Announcement

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

  • Deepening STATA knowledge. How to?

    Dear all,

    I have a silly question, considering the website where I am writing.

    I have been using STATA for years, but I have never had strong skills in Data Management (loops, macros, and so on).

    I wonder what the best way to improve my level would be. Courses? Specific books? I have been recommended "Data Management using STATA," second edition (2020). Do you know it?

    Additionally, an even sillier question: would you deepen your knowledge of only one tool (STATA) or would you advance your knowledge of multiple softwares in parallel (SAS, R...)?

    Please consider that I am a University Professor (Clinical Epidemiology) and I'mused to analyze observational studies. Moreover, I am often commissioned to analyze pharmacological clinical trials (where data management skills, Tables-Listing-Figures, etc., are necessary).

    Thank you very much for your support!

    Gianfranco

  • #2
    Various data management books are listed at

    https://www.stata.com/bookstore/data-management/

    the one you mention sounds very good to me.

    Your choice of a second language (if any) may depend on what other people in your field use. If I ever get ambitious I will probably try to get better with R. I occasionally get requests from people asking if there is an R or Python equivalent to some Stata program.

    But, I don’t feel too ambitious to learn a new program because Stata meets my needs very well, and it is also widely used by people I work with.

    if there is something you want to do, but Stata can’t do, then learn something that can do it. But there is a very good chance that Stata can do it, and you just have to learn how.
    -------------------------------------------
    Richard Williams, Notre Dame Dept of Sociology
    StataNow Version: 18.5 MP (2 processor)

    EMAIL: [email protected]
    WWW: https://www3.nd.edu/~rwilliam

    Comment


    • #3
      StataCorp offers free webinars that are good for a quick overview. It also offers netcourses. I liked netcourses 151.

      I have some materials for workshops I have given here: https://www.maartenbuis.nl/workshops/index.html
      ---------------------------------
      Maarten L. Buis
      University of Konstanz
      Department of history and sociology
      box 40
      78457 Konstanz
      Germany
      http://www.maartenbuis.nl
      ---------------------------------

      Comment


      • #4
        Thank you very much Maarten Buis and Richard Williams !

        Comment


        • #5
          I'd recommend a browse through the Data Management manual and past issues of the Stata Journal.

          Comment


          • #6
            I agree with Richard Williams when he says,
            if there is something you want to do, but Stata can’t do, then learn something that can do it. But there is a very good chance that Stata can do it, and you just have to learn how.
            From my own experience, I work with measurement models, often in a multilevel or longitudinal context. While Stata can handle these, I spent the time to learn Mplus, a program that is purpose built for structural equation modeling and can estimate all sorts of complex models that would be very difficult to estimate (or even do) in Stata. Another example is that I have taught graduate level multilevel modeling and many of my students wanted to use R because it was free and their department didn't pay for a Stata license. I used that as an occasion to teach myself R and am glad that I did that so; there are many incredible packages developed in R that have no equivalent in Stata (e.g., for network psychometrics).

            All that said, if Stata can do 95% of what you need, you can afford the license, and your clients and collaborators have no issues with you using Stata, then it is a far better use of your time to invest in learning things in Stata that you want to get better at.

            Comment


            • #7
              One of my Stata programs, xtdpdml, has an option to generate code for mPlus. I don’t know how mPlus does it, but it seems to work far more quickly than Stata, R, or any other program I am aware of.

              But, if I only had mPlus, I would have to learn how to do data manipulation, how to do post-estimation commands, etc. I would like for Stata sem to be faster, but for my needs it isn’t that big of a deal waiting a little longer for a program to run. But if I had more complex models, like Erik does, I would consider buying mPlus (which is fairly expensive) or see if R can do it better than Stata.

              I happen to be fortunate enough that other people get Stata for me! But if i wanted free and powerful, I would probably get R and a good book for learning it. But if i only needed free and basic, there are freebie SPSS clones and other simple packages around.

              Learning R is on my to do list. But it is on my list for AFTER i retire in a few years and need something to amuse myself!
              Last edited by Richard Williams; 29 May 2024, 17:17.
              -------------------------------------------
              Richard Williams, Notre Dame Dept of Sociology
              StataNow Version: 18.5 MP (2 processor)

              EMAIL: [email protected]
              WWW: https://www3.nd.edu/~rwilliam

              Comment


              • #8
                There is also a wealth of free online material, e.g.

                https://www.stata.com/links/resource...earning-stata/

                https://www.stata.com/links/

                Most my own stats and Stata notes are linked to at

                https://www3.nd.edu/~rwilliam/
                -------------------------------------------
                Richard Williams, Notre Dame Dept of Sociology
                StataNow Version: 18.5 MP (2 processor)

                EMAIL: [email protected]
                WWW: https://www3.nd.edu/~rwilliam

                Comment

                Working...
                X