Announcement

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

  • Vuong Test

    Dear all,
    I am interested in how to perform Vuong (1989) test to choose between test for a twop part model using twopm command and a Heckit model (following -heckman). Could you please provide more information how to proceed in stata?
    thank you

  • #2
    Dear Sabrina,

    Depending on what you want to do with your model, the Vuong test may not be appropriate in this context. So, for us to be able to help you, you need to provide more information on what you are doing and why you are doing it.

    All the best,

    Joao

    Comment


    • #3
      Thank you for your request.
      I want to test which of the models is suitable for my case study. my model is the following y = transfer of money (we have a lot of zero because there are many pesons that send money) my explanatory variables are the means of transfers, income ...
      I used two models the first is the Hurdel model using twopm stata command is the second is a model of Heckman.
      I want to know which one is suitable for my study using the Vuong test. I do not know the procedure of this test on stata
      Sabrine

      Comment


      • #4
        How do you estimate the second part in the tpm?

        Joao

        Comment


        • #5
          I estimate using GLM.
          Sabrine

          Comment


          • #6
            Yes, but what model did you estimate using GLM?

            Comment


            • #7
              Thank you for your request,
              I estimate the second part using GLM the first part i use a probit model. I use also a heckman model and i want to know which one is suitable for my study using the Vuong test. And I do't know the procedure.
              All the best
              Sabrine

              Comment


              • #8
                Dear Sabrine,

                You still do not say what model you estimate with GLM, but my guess is that the Vuong test would not be appropriate in your context because you do not really care about the correct specification of the likelihood function.

                Best wishes,

                Joao

                Comment


                • #9
                  Dear Joao,
                  I estimate with GLM the amount of money remitted (transfer) my explanatory variables are the means of transfers, income ... the first part of my two part model is a probit model y=1 If the househols send money
                  All the Best
                  Sabrine

                  Comment

                  Working...
                  X