Announcement

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

  • Lincom test Weibull analysis with dummyXdummy interaction

    Dear all,
    I searched for the forum for a solution to my problem and tried some fixes already, the issue is still there.
    I use Stata 16 and model my data with Weibull survival for dyads a-b. My keep independent variables are two dummies: dummy1 and dummy2.

    I generate my main regression results with the following command:
    streg i.dummy1##i.dummy2, d(weibull) coeflegend
    and run
    margins i.dummy1##i.dummy2 after that.

    I am interested in comparing the coefficients of the interaction (for combinations such as dummy1= 0 and dummy2=1 and vice versa).

    After typing in:

    . version 15: lincom 0.dummy1#.dummy2 - 0.dummy1#0.dummy

    The outputs is as follows:

    ( 1) - [_t]0b.dummy1#0b.dummy2 + [_t]0b.dummy1#1o.dummy2 = 0

    ------------------------------------------------------------------------------
    _t | Coef. Std. Err. z P>|z| [95% Conf. Interval]
    -------------+----------------------------------------------------------------
    (1) | 0 (omitted)
    ------------------------------------------------------------------------------

    I checked and it does not work for the Stata v.16 neither.

    I would appreciate your help in solving the issue.
    Agnieszka

  • #2
    streg i.dummy1##i.dummy2, d(weibull) coeflegend
    and run
    margins i.dummy1##i.dummy2 after that.
    If your lincom command is based on results from margins, you need to make these available using the -post- option.

    Code:
    margins i.dummy1##i.dummy2, post

    Otherwise, show your exact commands and the resulting Stata output (FAQ Advice #12).

    Comment


    • #3
      Dear Andrew,
      Thanks a lot for the reply. I guess it is the ", post" that I was missing. I just re-run it and it worked!! Thank you so much!
      Best
      Aga

      Comment

      Working...
      X