Announcement

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

  • Help in Stata

    Hello All,
    I am new to Stata. I am having some issues which I would like to discuss with you all and get your valuable suggestions:
    I am having data in the format:
    Company_name Year Var1 Var2 Var3......
    ABC 2010
    ABC 2011
    ABC 2012
    ABC 2013
    ABC 2014
    DEF 2010
    DEF 2011
    DEF 2012
    DEF 2013
    DEF 2014

    and so on.....
    This is panel model...I am getting confused regarding which model to use and which will give me the most significant results. Till now I got the most significant results in random effects model. I would also like to know is it always necessary to introduce industry dummies?

  • #2
    Also what other panel data models I can use here apart from fixed and random effects model

    Comment


    • #3
      If your goal is the most significant results, then you should probably dispense with random effects etc. altogether, and treat the observations as independent.

      Comment


      • #4
        Can you please make it a little bot clear your point..I am sorry because I am new...so I am not able to understand....my question is more or less below:

        I am having some questions which I am sure you people will be able to answer and help me out...
        I am having a small panel data set of 124 companies spanned over 6 years(2010-2015)...i tested between pooled OLS and fixed effects and got pooled OLS to be the one that I should use....further did a test between pooled ols and random effects and got random effects as the answer.....
        Now when I am running both pooled OLS and random effects, I am getting significant results for most of my explanatory variables but as soon as I am entering industry dummies the significance disappears for my most important explanatory variable and also other explanatory variables apart from one or two.. i am not sure what is happening and what kind of model should I be using...Need your suggestions....

        Comment


        • #5
          Sujata: this is the thread of Statalist that specialises in Mata. There appears to be nothing in your question regarding Mata. Hence I recommend that you post your question in the General forum. I also recommend re-reading the forum FAQ in advance of re-posting and reformulating your query in a way that will maximize the chances of getting a helpful response.

          Edit: I now see that you have posted in General as well. But my advice (also in Carlo's post) still stands
          Last edited by Stephen Jenkins; 27 Sep 2015, 06:51.

          Comment

          Working...
          X