[求助]面板数据做出来R方太小

gregorian

11615

收藏 2008-03-20

偶是菜鸟,照着别人的方法做了自己的数据,结果如下,高手帮我看下有没有补救方法,谢谢啊!

急~~~~~~~~~

. xtset comm year
       panel variable: comm (strongly balanced)
        time variable: year, 2002 to 2006
                delta: 1 unit

.
. sum tc vi

    Variable |       Obs        Mean    Std. Dev.       Min        Max
-------------+--------------------------------------------------------
          tc |      4020    .6430636    .1557206   .1548404   .9941447
          vi |      4020    .4504518    .2914263   3.03e-07   .9997568

.
. reg tc vi

      Source |       SS       df       MS              Number of obs =    4020
-------------+------------------------------           F( 1, 4018) =   26.68
       Model | .642918735     1 .642918735           Prob > F      = 0.0000
    Residual | 96.8134192 4018 .024094928           R-squared     = 0.0066
-------------+------------------------------           Adj R-squared = 0.0063
       Total | 97.4563379 4019 .024248902           Root MSE      = .15523

.
. xi:xtreg tc vi i.year,fe
i.year _Iyear_2002-2006 (naturally coded; _Iyear_2002 omitted)

Fixed-effects (within) regression Number of obs = 4020
Group variable: comm Number of groups = 804

R-sq: within = 0.0043                         Obs per group: min =         5
       between = 0.0082                                        avg =       5.0
       overall = 0.0046                                        max =         5

F(5,3211) = 2.79
corr(u_i, Xb) = -0.0768 Prob > F = 0.0160

------------------------------------------------------------------------------
          tc |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
          vi | -.0039905   .0018901    -2.11   0.035    -.0076965   -.0002845
_Iyear_2003 |   .0005467   .0006965     0.78   0.433    -.0008189    .0019122
_Iyear_2004 |   .0014861   .0006968     2.13   0.033     .0001199    .0028524
_Iyear_2005 | -.0002699    .000699    -0.39   0.699    -.0016404    .0011006
_Iyear_2006 | -.0004994   .0007047    -0.71   0.479    -.0018811    .0008822
       _cons |   .6446084   .0010294   626.22   0.000     .6425901    .6466267
-------------+----------------------------------------------------------------
     sigma_u |   .1553958
     sigma_e |   .0139442
         rho | .99201223   (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0:     F(803, 3211) =   616.04           Prob > F = 0.0000

.
. xi:xtreg tc vi i.year,re
i.year _Iyear_2002-2006 (naturally coded; _Iyear_2002 omitted)

Random-effects GLS regression Number of obs = 4020
Group variable: comm Number of groups = 804

R-sq: within = 0.0043                         Obs per group: min =         5
       between = 0.0082                                        avg =       5.0
       overall = 0.0043                                        max =         5

Random effects u_i ~ Gaussian Wald chi2(5) = 12.97
corr(u_i, X) = 0 (assumed) Prob > chi2 = 0.0237

------------------------------------------------------------------------------
          tc |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
          vi | -.0035127   .0018837    -1.86   0.062    -.0072046    .0001793
_Iyear_2003 |   .0005561    .000697     0.80   0.425    -.0008101    .0019223
_Iyear_2004 |   .0014971   .0006974     2.15   0.032     .0001303     .002864
_Iyear_2005 | -.0002522   .0006995    -0.36   0.718    -.0016233    .0011188
_Iyear_2006 | -.0004707   .0007052    -0.67   0.504    -.0018529    .0009114
       _cons |   .6443798   .0055537   116.03   0.000     .6334948    .6552648
-------------+----------------------------------------------------------------
     sigma_u | .15463201
     sigma_e |   .0139442
         rho | .99193376   (fraction of variance due to u_i)
------------------------------------------------------------------------------

.
. xttest0

Breusch and Pagan Lagrangian multiplier test for random effects

tc[comm,t] = Xb + u[comm] + e[comm,t]

        Estimated results:
                         |       Var     sd = sqrt(Var)
                ---------+-----------------------------
                      tc |   .0242489       .1557206
                       e |   .0001944       .0139442
                       u |   .0239111        .154632

        Test:   Var(u) = 0
                              chi2(1) = 7885.58
                          Prob > chi2 =     0.0000

.
. qui xtreg tc vi ,fe

. estimates store fe

. qui xtreg tc vi ,re

. estimates store re

. hausman fe

                 ---- Coefficients ----
             |      (b)          (B)            (b-B)     sqrt(diag(V_b-V_B))
             |       fe           re         Difference          S.E.
-------------+----------------------------------------------------------------
          vi |    -.003554    -.0030925       -.0004616        .0001518
------------------------------------------------------------------------------
                           b = consistent under Ho and Ha; obtained from xtreg
            B = inconsistent under Ha, efficient under Ho; obtained from xtreg

Test: Ho: difference in coefficients not systematic

                  chi2(1) = (b-B)'[(V_b-V_B)^(-1)](b-B)
                          =        9.25
                Prob>chi2 =      0.0024

.
end of do-file