Announcement

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

  • xtpcse estimation with AR(1) option

    Dear all:

    I have a time-series cross-section dataset. When I use xtpcse with AR(1) option "c(a)", I get the following message. I understand that I get the message because there are gaps in my dataset (unbalanced), and the AR(1) option assumes balanced data. But what does it mean for the results that Stata's xtpcse produces? The results cannot be trusted? Or they can be trusted, although estimated rho may not be appropriate?

    I thank you in advance for your kind help.

    Best,

    Taka Sakamoto

    -----------------------------------------
    . xtpcse mfppwt l4.redisp l4.lfamtotppp l4.lsecondarygdppop l4.lalmptotppp l4.igerdgov l4.popgrow l4.lcgdpopc if year
    > <2009,c(a) p

    Number of gaps in sample: 28
    (note: computations for rho restarted at each gap)
    (note: estimates of rho outside [-1,1] bounded to be in the range [-1,1])
    (note: at least one disturbance covariance assumed 0, no common time periods
    between panels)

    Prais-Winsten regression, correlated panels corrected standard errors (PCSEs)

    Group variable: id Number of obs = 215
    Time variable: year Number of groups = 17
    Panels: correlated (unbalanced) Obs per group:
    Autocorrelation: common AR(1) min = 2
    Sigma computed by pairwise selection avg = 12.647059
    max = 22
    Estimated covariances = 153 R-squared = 0.1254
    Estimated autocorrelations = 1 Wald chi2(7) = 19.00
    Estimated coefficients = 8 Prob > chi2 = 0.0082

    ----------------------------------------------------------------------------------
    | Panel-corrected
    mfppwt | Coef. Std. Err. z P>|z| [95% Conf. Interval]
    -----------------+----------------------------------------------------------------
    redispov |
    L4. | -.0512275 .0269678 -1.90 0.057 -.1040834 .0016284
    |
    lfamtotppp |
    L4. | .3697768 .365409 1.01 0.312 -.3464116 1.085965
    |
    lsecondarygdppop |
    L4. | .1587017 .1843462 0.86 0.389 -.2026102 .5200137
    |
    lalmptotppp |
    L4. | .4643055 .3519623 1.32 0.187 -.2255279 1.154139
    |
    igerdgov2 |
    L4. | 1.657015 .7539443 2.20 0.028 .1793118 3.134719
    |
    popgrow |
    L4. | -.2262848 .3279123 -0.69 0.490 -.8689812 .4164116
    |
    lcgdpopc |
    L4. | -3.314369 1.592007 -2.08 0.037 -6.434645 -.1940926
    |
    _cons | 32.64075 13.82515 2.36 0.018 5.54396 59.73755
    -----------------+----------------------------------------------------------------
    rho | .4644425
    ----------------------------------------------------------------------------------


  • #2
    I'm afraid I can't help you much on the content of your posting. You might have better luck if you provided Stata code in code delimiters, and sample data using dataex.

    What I can tell you is that the documentation says the estimates depend on the estimates of the autocorrelation parameter so if rho is not appropriate, the parameters are probably not consistent.

    Comment


    • #3
      Thank you for your reply. Could you tell me what "Stata code in code delimiters"? I apologize my ignorance.

      Thank you!

      Taka

      Comment

      Working...
      X