Announcement

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

  • Help/Interpretation of test/results

    Hello to all
    I'm seeking some helpto check if my interpretation is correct, i'm sorry about the lenght of this post, and for any mistakes in my english i will probably make.
    My research is to check if the military expenditure has any impact with the GDP from 1980 to 2013 by using VAR or VEC.
    First I checked the stationarity of the varibles, GDP (PIB in Portuguese, in the pictures) and MilitaryExpenditure ( GastoMilitarDefaIPCA2013, in portuese and in the pictures, sorry for the long name), i used Dickey-Fuller for the test, both variables were non-stationary, them i did first difference on both variables and they become stationary, both stationary of first order. Some criterias AK and BIA also suggested that i make first difference in the cointegration test. After that i did the Johansen Test of cointegration, and the result was that GDP(PIB) and MilitaryExpenditure(GastoMilitarDefaIPCA2013) has long run relationship, being cointegrated of Rank (1), with the "Star". Them after running VEC i checked the Error Correction Term, them checking the p-valor it didnt show the casuality in the long run. Now is one of my doubt in the short run, i couldnt find out if there is relationship or not.
    Them moving on to check if the model is well specified, i did the Lagrange Multiplier test for autocorrelation and it didnt show autocorrelation, another doubt, at last i did the Jarque-Bera test for error distribution, and my target model, that is MilitaryExpenditure effectiong the GDP, is normaly distributed. Again there is screenshots to see if my interpretation is correct i also didnt know if i did all correctly or if i jumped major steps.
    I'm sorry if im asking to much, i'm really lost here on the interpretation. I don't even know if this is the right place.
    Thank you in advantage.


    WARNING - BIG PICUTRE AHEAD - (https://i.imgur.com/pLtv137.png)





  • #2
    Dreyfuss:
    welcome to this list.
    Please re-post your query on the General forum (this one is for practicing purposes only) and take a look at the FAQ before re-posting (you will read that screenshots are highly discouraged, whereas using CODE delimiters to share with interested listers what you typed and what Stata gave you back is strongly recommended). Thanks.
    Kind regards,
    Carlo
    (Stata 18.0 SE)

    Comment


    • #3
      Ah, I'm so sorry Carlo.
      I will read the FAQ and re-post following your tips.
      Thank you!

      Comment

      Working...
      X