Announcement

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

  • ivreg2h warning

    Hi,

    I am trying to run an IV regression with internal instruments only (as per Lewbel 2012). My equation is specified as:

    ivreg2h sf36mh (disab_psych=) retiree_type1 retiree_type3 age agesq ///
    female partnered fem_part fem_type ///
    household_size dependants ///
    swave4 swave5 swave6 swave7 swave8 swave9 swave10 swave11 swave12 swave13 swave14 ///
    swave15 swave16 swave17 swave18, rob first endogtest(disab_psych) fe

    I get a warning that my variables have been centered (and they are 'forever changed' in the data). Why is this happening and can I stop this?

    Thanks.

  • #2
    Dear Jane,

    I have the same question. I don't want the variable centered in ivreg2h as I need to compare the estimate with other regression. Have you found a way to use the package but without automatic centering?

    Sergey
    Kind regards,
    Sergey Alexeev | ​The University of Sydney
    https://alexeev.pw/

    Comment


    • #3
      Hi Sergey,

      The best I can do it to use preserve and restore around the ivreg2h command...

      Best,
      Jane.

      Comment

      Working...
      X