Announcement

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

  • Testing overidentification with categorical covariates

    Hi everyone,

    I am using Stata Version 14.2 and currently run a 2SLS Regression using the ivprobit command with a binary outcome, one continuous endogenous variable, categorical and continuos covariates and three binary instruments. Now I want to test for overidentification using the overid command., but Stata returns an error message

    ivprobit clin_dia (chil_n = iv_sex iv_girl iv_twin) age i.educat i.wealth_quintile i.relig i.area, twostep first
    overid

    0: operator invalid
    r(198);


    However, when indicating the covariates to be continuous, I do not receive an error message:

    ivprobit clin_dia (chil_n = iv_sex iv_girl iv_2girls iv_twin) age educat wealth_quintile relig area, twostep first
    overid


    Is there a possibility to test for overidentification in a model that includes categorical covariates?

    I would really appreciate some help on this! Thanks a lot in advance
    Felix

    p.s.: this is my first post here, so I apologise in case I did not apply all the advices on posting to Statalist correctly.

  • #2
    Welcome to the Stata Forum / Statalist,

    I have no experience with the user-written (KIt Baum's) overid but, by reading the help files, it seems it won't allow factor notation.
    Best regards,

    Marcos

    Comment


    • #3
      Thanks a lot for your response!

      Is there an alternative to the overid command that allows factor notation?

      Best regards
      Felix

      Comment


      • #4
        Felix:
        as an aside to Marcos' helpful reply, you should probably omit by hand the variable -o.-mitted by Stata.
        Kind regards,
        Carlo
        (Stata 19.0)

        Comment

        Working...
        X