Announcement

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

  • Step by Step Stata Panel

    Dear Stata Users,

    I am quite a newbie to Stata and panel data. Actually when I take a look at books, articles, PDFs, or any such materials, I just come across econometrics statistics or codes, anytime. All I exactly want to do is to know how to proceed. For instance, one of my friends advised me first to run panel unit roots (maybe panel cointegration also, I don't know) and multicollinearity & then go on with selecting the appropriate model, i.e. fixed, random, or pooled one. But it might be also required to use dynamic panel rather than static one. Then she told me after deciding on which method to use, I shall run tests for heteroscedasticity, autocorrelation and similar violations and improve my standard errors, accordingly. However, in my opinion, there must be an order like there is in OLS. Moreover, I also want to use some dummies in my panel analyses, such as a crisis dummy. Shall I create time dummies for all the years I use in my data or just one dummy is enough? If the former is the correct way, can I use the dummies for all fixed, random, and pooled analyses? Do I need also create dummies for the cases or does it depend on my preferences? Actually my data has 15 years and 8 cases.

    Could you please tell me or provide to data to get simple tricks how to proceed? Thanks in advance!

  • #2
    Welcome to the Stata Forum/ Statalist.

    As an old saying, first things first. My advice is to start by reading some parcels of the Stata Manual. They are free. What is more, you can access them directly from your Stata. Also, you may wish to take a look at the video tutorials at Stata YouTube Channel.

    To end, I assume that ‘Newbie Panel Data’ is not your real name. I also assume that you want to act accordingly in terms of following the FAQ advice. Being this so, please click on the ‘contact us’ button and inform your name as well as family name.

    Please read the FAQ. There’s great advice as well as information to be found there.

    Hopefully that helps.
    Last edited by Marcos Almeida; 12 Dec 2018, 10:34.
    Best regards,

    Marcos

    Comment


    • #3
      https://www.statalist.org/forums/help#realnames explains that we ask for full real names. Either you skipped reading of the FAQ Advice, or else decided that request doesn't apply to you, Your choice, and our choice whether to answer you too if you ignore requests that have been thought through and are explained.

      A much bigger deal than that is what are you asking. You are telling us that you're working with panel data and giving some details, but not much else. You are naturally aware of textbooks and courses but seem to expect that all the important stuff is what people are now going to explain in posts with a few minutes of their time. It doesn't work like that. As Marcos Almeida is gently hinting, you need to do the reading and then start doing it. Otherwise put, suppose I have a friend who does nuclear physics or microbiology and ask them to explain their field. They would humour me with a Mickey Mouse summary, but they aren't going to start explaining how to do it.

      This no doubt won't seem friendly and is certainly not what you are asking for. But look around the site and you'll find that precise and concise questions based on data and code often get very full answers, fast and free.

      Comment

      Working...
      X