Announcement

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

  • When is the cross-sectional data better than panel data?

    Dear all,

    When is the cross-sectional data better than panel data? That might be a weird and strange question, I personally think panel is the best choice but I still want to know if there is some expectations or unique value of using cross-sectional data. Say, in which situation do you prefer to use cross-sectional data than panel data (assume that both of them are available)?

    Best,
    David

  • #2
    David:
    if you have multiple waves of data (i.e. repeated measurements on the same units equally spaced stretching over a given timespan), the natural choice would be panel data analysis.
    However, there might be instances when a pooled OLS shoul be preferred vs -xtreg, fe- (see for more details on this topic the F test of ui = 0 at the foot of -xtreg, fe- outcome table, Example 2, -xtreg- entry, Stata .pdf manual).
    Kind regards,
    Carlo
    (Stata 19.0)

    Comment


    • #3
      Cross-posted at http://stats.stackexchange.com/quest...han-panel-data

      Please note our cross-posting policy http://www.statalist.org/forums/help#crossposting

      8. May I cross-post to other forums?

      People posting on Statalist may also post the same question on other listservers or in web forums. There is absolutely no rule against doing that.
      But if you do post elsewhere, we ask that you provide cross-references in URL form to searchable archives. That way, people interested in your question can quickly check what has been said elsewhere and avoid posting similar comments. Being open about cross-posting saves everyone time.
      If your question was answered well elsewhere, please post a cross-reference to that answer on Statalist.

      Comment


      • #4
        If you study something that does not change over time, than collecting a panel study is a huge waste of resources. Sometimes panel attrition is a big enough problem, to make them of little use.

        On a pragmatic level, there is a huge free rider problem when it comes to collecting panel data. It often takes many years before a panel dataset can be analyzed as a panel. An early career researcher would thus have a long time without publications or output that is recognized when hiring. That is obviously not good for her/him. A more senior researcher will have to worry less about that, but she/he will have to consider that she/he will probably retire before she/he can really use the data.

        However, I think the real answer is that no single study will give a definative answer a given question. Instead you need to think of a given research project as a small step in a much larger project. The researchers try to answer the same (or similar) question using different methods or under different circumstances. Each of these studies has their strengths and weaknesses. In many cases the strengths and weaknesses are complementary, i.e. the strength of one type of study is the weakness of another and vice versa. So after a while we will be able to summarize the different projects and hopefully get a coherent picture. So the aim should not be to device the perfect study but to device a study that adds something to the larger debate. So it is not about which type of study is good or bad, but we need diversity of approaches in order to assess the robustness of our claims.
        ---------------------------------
        Maarten L. Buis
        University of Konstanz
        Department of history and sociology
        box 40
        78457 Konstanz
        Germany
        http://www.maartenbuis.nl
        ---------------------------------

        Comment


        • #5
          Originally posted by Nick Cox View Post
          Dear Nick,

          Thank you for the notice very much! I'll add the cross-reference next time (if any).

          Best,
          David

          Comment


          • #6
            Originally posted by Carlo Lazzaro View Post
            David:
            if you have multiple waves of data (i.e. repeated measurements on the same units equally spaced stretching over a given timespan), the natural choice would be panel data analysis.
            However, there might be instances when a pooled OLS shoul be preferred vs -xtreg, fe- (see for more details on this topic the F test of ui = 0 at the foot of -xtreg, fe- outcome table, Example 2, -xtreg- entry, Stata .pdf manual).
            Dear Carlo,

            Yes, you're right, since I'm unfamiliar with pooled OLS, I will learn it more. Thanks for the advice.

            Best,
            David

            Comment


            • #7
              Originally posted by Maarten Buis View Post
              If you study something that does not change over time, than collecting a panel study is a huge waste of resources. Sometimes panel attrition is a big enough problem, to make them of little use.

              On a pragmatic level, there is a huge free rider problem when it comes to collecting panel data. It often takes many years before a panel dataset can be analyzed as a panel. An early career researcher would thus have a long time without publications or output that is recognized when hiring. That is obviously not good for her/him. A more senior researcher will have to worry less about that, but she/he will have to consider that she/he will probably retire before she/he can really use the data.

              However, I think the real answer is that no single study will give a definative answer a given question. Instead you need to think of a given research project as a small step in a much larger project. The researchers try to answer the same (or similar) question using different methods or under different circumstances. Each of these studies has their strengths and weaknesses. In many cases the strengths and weaknesses are complementary, i.e. the strength of one type of study is the weakness of another and vice versa. So after a while we will be able to summarize the different projects and hopefully get a coherent picture. So the aim should not be to device the perfect study but to device a study that adds something to the larger debate. So it is not about which type of study is good or bad, but we need diversity of approaches in order to assess the robustness of our claims.
              Dear Maarten,

              I agree with you. Thanks for your inspiring answer.

              Best,
              David

              Comment

              Working...
              X