Announcement

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

  • Hausman test/ IIA

    Good Morning all,
    I'm performing the IIA test for the first time. Obviously Im using a multinomial logit regression with 3 categories for the dependent variable. The default outcome base is 3. After estimating the parameters for the model, I perform the iia test with the command : mlogtest, iia base. I recieve an error message saying there is a problem determining number of categories. See the output below and I would be happy to recieve your advice. Thanks!

    mlogtest, iia base

    Problem determining number of categories.

    **** Hausman tests of IIA assumption

    Ho: Odds(Outcome-J vs Outcome-K) are independent of other alternatives.
    You used the old syntax of hausman. Click here to learn about the new syntax.

    (storing estimation results as _HAUSMAN)

    Omitted chi2 df P>chi2 evidence

    1 -8.648 6 1.000 for Ho
    2 -3.305 4 1.000 for Ho
    3 7.400 5 0.193 for Ho


    **** Small-Hsiao tests of IIA assumption

    Ho: Odds(Outcome-J vs Outcome-K) are independent of other alternatives.
    equation 3 not found
    r(303);
    Last edited by Benito D Isaac; 05 Aug 2015, 10:19.

  • #2
    I have performed the test with the command: hausman . , alleqs Here is the result I have this time:

    hausman all . , alleqs

    ---- Coefficients ----
    (b) (B) (b-B) sqrt(diag(V_b-V_B))
    all . Difference S.E.

    1
    sexe 15.99035 20.55081 -4.560464 804.3561
    influence~ee -1.273046 -1.147455 -.1255915 .
    influence~ge 1.045593 1.382695 -.3371018 .
    interet_in~e .1256635 -.2919401 .4176036 .
    domaine_ex~e .5012915 .412696 .0885955 .

    2
    sexe -1.413205 -1.455074 .0418687 .
    influence~ee -.5156961 -.4336147 -.0820814 .
    influence~ge .3465675 .7842892 -.4377217 .
    interet_in~e -.7119593 -.2991291 -.4128302 .
    domaine_ex~e .5681589 .7039288 -.1357699 .

    b = consistent under Ho and Ha; obtained from mlogit
    B = inconsistent under Ha, efficient under Ho; obtained from mlogit

    Test: Ho: difference in coefficients not systematic

    chi2(10) = (b-B)'[(V_b-V_B)^(-1)](b-B)
    = -6.82 chi2<0 ==> model fitted on these
    data fails to meet the asymptotic
    assumptions of the Hausman test;
    see suest for a generalized test


    Comment

    Working...
    X