Announcement

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

  • Collinearity issue with panel dataset

    I have an unbalanced panel dataset.
    I am trying to run "xtrc" random coefficient model and am getting the following error that says that variables were dropped due to collinearity. I am also attaching the correlation table, which clearly shows that the correlation between revt_log (DV) and xrd_lag1_log (IV) is 0.2439 which is not high at all so the model should have run. So why am I getting the said error when running xtrc?
    [All variables are log-transformed, if that is somehow related to the issue. Also, I am not just getting this error with the IV -xrd_lag1_log but with all other variables when I run them individually. That means the error is not specific to a specific IV - xrd_lag1_log]

    Thanks in advance!

    Click image for larger version

Name:	output.jpg
Views:	1
Size:	69.0 KB
ID:	1364653

  • #2
    Please read the FAQ on questions and dataex.

    As I read the error message, xtrc dies if you drop observations due to colinearity. It looks like the colinearity is at the panel level - it says you need to be able to do ols on each panel without dropping variables due to colinearity. If you ran ols by panel, you'd find some panels that have dropped variables due to colinearity.

    Comment


    • #3
      Thank you for your reply. I have further questions-

      Q1) You said,"If you ran ols by panel, you'd find some panels that have dropped variables due to collinearity." <- could you please elaborate what you mean here? Is there a command within the xtrc program to do so? The Stata manual on xtrc program did not mention anything about doing ols by panels.

      Q2) Is there a way to identify which panels were dropped (being too small) as shown in the error output?

      Q3) Please suggest any resource (other than Stata manual) on panel data regression that I can use to learn more on that topic if you know of some good ones. I am trying to learn more on xtrc, xtreg, xtmixed programs and how to know which is a better fit.

      Thank you in advance!
      Last edited by cherry singhal; 18 Nov 2016, 12:16.

      Comment


      • #4
        Cherry.
        as far Q3) is concerned, you may want to take a look at http://www.stata.com/bookstore/econometrics/index.html.
        For Stata users dealing with panel data econometrics, the following one is a valuable textbook: http://www.stata.com/bookstore/micro...ata/index.html.
        Last but not least, Stata .pdf manual is, as Always, a good place to start from.
        Kind regards,
        Carlo
        (Stata 18.0 SE)

        Comment

        Working...
        X