Announcement

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

  • constructing a vector of variables in order to make panel estimation

    Hello everyone I have a set of indicators that represent my control variables I would like to create a vector of these variables in order to make estimates on panel I do not know how to do it thank you for your help

  • #2
    The command will do this for you. For a panel estimator, see

    Code:
    help xtreg

    Comment


    • #3
      Originally posted by Andrew Musau View Post
      The command will do this for you. For a panel estimator, see

      Code:
      help xtreg
      my problem is in constructing the vector of variable

      Comment


      • #4
        i don't really know how to do

        Comment


        • #5
          Do you want to create a variable from several indicators? Present a data example using dataex and explain what is needed, e.g., by copying and pasting the output from the following:

          Code:
          dataex in 1/20

          See FAQ Advice #12 for more details.

          Comment


          • #6
            hello!
            I would like to build indexes of the financial development from (by the method of principal component analysis if possible) I join here below my excel file the variables in green are those which I wish to use to build the first index on the financial development, the variables in yellow are those for the second index, the variables in orange are those for the control variables
            Attached Files

            Comment


            • #7
              See the PDF entry of pca in Stata

              Code:
              help pca
              or if you can get your hands on Alan Acock's textbook (linked), there is a chapter in there that deals with this topic.

              Comment


              • #8
                my dependent variable is I(1) what to know wich model i am going to choose

                Comment


                • #9
                  For a panel with \(N > T\) and the outcome variable is continuous, as I advised in #2, see

                  Code:
                  help xtreg
                  With a binary outcome variable

                  Code:
                  help xtlogit
                  With an ordinal outcome variable

                  Code:
                  help xtologit
                  or

                  Code:
                  search feologit
                  or if you have Stata 17 and want to estimate a multinomial logit model

                  Code:
                  help xtmlogit

                  Your posts are difficult to follow as it is not clear what you want to do. Specific questions are easier to answer, but hopefully the manual entries of the referenced commands can give you some ideas on how to proceed.

                  Comment


                  • #10
                    In fact I have panel data and after the unit root test I see that I have variables that are stationary in first difference including my dependent variable. Other variables are stationary in second and third difference so I don't know which model to choose either a VAR or a VEC or an ARDL. If not, what should I do? or I have to eliminate the variables that are I(2) I(3) I stay with the I(1) variables I do the cointegration test and I work with a VEC

                    Comment


                    • #11
                      If you have \(N>>T\), my advice in #9 still applies. With large \(T\), start a new thread with a more informative title and ask the question. Others will be better placed to help you.

                      Comment


                      • #12
                        Originally posted by Andrew Musau View Post
                        If you have \(N>>T\), my advice in #9 still applies. With large \(T\), start a new thread with a more informative title and ask the question. Others will be better placed to help you.
                        \(T>N\)

                        Comment

                        Working...
                        X