unfortunately, my reference model is estimated in SAS and the coefficients returned by xtreg are slightly different. Sadly, using model coefficients estimated in SAS is a requirement that I have to comply with.

    Given such premises, any solution would be really appreciated.

    Thank you in advance!

    Edoardo

  • #2
    Cross-posted at https://stackoverflow.com/questions/...he-input-model

    Comment


    • #3
      HTML Code:
      https://www.stata.com/statalist/archive/2002-12/msg00054.html
      not sure what xtsktest will use. probably need to overwrite e(b) and e(V).

      Comment

      Working...
      X