Announcement

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

  • Adjusting standard errors for multiple hypothesis testing

    Good morning everybody.
    I am running an analysis in which I test the effect of education on a large number of outcomes thus I would like, as a robustness check, to adjust standard errors for multiple hypothesis testing as done in Erten, Keskin (2016) following Simes (1986), i.e. using a Bonferroni procedure.
    If I have been clear, I am looking for a command or a routine to implement that kind of correction in Stata. Does anybody knows something about that?
    Thank you
    Last edited by Alessandro Sovera; 28 Apr 2017, 08:41.

  • #2
    .
    Last edited by Alessandro Sovera; 28 Apr 2017, 08:46.

    Comment


    • #3
      Alessandro: Check out "help test" with the mtest(...) option.

      Comment


      • #4
        Originally posted by John Mullahy View Post
        Alessandro: Check out "help test" with the mtest(...) option.
        Nice, thanks. Using this option I will obtain the corrected p-values for the test I need. Thus, if I want to present both the unadjusted p-value and the corrected one in a regression output table, I just have to run
        Code:
        test varlist, mtest(bonferroni)
        for all the coefficient I have and the result would be the p-value of a significance test on the specified coefficient. Am I right?

        Comment

        Working...
        X